The globalization of technologicalinnovation denition and evidence

Daniele Archibugi and Simona IammarinoItalian National Research Council and University of Rome lsquoLa Sapienzarsquo

ABSTRACT

The concept of globalization of innovation is the zip between two funda-mental phenomena of modern economies the increased internationalintegration of economic activities and the raising importance of knowl-edge in economic processes The paper singles out three differentcomponents of the globalization of innovation (1) the internationalexploitation of nationally generated innovations (2) the global generationof innovations by MNEs and (3) global techno-scientic collaborationsEmpirical evidence on these three categories is here presented suggestingthat the relevance of global forces in innovation is rapidly increasingalthough at a different pace for each of the three ongoing processes

KEYWORDS

Technological change multinational corporations strategic technologyagreements

GLOBALIZATION AND INNOVATION

The notion of globalization of innovation similarly to that of nanceproduction culture and information is now diffuse Scholars govern-ments and international organizations have attempted to assess thechanges that have occurred in innovative activities due to an ever increas-ingly globalized societyGlobalization is not a single phenomenon but a catch-all concept to

describe a wide range of forces It has been dened very differentlyaccording to the social science within which it is applied Paul Streeten(1996) has half in jest collected the various denitions in the literatureHere we have applied a rather wide denition of globalization whichconforms to that provided by Giddens (1990 64) lsquothe intensication ofworld-wide social relations which link distant localities in such a waythat local happenings are shaped by events occurring many miles away

Review of International Political Economy 91 March 2002 98ndash122

123456789101112131415161718192021222324252627282930313233

341135363738394041424344

Folio Review of International Political EconomyISSN 0969-2290 printISSN 1466-4526 online copy 2002 Taylor amp Francis Ltd

httpwwwtandfcoukDOI 10108009692290110101126

and vice versarsquo Thus by lsquoglobalizationrsquo we mainly refer to a high (andincreasing) degree of interdependency and interrelatedness amongdifferent and geographically dispersed actors In principle thereforethere might be a higher globalization even with the same level of inter-nationalization (Archibugi and Iammarino 1998 Cantwell andIammarino 1998)The term is used to describe the phenomenon of lsquoglobalizationrsquo expe-

rienced by the world of invention and innovation Strictly speaking theeconomic application of new ideas and knowledge is not only lsquotechnicalrsquoinsofar as it can also be organizational managerial institutional Thenew lsquogeneral purposersquo technologies such as ICTs biotechnology newmaterials etc have been shown to intensify the sciencendashtechnology interface and to be inextricably associated with the complex processesof organizational institutional and infrastructural change (Freeman1994) In its most modest use ndash which is also the easiest to be recordedand thus quantied ndash the expression lsquoglobalization of innovationrsquo is shorthand for the increasing international scope of the generation and diffu-sion of technologies That technology in the sense of knowledge directedtowards the solution of specic human problems is transmitted fromone culture to another or from one society to another is certainly not anovelty Even though learning processes are long and cumbersome tech-nological knowledge transmission among peoples has met less resistancethan occurred in the cases of cultural religious social or political habitsTechnology has always constituted a fertile meeting place for differentsocieties If the assimilation and transfer of technology required lengthytime spans in the past today it takes place with a much higher inten-sity and speedNew technologies play a fundamental part in making globalization

possible Without aeroplanes telephones satellites computers and tele-visions it would not be possible to transfer information from one placeto another thus allowing for the speed and the intensity which charac-terize the modern world These give rise to a rate of diffusion and transferof knowledge which is greatly superior to that of the past In otherwords it was the new technologies that allowed the emergence of thelsquoglobal villagersquoThe coming of a society based on knowledge has proceeded hand in

hand with the enlargement of markets and the intensication ofexchange International trade and direct investments abroad havesubstantially increased thus rendering the national economic systemsincreasingly integrated with each other The pace of globalization andthat of technological change have in fact been strictly interrelated andfrom a long-term perspective it appears less important to establish whichone should be considered responsible for triggering the other rather thanto establish that they mutually enforced each other However to what

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 99

extent is the generation transfer and diffusion of innovations trans-formed by the globalization which they themselves facilitate Thepresumption here is that for many years there has been a circular processin which new technologies act as a lsquolubricantrsquo for economic and socialglobalization In turn globalization while facilitating the circulation ofpeople goods capital and above all ideas and knowledge allows forthe sustenance of a historically unprecedented rate of technologicalchange The concept of globalization of innovation thus comes to be thezip between the two fundamental phenomena of modern economies theincreased international integration of economic activities and the raisingimportance of knowledge in economic processesIn this paper we present some indicators on the empirical relevance

of the globalization of innovation among the most developed countriesUltimately the question we ask is does the globalization of innovativeactivities exist and to what extent We thought it appropriate to referto a previously outlined taxonomy of the globalization of innovation(Archibugi and Michie 1995) since we are convinced that this taxonomyis a useful lter through which to interpret the phenomenon consideredhere

A TAXONOMY OF THE GLOBALIZATION OF INNOVATION

Three main categories of the globalization of innovation were identied(Archibugi and Michie 1995 1997a) (1) the international exploitation oftechnology produced on a national basis (2) the global generation ofinnovations and (3) global technological collaborations The three cate-gories are complementary and not mutually exclusive both at rm andcountry level Firms especially large ones generate innovations in alldifferent ways described here From a historical point of view thesecategories emerged in three successive stages even though the secondand the third added to rather than substituted for the oldest one Thecategories of this taxonomy are contained in Figure 1

The international exploitation of technology produced on a national basis

The rst category includes innovatorsrsquo attempts to obtain economicadvantages through the exploitation of their own technological compe-tence in markets other than the domestic one We have preferred to labelthis category as lsquointernationalrsquo rather than lsquoglobalrsquo as the actors intro-ducing the innovations preserve in the main their national identity evenwhen the innovations are diffused and sold in multiple countries or thenecessary knowledge has been sourced elsewhere Clearly the distinction

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 100

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

1111011121314151617

101

Figure 1 A taxonomy of the globalization of innovation

Source Elaboration on Archibugi and Michie (1995)

Categories

International exploitationof nationally producedinnovations

Global generation ofinnovations

Global techno-scientificcollaborations

Actors

Profit-seeking firmsandindividuals

Multinationalfirms

Universitiesand publicresearchcentres

National andmultinationalfirms

Forms

Exports of innovative goods

Cession of licenses andpatents

Foreign production of innovative goods internallydesigned and developed

RampD and innovative activitiesboth in the home and the hostcountries

Acquisitions of existing RampDlaboratories or green-field RampDinvestment in host countries

Joint scientific projects

Scientific exchangessabbatical years

International flows of students

Joint-ventures for specificinnovative projects

Productive agreements withexchange of technicalinformation andor equipment

between lsquointernationalrsquo and lsquoglobalrsquo becomes rather blurred when takinginto account the huge intra-rm share of international trade and theincreasing relevance of Global Production Networks However this con-rms once more the differentiation between lsquointernationalizedrsquo activities(carried out in more than one country) and lsquoglobalizedrsquo processes (inter-dependent and integrated across space)Firms have incentives to expand their market range but their prod-

ucts might be unwelcome in host countries Innovative products areoften admitted into importing countries in the temporary absence ofsatisfactory internal surrogates when they represent radical advancesfor example at the beginning of the 1960s many countries did their bestto import the rst computers Such innovative products do not competewith those of local rms in the short run A non-hostile receipt of inno-vative products is all the more likely the more similar the income leveland the closer the commercial integration between the innovating countryand the importing country Both the exporting economy and theimporting one have an interest in the exchange of products (startingfrom those with a higher innovative content) if such an exchange occurswithin a framework of comparative advantages and for products withsimilar technological intensity However as soon as the rms of twocountries are able to produce similar products competition tends to befar ercer than that typically encountered for traditional products asinnovative goods are deemed to be of strategic importance (cf Pianta1988 Tyson 1992 Scherer 1992) Competitive struggles today involvesemiconductors and aeroplanes much more than corn wine and pota-toes Besides it is easily predictable that international rivalry will involvemore and more technology-intensive eldsInternational trade is not the only way through which an innovative

rm can benet from its technological competence it is possible that theinnovator nds it more advantageous to sell the innovation disembodiedie to licence it to foreign rms This strategy is all the more convenientwhen there are various types of obstacles to international trade as forexample in the case of (1) high transportation costs (2) barriers toimports (3) high wage differentials between the innovating country andthe importing country which would render the cost of the new producttoo high for the income level of the imitating country However it isnot always possible to licence technology to third countries In order fora market for disembodied technology to exist such a technology shouldbe of a codied nature and the acquiring country should have anadequate capacity to absorb it (Bell and Pavill 1997) To be effective thetransfer of technology especially from North to South needs more strin-gent co-operation forms such as those described in the third category ofthe taxonomy (technological collaborations)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 102

Another signicant way of exploiting innovations in foreign marketsis through foreign direct investment (FDI) The conditions allowinginternational production are known availability of capital a willingnessto geographically exploit ownership technological and organizationaladvantages are required on behalf of the investing rm (Dunning 1993)Economic and institutional stability and a sufcient level of economicdevelopment or in other words location advantages are required onbehalf of the host country It should be remembered that this rst cate-gory only includes the productive activity operated in host countrieswhich does not entail the creation of additional local technologicalcapacity if this were to be the case we would be moving from the rstto the second category of this taxonomy

The global generation of innovations

The second category is the global generation of innovations whichincludes innovations conceived on a global scale from the moment theyare generated Only innovations created by multinational enterprises(MNEs) are contained in this category With very few exceptions (suchas Shell and Unilever) it is easy to identify the country of origin of suchcompanies so much that to some they appear as national enterpriseswith multinational operations (Hu 1992)MNEs have often their own internal innovative network with units

based in different countries An efcient management of these geograph-ically dispersed RampD and technical centres would imply that thesecentres do not simply provide inputs to the local production units butthat they are integrated into the overall innovative strategy of the MNEsA substantial body of theoretical and empirical research has investigatedhow companies organize their internal innovative centres (Howells 1997Pearce and Singh 1992 Florida 1997 Grandstrand et al 1992 Zander1999) Bartlett and Ghoshal (1990) have singled out three main strate-gies which can be implemented by MNEs whose signicance variesacross countries industries and companies

Centre-for-globalThis is the traditional lsquooctopusrsquo view of the multinational corporation asingle lsquobrainrsquo located within the company headquarters concentrates thestrategic resources (top management planning and technological exper-tise) and distributes impulses to the lsquotentaclesrsquo (that is the subsidiaries)scattered across host countries Even when some RampD is undertakenabroad this is basically concerned with adapting products to the needsof the local users

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 103

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

extent is the generation transfer and diffusion of innovations trans-formed by the globalization which they themselves facilitate Thepresumption here is that for many years there has been a circular processin which new technologies act as a lsquolubricantrsquo for economic and socialglobalization In turn globalization while facilitating the circulation ofpeople goods capital and above all ideas and knowledge allows forthe sustenance of a historically unprecedented rate of technologicalchange The concept of globalization of innovation thus comes to be thezip between the two fundamental phenomena of modern economies theincreased international integration of economic activities and the raisingimportance of knowledge in economic processesIn this paper we present some indicators on the empirical relevance

of the globalization of innovation among the most developed countriesUltimately the question we ask is does the globalization of innovativeactivities exist and to what extent We thought it appropriate to referto a previously outlined taxonomy of the globalization of innovation(Archibugi and Michie 1995) since we are convinced that this taxonomyis a useful lter through which to interpret the phenomenon consideredhere

A TAXONOMY OF THE GLOBALIZATION OF INNOVATION

Three main categories of the globalization of innovation were identied(Archibugi and Michie 1995 1997a) (1) the international exploitation oftechnology produced on a national basis (2) the global generation ofinnovations and (3) global technological collaborations The three cate-gories are complementary and not mutually exclusive both at rm andcountry level Firms especially large ones generate innovations in alldifferent ways described here From a historical point of view thesecategories emerged in three successive stages even though the secondand the third added to rather than substituted for the oldest one Thecategories of this taxonomy are contained in Figure 1

The international exploitation of technology produced on a national basis

The rst category includes innovatorsrsquo attempts to obtain economicadvantages through the exploitation of their own technological compe-tence in markets other than the domestic one We have preferred to labelthis category as lsquointernationalrsquo rather than lsquoglobalrsquo as the actors intro-ducing the innovations preserve in the main their national identity evenwhen the innovations are diffused and sold in multiple countries or thenecessary knowledge has been sourced elsewhere Clearly the distinction

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 100

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

1111011121314151617

101

Figure 1 A taxonomy of the globalization of innovation

Source Elaboration on Archibugi and Michie (1995)

Categories

International exploitationof nationally producedinnovations

Global generation ofinnovations

Global techno-scientificcollaborations

Actors

Profit-seeking firmsandindividuals

Multinationalfirms

Universitiesand publicresearchcentres

National andmultinationalfirms

Forms

Exports of innovative goods

Cession of licenses andpatents

Foreign production of innovative goods internallydesigned and developed

RampD and innovative activitiesboth in the home and the hostcountries

Acquisitions of existing RampDlaboratories or green-field RampDinvestment in host countries

Joint scientific projects

Scientific exchangessabbatical years

International flows of students

Joint-ventures for specificinnovative projects

Productive agreements withexchange of technicalinformation andor equipment

between lsquointernationalrsquo and lsquoglobalrsquo becomes rather blurred when takinginto account the huge intra-rm share of international trade and theincreasing relevance of Global Production Networks However this con-rms once more the differentiation between lsquointernationalizedrsquo activities(carried out in more than one country) and lsquoglobalizedrsquo processes (inter-dependent and integrated across space)Firms have incentives to expand their market range but their prod-

ucts might be unwelcome in host countries Innovative products areoften admitted into importing countries in the temporary absence ofsatisfactory internal surrogates when they represent radical advancesfor example at the beginning of the 1960s many countries did their bestto import the rst computers Such innovative products do not competewith those of local rms in the short run A non-hostile receipt of inno-vative products is all the more likely the more similar the income leveland the closer the commercial integration between the innovating countryand the importing country Both the exporting economy and theimporting one have an interest in the exchange of products (startingfrom those with a higher innovative content) if such an exchange occurswithin a framework of comparative advantages and for products withsimilar technological intensity However as soon as the rms of twocountries are able to produce similar products competition tends to befar ercer than that typically encountered for traditional products asinnovative goods are deemed to be of strategic importance (cf Pianta1988 Tyson 1992 Scherer 1992) Competitive struggles today involvesemiconductors and aeroplanes much more than corn wine and pota-toes Besides it is easily predictable that international rivalry will involvemore and more technology-intensive eldsInternational trade is not the only way through which an innovative

rm can benet from its technological competence it is possible that theinnovator nds it more advantageous to sell the innovation disembodiedie to licence it to foreign rms This strategy is all the more convenientwhen there are various types of obstacles to international trade as forexample in the case of (1) high transportation costs (2) barriers toimports (3) high wage differentials between the innovating country andthe importing country which would render the cost of the new producttoo high for the income level of the imitating country However it isnot always possible to licence technology to third countries In order fora market for disembodied technology to exist such a technology shouldbe of a codied nature and the acquiring country should have anadequate capacity to absorb it (Bell and Pavill 1997) To be effective thetransfer of technology especially from North to South needs more strin-gent co-operation forms such as those described in the third category ofthe taxonomy (technological collaborations)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 102

Another signicant way of exploiting innovations in foreign marketsis through foreign direct investment (FDI) The conditions allowinginternational production are known availability of capital a willingnessto geographically exploit ownership technological and organizationaladvantages are required on behalf of the investing rm (Dunning 1993)Economic and institutional stability and a sufcient level of economicdevelopment or in other words location advantages are required onbehalf of the host country It should be remembered that this rst cate-gory only includes the productive activity operated in host countrieswhich does not entail the creation of additional local technologicalcapacity if this were to be the case we would be moving from the rstto the second category of this taxonomy

The global generation of innovations

The second category is the global generation of innovations whichincludes innovations conceived on a global scale from the moment theyare generated Only innovations created by multinational enterprises(MNEs) are contained in this category With very few exceptions (suchas Shell and Unilever) it is easy to identify the country of origin of suchcompanies so much that to some they appear as national enterpriseswith multinational operations (Hu 1992)MNEs have often their own internal innovative network with units

based in different countries An efcient management of these geograph-ically dispersed RampD and technical centres would imply that thesecentres do not simply provide inputs to the local production units butthat they are integrated into the overall innovative strategy of the MNEsA substantial body of theoretical and empirical research has investigatedhow companies organize their internal innovative centres (Howells 1997Pearce and Singh 1992 Florida 1997 Grandstrand et al 1992 Zander1999) Bartlett and Ghoshal (1990) have singled out three main strate-gies which can be implemented by MNEs whose signicance variesacross countries industries and companies

Centre-for-globalThis is the traditional lsquooctopusrsquo view of the multinational corporation asingle lsquobrainrsquo located within the company headquarters concentrates thestrategic resources (top management planning and technological exper-tise) and distributes impulses to the lsquotentaclesrsquo (that is the subsidiaries)scattered across host countries Even when some RampD is undertakenabroad this is basically concerned with adapting products to the needsof the local users

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 103

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

1111011121314151617

101

Figure 1 A taxonomy of the globalization of innovation

Source Elaboration on Archibugi and Michie (1995)

Categories

International exploitationof nationally producedinnovations

Global generation ofinnovations

Global techno-scientificcollaborations

Actors

Profit-seeking firmsandindividuals

Multinationalfirms

Universitiesand publicresearchcentres

National andmultinationalfirms

Forms

Exports of innovative goods

Cession of licenses andpatents

Foreign production of innovative goods internallydesigned and developed

RampD and innovative activitiesboth in the home and the hostcountries

Acquisitions of existing RampDlaboratories or green-field RampDinvestment in host countries

Joint scientific projects

Scientific exchangessabbatical years

International flows of students

Joint-ventures for specificinnovative projects

Productive agreements withexchange of technicalinformation andor equipment

between lsquointernationalrsquo and lsquoglobalrsquo becomes rather blurred when takinginto account the huge intra-rm share of international trade and theincreasing relevance of Global Production Networks However this con-rms once more the differentiation between lsquointernationalizedrsquo activities(carried out in more than one country) and lsquoglobalizedrsquo processes (inter-dependent and integrated across space)Firms have incentives to expand their market range but their prod-

ucts might be unwelcome in host countries Innovative products areoften admitted into importing countries in the temporary absence ofsatisfactory internal surrogates when they represent radical advancesfor example at the beginning of the 1960s many countries did their bestto import the rst computers Such innovative products do not competewith those of local rms in the short run A non-hostile receipt of inno-vative products is all the more likely the more similar the income leveland the closer the commercial integration between the innovating countryand the importing country Both the exporting economy and theimporting one have an interest in the exchange of products (startingfrom those with a higher innovative content) if such an exchange occurswithin a framework of comparative advantages and for products withsimilar technological intensity However as soon as the rms of twocountries are able to produce similar products competition tends to befar ercer than that typically encountered for traditional products asinnovative goods are deemed to be of strategic importance (cf Pianta1988 Tyson 1992 Scherer 1992) Competitive struggles today involvesemiconductors and aeroplanes much more than corn wine and pota-toes Besides it is easily predictable that international rivalry will involvemore and more technology-intensive eldsInternational trade is not the only way through which an innovative

rm can benet from its technological competence it is possible that theinnovator nds it more advantageous to sell the innovation disembodiedie to licence it to foreign rms This strategy is all the more convenientwhen there are various types of obstacles to international trade as forexample in the case of (1) high transportation costs (2) barriers toimports (3) high wage differentials between the innovating country andthe importing country which would render the cost of the new producttoo high for the income level of the imitating country However it isnot always possible to licence technology to third countries In order fora market for disembodied technology to exist such a technology shouldbe of a codied nature and the acquiring country should have anadequate capacity to absorb it (Bell and Pavill 1997) To be effective thetransfer of technology especially from North to South needs more strin-gent co-operation forms such as those described in the third category ofthe taxonomy (technological collaborations)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 102

Another signicant way of exploiting innovations in foreign marketsis through foreign direct investment (FDI) The conditions allowinginternational production are known availability of capital a willingnessto geographically exploit ownership technological and organizationaladvantages are required on behalf of the investing rm (Dunning 1993)Economic and institutional stability and a sufcient level of economicdevelopment or in other words location advantages are required onbehalf of the host country It should be remembered that this rst cate-gory only includes the productive activity operated in host countrieswhich does not entail the creation of additional local technologicalcapacity if this were to be the case we would be moving from the rstto the second category of this taxonomy

The global generation of innovations

The second category is the global generation of innovations whichincludes innovations conceived on a global scale from the moment theyare generated Only innovations created by multinational enterprises(MNEs) are contained in this category With very few exceptions (suchas Shell and Unilever) it is easy to identify the country of origin of suchcompanies so much that to some they appear as national enterpriseswith multinational operations (Hu 1992)MNEs have often their own internal innovative network with units

based in different countries An efcient management of these geograph-ically dispersed RampD and technical centres would imply that thesecentres do not simply provide inputs to the local production units butthat they are integrated into the overall innovative strategy of the MNEsA substantial body of theoretical and empirical research has investigatedhow companies organize their internal innovative centres (Howells 1997Pearce and Singh 1992 Florida 1997 Grandstrand et al 1992 Zander1999) Bartlett and Ghoshal (1990) have singled out three main strate-gies which can be implemented by MNEs whose signicance variesacross countries industries and companies

Centre-for-globalThis is the traditional lsquooctopusrsquo view of the multinational corporation asingle lsquobrainrsquo located within the company headquarters concentrates thestrategic resources (top management planning and technological exper-tise) and distributes impulses to the lsquotentaclesrsquo (that is the subsidiaries)scattered across host countries Even when some RampD is undertakenabroad this is basically concerned with adapting products to the needsof the local users

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 103

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

between lsquointernationalrsquo and lsquoglobalrsquo becomes rather blurred when takinginto account the huge intra-rm share of international trade and theincreasing relevance of Global Production Networks However this con-rms once more the differentiation between lsquointernationalizedrsquo activities(carried out in more than one country) and lsquoglobalizedrsquo processes (inter-dependent and integrated across space)Firms have incentives to expand their market range but their prod-

ucts might be unwelcome in host countries Innovative products areoften admitted into importing countries in the temporary absence ofsatisfactory internal surrogates when they represent radical advancesfor example at the beginning of the 1960s many countries did their bestto import the rst computers Such innovative products do not competewith those of local rms in the short run A non-hostile receipt of inno-vative products is all the more likely the more similar the income leveland the closer the commercial integration between the innovating countryand the importing country Both the exporting economy and theimporting one have an interest in the exchange of products (startingfrom those with a higher innovative content) if such an exchange occurswithin a framework of comparative advantages and for products withsimilar technological intensity However as soon as the rms of twocountries are able to produce similar products competition tends to befar ercer than that typically encountered for traditional products asinnovative goods are deemed to be of strategic importance (cf Pianta1988 Tyson 1992 Scherer 1992) Competitive struggles today involvesemiconductors and aeroplanes much more than corn wine and pota-toes Besides it is easily predictable that international rivalry will involvemore and more technology-intensive eldsInternational trade is not the only way through which an innovative

rm can benet from its technological competence it is possible that theinnovator nds it more advantageous to sell the innovation disembodiedie to licence it to foreign rms This strategy is all the more convenientwhen there are various types of obstacles to international trade as forexample in the case of (1) high transportation costs (2) barriers toimports (3) high wage differentials between the innovating country andthe importing country which would render the cost of the new producttoo high for the income level of the imitating country However it isnot always possible to licence technology to third countries In order fora market for disembodied technology to exist such a technology shouldbe of a codied nature and the acquiring country should have anadequate capacity to absorb it (Bell and Pavill 1997) To be effective thetransfer of technology especially from North to South needs more strin-gent co-operation forms such as those described in the third category ofthe taxonomy (technological collaborations)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 102

Another signicant way of exploiting innovations in foreign marketsis through foreign direct investment (FDI) The conditions allowinginternational production are known availability of capital a willingnessto geographically exploit ownership technological and organizationaladvantages are required on behalf of the investing rm (Dunning 1993)Economic and institutional stability and a sufcient level of economicdevelopment or in other words location advantages are required onbehalf of the host country It should be remembered that this rst cate-gory only includes the productive activity operated in host countrieswhich does not entail the creation of additional local technologicalcapacity if this were to be the case we would be moving from the rstto the second category of this taxonomy

The global generation of innovations

The second category is the global generation of innovations whichincludes innovations conceived on a global scale from the moment theyare generated Only innovations created by multinational enterprises(MNEs) are contained in this category With very few exceptions (suchas Shell and Unilever) it is easy to identify the country of origin of suchcompanies so much that to some they appear as national enterpriseswith multinational operations (Hu 1992)MNEs have often their own internal innovative network with units

based in different countries An efcient management of these geograph-ically dispersed RampD and technical centres would imply that thesecentres do not simply provide inputs to the local production units butthat they are integrated into the overall innovative strategy of the MNEsA substantial body of theoretical and empirical research has investigatedhow companies organize their internal innovative centres (Howells 1997Pearce and Singh 1992 Florida 1997 Grandstrand et al 1992 Zander1999) Bartlett and Ghoshal (1990) have singled out three main strate-gies which can be implemented by MNEs whose signicance variesacross countries industries and companies

Centre-for-globalThis is the traditional lsquooctopusrsquo view of the multinational corporation asingle lsquobrainrsquo located within the company headquarters concentrates thestrategic resources (top management planning and technological exper-tise) and distributes impulses to the lsquotentaclesrsquo (that is the subsidiaries)scattered across host countries Even when some RampD is undertakenabroad this is basically concerned with adapting products to the needsof the local users

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 103

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

Another signicant way of exploiting innovations in foreign marketsis through foreign direct investment (FDI) The conditions allowinginternational production are known availability of capital a willingnessto geographically exploit ownership technological and organizationaladvantages are required on behalf of the investing rm (Dunning 1993)Economic and institutional stability and a sufcient level of economicdevelopment or in other words location advantages are required onbehalf of the host country It should be remembered that this rst cate-gory only includes the productive activity operated in host countrieswhich does not entail the creation of additional local technologicalcapacity if this were to be the case we would be moving from the rstto the second category of this taxonomy

The global generation of innovations

The second category is the global generation of innovations whichincludes innovations conceived on a global scale from the moment theyare generated Only innovations created by multinational enterprises(MNEs) are contained in this category With very few exceptions (suchas Shell and Unilever) it is easy to identify the country of origin of suchcompanies so much that to some they appear as national enterpriseswith multinational operations (Hu 1992)MNEs have often their own internal innovative network with units

based in different countries An efcient management of these geograph-ically dispersed RampD and technical centres would imply that thesecentres do not simply provide inputs to the local production units butthat they are integrated into the overall innovative strategy of the MNEsA substantial body of theoretical and empirical research has investigatedhow companies organize their internal innovative centres (Howells 1997Pearce and Singh 1992 Florida 1997 Grandstrand et al 1992 Zander1999) Bartlett and Ghoshal (1990) have singled out three main strate-gies which can be implemented by MNEs whose signicance variesacross countries industries and companies

Centre-for-globalThis is the traditional lsquooctopusrsquo view of the multinational corporation asingle lsquobrainrsquo located within the company headquarters concentrates thestrategic resources (top management planning and technological exper-tise) and distributes impulses to the lsquotentaclesrsquo (that is the subsidiaries)scattered across host countries Even when some RampD is undertakenabroad this is basically concerned with adapting products to the needsof the local users

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 103

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 104

Local-for-localEach subsidiary of the rm develops its own technological know-howto serve local needs The interactions among subsidiaries are at leastfrom the viewpoint of developing technological innovations rather weakOne the contrary subsidiaries are integrated into the local fabric Thismay occur with conglomerate rms but also in the case of MNEs whichfollow a strategy of technological diversication through tapping intothe competence of indigenous rms

Local-for-globalThis is the case of multinational corporations which rather than con-centrating their technological activities in the home country distributeRampD and technological expertise in a variety of host locations This allowsthe company to develop each part of the innovative process in the mostsuitable environment semiconductors in Silicon Valley automobile com-ponents in Turin software in India The effectiveness of such a strategyrelies on the intensity of intra-rm information owsIn general it has been increasingly observed the emergence of the

trend for MNEs to establish internal (intra-rm) and external (inter-rm)networks for the generation of technological innovation Indeed it hasbeen the development of such cross-border corporate integration andintra-border inter-rm relationships ndash as new forms of technologicalgovernance ndash to make consider the MNE as the key-ring between thelsquolocalrsquo and the lsquoglobalrsquo (Cantwell and Iammarino 2001)

The global technological collaborations

In recent times a third type of globalization of innovative activities hasmade a forceful entry on the scene This in some ways is interme-diate to the two preceding categories Technological collaborations occurwhen two (or more) different rms decide to establish a joint venturewith the aim of developing technical knowledge andor products Threeconditions need to be respected (1) the joint venture should be some-thing more than an occasional and informal collaboration (2) rmspreserve their ownership and (3) the bulk of the collaboration is relatedto sharing know-how andor the generation of new products andprocesses (Mowery 1992) We have witnessed an increasing number ofagreements between rms for the communal development of specictechnological discoveries (Hagedoorn and Schakenraad 1993) Suchcollaborations often take place among rms of the same countrybut in many cases they involve rms located in two or more count-ries thus emerging as authentically global These forms of collaboration for technological advances have promoted a variety of mechanisms for

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

the division of costs and the exploitation of results In a way the neces-sity to reduce the costs of innovation ndash and to cope with its increasingcomplexity ndash has created new industrial organization forms and newownership structures which today are expanding beyond the simpletechnological sphere (Dodgson 1993)However it was not the private sector that discovered this form of

knowledge transmission The academic world has always had a transna-tional radius of action knowledge is traditionally transmitted from onescholar to another and thus disseminated without always requiring pecu-niary compensation

EMPIRICAL RELEVANCE OF THE THREE DIMENSIONSOF THE GLOBALIZATION OF INNOVATION

How important are the three aspects of the globalization of innovationidentied above And more importantly what are the tendenciescurrently predominating In order to answer these questions it is neces-sary to nd appropriate measuring devices None of the availableindicators entirely represents the three aforementioned categories Someindicators do not represent them totally (in the sense that they excludesignicant parts of the phenomenon) others do not represent them exclu-sively (in the sense that they include phenomena that are not part of theobject treated) There are further indicators that represent the phenom-enon neither totally nor exclusively In spite of these limitations thissection reviews the available empirical evidence

The evidence on the international exploitation of technology

The rst indicator of international exploitation of technology is repre-sented by international trade ows Although this is a heterogeneousindicator which includes both innovative and non-innovative productsit is clear that trade is a fundamental means for the international diffu-sion of innovations especially embodied innovations During thepost-war period trade has been growing constantly the export ratio ofgoods and services to GDP in advanced countries went from 94 percentin 1970 to 209 percent in 1995 (OECD 1996a) While all categories ofcommodities embody knowledge this is greater in sectors with thehighest technological content Indeed as shown by Guerrieri and Milana(1995) the sectors in which trade has grown most rapidly are those withthe highest technological content Among these the electronics industryis outstanding as its growth rate has been double with respect to thatof total manufacturing As a whole high tech products which consti-tuted 95 percent of world trade in 1970 represented more than 29 percentin 1995 (Guerrieri 1999) The technologyndashtrade causal relationship is

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 105

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

often a two-way one (Pietrobelli and Samper 1997) On the one handtechnological competence has a positive impact on exports and compet-itiveness on the other international trade boosts the generation and thetransfer of innovations thus giving rise to cumulative causation mech-anisms Figures 2 and 3 show that the correlation between the RampDintensity (measured by the ratio RampDvalue added) and the degree ofinternationalization (measured by the ratio exportsvalue added) for thesix most industrialized countries was remarkably higher in 1996 than in1975 conrming that the link between technological intensity and inter-nationalization has been considerably strengthened over timeScience-based sectors ndash such as Aircraft (ISIC 3845) Professional Goods

(ISIC 385) and especially Ofce and Computing Machinery (ISIC 3825)ndash show a remarkable increase in the degree of internalization endorsingalso the fact that the technologyndashtrade relationship holds particularly fortechnology-intensive areas of production (see also Daniels 1997) Thenotable exception is Drugs and Medicines (ISIC 3522) In this sector astronger technological intensity does not correspond to an increase ininternationalization the position of the sector with respect to the X-axisremained basically unchanged over the two decades This is likely to bedue to the fact that the international exploitation of technological capa-bilities in this sector takes place mainly through foreign direct investmentThe number of patents registered abroad can be considered an indi-

cator of the will to exploit in foreign markets innovations both embodiedin products (a product is patented in order to prevent others fromproducing a similar good and thus to cover all the existing market) anddisembodied (an innovation is patented in order to licence it) Table 1reports the annual rates of growth of selected technological indicatorsfor the main OECD countries in two different periods It shows thatindustrial RampD and resident patents (ie the patent applications of theinventors in their home country) have grown at a moderate pace andsometimes have even experienced a negative rate of change On thecontrary non-resident patents (ie the patent applications of foreigninventors in the country which show to what extent a country has beenlsquoinvadedrsquo by foreigners) and external patents (ie national inventorspatenting abroad which show to what extent a country is lsquoinvadingrsquoother countries) have registered remarkable rates of growth particularlyduring the most recent decade (1987ndash97)Table 2 reports further elaboration on the same data The rst two

columns show the average number of external patents for each residentpatent in 1987 and 1997 Each patent application can in fact be extendedin several countries While a patent application was extended on averagein 13 countries in 1987 it was extended in as many as 62 countries in1997 (OECD 2000a) This is not necessarily due to the increase ofresources devoted to science and technology As shown in columns 5

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 106

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 107

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08

Exportsvalue added

1975

RSq = 03280through origin

10 12 14 16 18

3845

3832

3522

3842+384

3841382+3825

3842

372355+356

3713236

361381

351+352

3385383ndash3832

353+354

3825

Figure 2 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OECD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products +

plastic products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UK US

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 108

04

03

02

Ramp

Dv

alue

add

ed

01

0000 02 04 06 08 10 12 14

Exportsvalue added

1996

RSq = 06674through origin

16

382ndash3825383ndash3832

3522

3832385

3845

3825

3842+384

3843

351+352

32

372

37131 381365

355+356353+354

33841

Figure 3 RampD intensity and internationalization

Note All variables calculated at constant US$ and PPPSource OECD STAN Database 1999 OFCD RampD Expenditure in Industry OECDBasic Science and Technology Statistics 2000KeySectors

3000 Total manufacturing31 Food beverages amp tobacco32 Textiles apparel amp leather352+351ndash3522 Industrial chemicals +

(other chemicals ndash drugs amp medicines)

3522 Drugs amp medicines353+354 Petroleum reneries +

petroleum amp coal products355+356 Rubber products + plastic

products36 Non-metallic mineral products371 Iron amp steel372 Non-ferrous metals381 Metal products382ndash3825 Non-electrical machinery ndash

ofce amp computing machinery

3825 Ofce amp computing machinery383ndash3832 Electrical machinery ndash

radio TV amp communication equipment

3832 Radio TV amp communication equipment

3841 Shipbuilding amp repairing3843 Motor vehicles3845 Aircraft3842+3844+3849 Railroad equipment

+ motorcycles amp bicycles + transport equipment nec

385 Professional goods

CountriesFrance Germany Italy Japan UKUS

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

Table 1

Rates of grow

th of industrial R

ampD and

paten

ting

in the OECD cou

ntries

Annual average growth rates (percent)

Countries

Industrial RampD (1)

Resident patents (2)

Non-resident patents (3)

External patents (4)

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

1970ndash80

1987ndash97

US

20

24

ndash20

52

50

50

ndash06

221

Japa

n61

38

51

11

ndash08

66

55

141

Aus

tria

98a

46e

03

ndash19

34

116

14

136

Belgium

67c

29

ndash30

15

ndash01

105

05

170

Den

mark

38

58

17

26

ndash03

242

10

244

Finlan

d68c

68

47

23

07

259

57

273

Fran

ce37

20

ndash24

04

02

63

30

136

German

y49a

04

ndash07

31

08

61

17

130

Greece

na

61f

ndash08

ndash124j

24

198

na

242

Irelan

d52c

137

68

11

49

358

67

229

Italy

36

ndash07

na

ndash17

kna

82m

18

128

Nethe

rlan

ds

14

12

ndash21

07

15

91

01

185

Portuga

l46d

62f

ndash64

15

ndash05

387

ndash242

176

Spain

127

41

ndash45

24

02

135

13

231

Swed

en59c

45

ndash05

15

25

99

30

219

UK

30b

ndash01

ndash24

ndash10

08

60

ndash17

186

Norway

73

17

ndash27

29

ndash01

132

08

348

Switze

rlan

d08a

ndash01

gndash31

ndash34

22

102

ndash13

134

Aus

tralia

na

74

52

14

ndash20

98

67

186

Can

ada

55

41

ndash11

27

ndash21

60

ndash05

252

Notes

na =

not ava

ilable

a 1970ndash81

b 1972ndash

81c 1971ndash81

d19

71ndash8

0e 198

5ndash93

f 198

6ndash97

g 1986ndash

96j 1987ndash

96k 199

2ndash96

m1992

ndash97

(1)Million US$

at 1995 PPP

(2)Residen

t pa

tents inv

entors in their ho

me coun

try

(3)Non

-residen

t pa

tents foreign

inv

entors in the coun

try

(4)Ex

ternal paten

ts n

ationa

l inve

ntors pa

tenting ab

road

Source C

alcu

lation

s on

OEC

D MST

I (200

0)

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

and 6 the ratio of resident patents per unit of industrial RampD declinedfrom 1987 to 1997 in almost all countries (with the notable exceptionsof the US Germany and Norway where the increase was anyway negli-gible) On the contrary the ratio of external patents per unit of industrialRampD grew dramatically in the same period (see columns 7 and 8 ofTable 2) Columns 3 and 4 report the ratio between non-resident andresident patents A ratio equal to 1 shows that the number of patented

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 110

Table 2 Relation between industrial RampD resident non-resident and externalpatents OECD countries 1987 1997

Countries External Non-resident Resident Externalpatents (4) patents (3) patents (2) Patents (4)

divide divide divide divide

Resident Resident Industrial Industrialpatents (2) patents (2) RampD (1) RampD (1)

1987 1997 1987 1997 1987 1997 1987 1997

US 27 148 095 093 06 08 15 112Japan 03 11 011 019 74 55 22 63Austria 28 139 1095 4507 25b 14d 70b naBelgium 58 353 3692 9326 03 03 26 110Denmark 58 464 762 6198 12 09 74 444Finland 31 345 349 3412 20 13 63 453France 37 140 374 699 10 08 35 107Germany 35 101 146 199 12 16 45 165Greece 02 54 732 12773a 170c 22 22 132Ireland 13 119 396 10208 37 10 55 134Italy na 121 na 1014a na 11e 36 137Netherlands 77 435 1402 3401 07 06 52 299Portugal 16 85 3702 114822 06c 04 12 45g

Spain 14 97 1243 3831 11 09 15 91Sweden 49 386 831 2012 12 08 63b 356g

UK 24 176 262 552 14 13 35 228Norway 27 500 837 2393 10 11 29b 664g

Switzerland 71 379 754 3187 11c 07f 73 293Australia 20 92 207 497 47 25 87 239Canada 36 354 1053 1481 06 05 18 143

Notesna = not available a1996 b1985 for RampD c1986 for RampD d1993 for RampDe1993 for patents f1996 for RampD g1995 for RampD(1) Million US$ at 1995 PPP(2) Resident Patents inventors in their home country(3) Non Resident Patents foreign inventors in the country(4) External Patents national inventors patenting abroad External Patent Applications in year t divided by industrial RampD in year t-1Source Calculations on OECD MSTI (2000)

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

inventions generated in the country is equal to the number of foreigninventions for which patent protection is sought in the country In smallcountries foreign patents strongly outnumber the domestic ones onlyJapan and the US have a number of domestic patents which is greaterthan the foreign one With the exception of the latter country alleconomies increased their dependency from abroad in the period1987ndash97US inventors and rms have considerably increased their penetration

in external markets as shown by the doubling of the ratio of externalto resident patents The same ratio is particularly high for technologi-cally dynamic small and medium sized countries such as theNetherlands Denmark Belgium Switzerland and all the Scandinavianeconomies The case of Japan is not particularly signicant the countryhas a large number of domestic inventions since its patent system is notcomparable to that of other countries The Technology Balance ofPayments (TBP) ndash which reports data on nancial ows connected tothe use of patents licences trademarks inventions etc ndash is another indi-cator of the increased internationalization of innovative activitiesespecially of disembodied technical know-how The nancial transac-tions measured by the TBP include those occurring both betweendifferent rms and between different subsidiaries of the same multina-tional corporation International exchanges of technological know-howand services have increased with respect to the internal business RampDexpenditure (cf OECD 1999a) With the notable exception of Japan andFrance both payments and receipts for technology recorded substantialannual rates of growth in the period 1987ndash97 ndash 108 percent and 138percent respectively on average for the G6 (OECD 2000a) This suggestsa growing interdependence between the national-based innovative activ-ities and the transferacquisition of technology to and from abroadWhat are the reasons underlying the substantial increase of the need

for innovative rms to extend the geographical dimension of theirmarket This seems to be directly linked to the increasing costs of inno-vation on the one hand and to the reduction in the life cycle of productson the other Given that innovations are becoming increasingly costlyand rapidly obsolete innovators must be in the position to commer-cialize them in increasingly large marketsFrom a geographical point of view as we have seen the countries

which are most involved in this form of globalization are the smallestand the most technologically dynamic ndash in other words those showinga higher degree of international integration The limited dimension oftheir domestic market in fact has always induced rms to search abroadfor a market for their products in particular for products requiring higherinvestment Small and medium sized countries have greater difcultiesin promoting innovative programmes on a large scale unless they have

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 111

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

access (guaranteed whereas possible by intergovernmental agreementsor by the existence of customs unions) to foreign markets (Molero 1995)

The evidence on the global generation of innovations

This category only includes multinational enterprises and depending onthe strategy they follow their efforts to generate innovations combiningthe expertise of their afliates in more than one foreign location A rstindicator of this category is represented by the distribution of MNEsrsquoRampD between the home and the host country the data for selected coun-tries are reported in Table 3Columns 1 and 2 report the distribution of RampD in manufacturing

performed within each country by type of ownership of the rm (foreignor national) This shows to what extent countries have been lsquoattractiversquofor RampD-related foreign direct investment The data show that the RampDperformed by foreign subsidiaries accounts for more than 20 percent oftotal RampD in manufacturing in Canada the Netherlands and the UKThe role of foreign rms is equally signicant in the majority of advancedeconomies The only country with a very low share of RampD performedby foreign rms is Japan in this country as much as 99 percent of RampDin manufacturing is nanced by Japanese-owned companies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 112

Table 3 Distribution and intensity of RampD in manufacturing industries byNational Firms and Foreign Afliates Main OECD countries 1996 and 1994

of National total RampD intensities1996 1994

Countries Foreign afliates National rms Foreign afliates National rms

US 120 880 25 25Japan 09 991 12 25Germany 164(1) 836(1) 32 63France 186 814 18(2) 27(2)

UK 395 605 15 19Netherlands 240 760 08(3) 27(3)

Sweden 187 813 24 38Finland 115(4) 885(4) 26 25Canada 403 597 09 17

Notes(1) 1995(2) 1991(3) 1993(4) 1997 Ratio between RampD expenditure and turnoverSource OECD 1999b

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

To a large extent there is a link between RampD and the production ofMNEsrsquo foreign afliates although this is far from being uniform acrosscountries Columns 3 and 4 of Table 3 report the RampD intensities (iethe ratio of RampD expenditure to turnover) of foreign afliates andnational rms While in the US national and foreign afliated have thesame propensity to invest in RampD in all other countries with the excep-tion of Finland the propensity of foreign afliates to nance RampD ismuch lower than for national rms National governments are particu-larly interested to acquire such information since this allows them toknow if inward foreign direct investment contributes on a par withdomestic capital to the creation of the national technological competence(for the policy implications of the globalisation of innovation seeArchibugi and Iammarino 1999) Overall the data conrm the wide-spread belief that in the 1990s MNEs were still more prone to locatetheir RampD facilities in the home country rather than in foreign loca-tionsSo far we have looked at the inward ows of investment in RampD A

specular perspective is represented by the outward ows of RampD invest-ment Unfortunately these data are not available for all countriesHowever the US government since the 1960s has collected data on theRampD performed abroad by its MNEs because of the general concernthat skill-intensive jobs could be displaced abroad These data show thaton average over the 1990s slightly more than 10 percent of the RampD ofUS rms is executed abroad and that the share has slightly increasedover time (NSF 2000 Dalton and Serapio 1995)Outward ows of investment related to the generation of innovation

can also be identied by looking at the patents owned by multinationalcorporations but generated in host countries This allows us to take intoaccount a larger number of countries Each patent record provides infor-mation on the address of the inventor and the name of the owner (whichin most cases is a corporate group) thus allowing the identication ofthe geographical location of both of them On the basis of a signicantsample of large innovative rms during the period 1992ndash96 Patel andVega (1997) showed that 874 percent of their patented inventions wasgenerated in the rmrsquos country of origin and only 126 percent insubsidiaries located abroad (see Table 4) Although there is an increasebetween 1979ndash84 and 1992ndash96 (see also Patel 1995) this is not sufcientto state a radical intensication of the phenomenonHowever as already shown by RampD data there are signicant cross-

country differences Large Japanese rms generate 974 percent of theirpatented inventions in their country whereas American rms concen-trate in the US a relatively smaller share (92 percent) European rmsshow a greater tendency towards decentralization considering Europeas a lsquosingle marketrsquo the share of patents generated outside the continent

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 113

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

is equal to 227 percent Looking at individual countries the propensityof rms to generate innovations abroad is even greater large rms basedin Belgium the Netherlands Switzerland and the UK have more than ahalf of their patents in their subsidiaries abroad German Italian FrenchSwedish and Norwegian multinationals on the contrary have a largershare of their innovations produced at home Using the same dataCantwell (1995) and Cantwell and Kosmopoulou (2000) also consideredlong-term trends They showed that the innovative activities carried outin the subsidiaries located abroad of a sample of North American andEuropean rms more than quadrupled going from 4 percent in the period1920ndash24 to 165 percent in the period 1991ndash95 The analyses of Cantwelland Patel and Vega also allow the identication of the countries in whichsuch rms tend to decentralize their innovative activities At an aggre-gate level more than 90 percent of such activities is hosted by the USWestern Europe and Japan conrming that the globalization of innova-tion by multinational rms has rather to be seen as a process of lsquotriadi-sationrsquo Not even the East Asian dynamic economies have managed toabsorb a substantial share of multinationalsrsquo RampDAs far as the sectoral dimension if concerned rms operating in indus-

tries with higher technological opportunities (Computers Scientic

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 114

Table 4 Geographic location of the US patenting activity of large rmsaccording to their country of origin 1992ndash96 (percentage shares)

Of which hosted in

Nationality Home Abroad US Japan Europe

US 920 80 ndash 11 53Japan 974 26 19 ndash 06Europe 773 227 211 06 ndashGermany FR 782 218 141 07 65France 654 346 189 04 142UK 476 524 381 05 120Italy 779 221 120 00 95Netherlands 401 599 309 09 274Belgium 332 668 140 00 526Sweden 640 360 194 02 142Austria 906 94 22 00 72Finland 712 288 52 00 235Switzerland 420 580 312 09 250Norway 630 370 15 00 333All Companies 874 126 55 06 55

Source Patel and Vega (1997)

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

Instruments Aeronautics Motor Vehicles) show a strong propensity toconcentrate their technological activities in their country of origin Therms with the greatest innovative activity in host countries are thosebelonging to traditional sectors drink and tobacco food building mate-rials other transport mining and petroleum industries (cf Patel 1995tab 6 150 Pavitt and Patel 1998) In the case of natural resources thistendency can be explained by the necessity to situate technological activ-ities locally But also many industries producing consumer goods needtechnological competence in the country of production in order to satisfyboth consumer tastes and national legislative standards An intermediatecase is represented by pharmaceutical and chemical rms in which thepropensity towards the global generation of innovations is above theaverage This might be related to institutional factors rather than strictlytechnological ones it is convenient for rms which are constrained bygovernmental regulations such as the pharmaceutical producers toperform their RampD activities locally so that their products can conformto national standards and satisfy the needs of special lsquoclientsrsquo such asthe governments (Haringkanson 1992)The empirical evidence considered up to this point has concentrated

on two indicators RampD and patents which capture the most importantand codied technological activities However a question emerges as towhether the globalization of multinational enterprises is greater for tech-nological activities which are less formal but equally important for thermsrsquo competitive strategy Multinational corporations in fact transferknowledge to subsidiaries at more than one level These activities includetechnical assistance the often informal exchange of technondashscienticinformation the transmission of organizational and managerial methodsetc They are connected to production and it is thus reasonable to assumethat they should be directly related to direct investment abroad both inproduction and innovationIt is worth noting moreover that the reported indicators capture only

a small part of innovations in a sector which is becoming both increas-ingly important in technological change and globalized software As itis transferred at very low cost some rms have a tendency to sub-contract it to centres in countries with much lower labour costs thantheir own (Antonelli 1991) and to satisfy their own software needs bytight interactions between headquarters subsidiaries and specializedsuppliers However there is still no empirical research quantifying theimportance of such a phenomenonWe may conclude that each member of the triad is differently affected

by this form of globalization of innovation Japan does not participatesubstantially to the global generation of innovation on the one handforeign rms are still reluctant to locate RampD facilities in Japan on theother hand Japanese rms are reluctant to decentralize RampD facilities

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 115

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

abroad Both inward and outward RampD and knowledge-related foreigndirect investment largely contaminate the US The most dynamic situa-tion is to be found in the European countries where a substantial partof national technological competencies is performed by foreign-ownedafliates and where national rms are more and more locating their RampDfacilities both in other European countries and in North America Themost signicant data however are probably the sectoral ones Theyshow that contrary to what occurs in the rst category traditional indus-tries are still more globalized than high-tech ones

The empirical evidence on the global technological collaborations

The available information on global technological collaborations is morefragmented This is partly attributable to the nature of the phenomenonwhich is less easily quantiable than the other two categories First everycollaboration has a different economic and technological signicanceand it is difcult to merge them into a homogeneous unit of measure-ment Second the nature of the collaborations precisely because of theirintermediate form is not easily identiable A precious source of infor-mation is the Merit database on strategic technological alliances (cfHagedoorn and Schakenraad 1990 1993 Hagedoorn 1996) This showsthat the new strategic alliances for technological purposes have substan-tially increased since 1970 to this day and they are particularly relevantin crucial technological areas such as biotechnology new materials andespecially information technologies Although it is not possible to esti-mate the total expenditure on innovation associated with thesecollaborations they turn out to be a relatively new phenomenon whichis particularly signicant for those industries in which technologicalchange has been more intense and where the risks connected to inno-vation are higher Agreements crossing national boundaries constituteby now almost 60 percent of the registered ones Among these around40 percent involves the North AmericandashEuropendashJapan triad whereasthose involving countries outside the triad (mainly Southeast Asian coun-tries) have exceeded 20 percent during the 1990s In spite of this asemerges from an in depth reading of a review of the literature promotedby UNCTAD (Pietrobelli 1996) rms in developing countries are onlymarginally involved in such collaborationsAs far as the total international strategic agreements are concerned

there has been a considerable increase in the second half of the 1980swhich was apparently stabilized during the 1990s (see Table 5) Thefastest growth has been registered by the collaborations between Europeand the US especially in the biotechnology sector The number of collab-orations established by Japanese rms still remains rather limited eventhough it is increasing especially in the information technologies sector

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 116

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

The increase registered by the intra-European agreements is insteadattributable mainly to the biotechnology sector the latter has recordedthe most signicant growth in the number of international alliances bothacross and within geographical macro-regionsStrategic agreements among rms do not cover entirely the phenom-

enon of global collaborations As stated above the academic worldestablished these collaborations well before the business world The acad-emic world has also an inuence over industry and its globalization actsas a vehicle for the transfer of knowledgeAmong the forms contributing to the dissemination of knowledge we

can refer to the increasing number of students attending specializationcourses in foreign countries They represent an uninterrupted channelfor the transfer of scientic and technical knowledge both for developedand for developing countries In the most advanced countries the num-ber of foreign students enrolled in higher education (university level)had a surprising growth rate over the 1980s and the rst part of the1990s Apart from the strong inows registered especially in the smallScandinavian economies the highest growth rates have occurred in theAsianndashPacic area where the number of foreign students registered inhigher education increased at an average annual rate of around 15 per-cent in Japan and 12 percent in Australia The area of origin for this lat-ter case is the Asian continent itself a proof of the fact that learning andknowledge processes are characterized by cultural elements whose sim-ilarities are more likely to manifest within the same macro-area (cfIammarino and Michie 1998) Furthermore the inward ows of foreign

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 117

Table 5 Number of international strategic technology alliances by technologicaleld 1980ndash98

1980ndash84 1985ndash89 1990ndash94 1995ndash98

Total 1286 2540 2477 2655Information technology 469 927 1132 1135Biotechnology 230 499 490 633All other 587 1114 855 887Across regions 709 1306 1191 1193Information technology 258 438 490 463Biotechnology 99 216 261 322All other 352 652 440 408Within regions 577 1234 1286 1462Information technology 211 489 642 672Biotechnology 131 283 229 311All other 235 462 415 479

Source National Science Foundation (2000) from J Hagedoorn MERIT Co-operativeAgreements and Technology Indicators data-base

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

students are even more consistent at the postgraduate level In the USfor example 24 percent of students attending postgraduate courses in1994 came from other countries a percentage that has grown constantlyover time (UNESCO 1996) Thus it is not surprising that universitiesand other public research centres have now even started to operate directinvestments abroad establishing branches in other countries (Malerba etal 1991) We can note the paradox that while rms are imitating uni-versities and developing their know-how through technological collabo-rations some universities are imitating rms by becoming multinationalsThe intensity of international scientic collaborations can also be

measured through the number of articles written in collaboration byacademics of different countries in just a decade the share of interna-tionally co-authored papers in the world has almost doubled (NSF 2000)Even though the majority of scientists continues to work in strict collab-oration with fellow countrymen direct international collaborations areacquiring an increasing weight also facilitated by the diffusion ofInternet This is evident to a substantial extent in the European coun-tries where the share of internationally co-authored scientic articles asa percentage of total co-authored articles was much higher than in theUS or Japan in both 1986ndash88 and 1995ndash97 (see Table 6) The importanceof a global academia would be certainly greater if reference were to bemade to the acquisition of information from abroad through scienticliterature congresses conferences or personal contactsDoes the empirical evidence on the techno-scientic collaborations pro-

vide a conclusive answer as to the relevance of global technological col-laborations They started to appear systematically among rms not morethan a quarter of a century ago but they are rmly established todayrepeating it would seem what occurred in the academic world in theremote past They mainly concern the technological areas with highestopportunities and which are closest to basic research whereas they areless common in traditional sectors Even though the bulk of them involvesessentially the Triad countries a certain vitality has emerged in the newindustrialised countries of East Asia since the beginning of the 1990s

CONCLUSIONS

In this paper we have shown that the globalization of innovation is nota single phenomenon but a catch-all concept to describe a wide rangeof forces The attempt to estimate their weight according to geograph-ical location and industrial sectors shows that the importance of globalforces in innovation is rapidly increasing although at a different pacefor each of the three ongoing processesThe dimensions of globalization summarized in the taxonomy have

not affected the various world regions at the same time and with the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 118

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

same intensity The expansion of global forces has instead remainedcircumscribed to the most developed part of the world up to now somuch so as to have been dened a process of lsquotriadisationrsquo in otherwords of increasing polarization of economic and innovative activitiesin the Triad economies

ACKNOWLEDGEMENTS

We wish to thank Leo Nascia and Lorenzo De Julio for computing assis-tance We gratefully acknowledge the nancial support of the EuropeanCommission Daniele Archibugi under the STRATA project lsquoTheRelationships between Strategies of Multinational Companies and theNational Systems of Innovationrsquo (Contract No HPV1ndashCTndash1999ndash0003) andSimona Iammarino under the TMR Marie Curie Research TrainingProgramme (Contract No ERBFMBICT961062)

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 119

Table 6 Percentage of internationally co-authored papers published in selectedcountries in all papers

Country 1986ndash88 1995ndash97 Absolute growth

US 98 180 84Japan 81 152 88European UnionUK 167 293 75Germany 207 337 63France 222 356 60Italy 240 353 47Netherlands 213 360 69Sweden 240 394 64Denmark 259 443 71Finland 209 361 73Belgium 312 466 49Austria 271 436 61Ireland 289 419 45Spain 188 322 71Greece 276 383 39Portugal 376 508 35World 78 148 90

Note The world totals appear lower than those of individual countries because for worldtotals each internationally co-authored paper is counted only once while each collabo-rating country is assigned one paper In 1997 each internationally co-authored paperinvolved an average of 222 countriesSource National Science Foundation (2000)

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

REFERENCES

Antonelli C (1991) The Diffusion of Advanced Telecommunications in DevelopingCountries Paris OECD Development Centre

Archibugi D Howells J and Michie J (eds) (1999) Innovation Systems in theGlobal Economy Cambridge Cambridge University Press

Archibugi D and Iammarino S (1999) lsquoThe policy implications of the globali-sation of innovationrsquo Research Policy 28 317ndash36

Archibugi D and Michie J (1995) lsquoThe globalisation of technology a newtaxonomyrsquo Cambridge Journal of Economics 19 1 121ndash40

Archibugi D and Michie J (1997a) lsquoTechnological Globalisation or NationalSystems of Innovationrsquo in D Archibugi J Michie (eds) (1997b)

Archibugi D and Michie J (eds) (1997b) Technology Globalisation and EconomicPerformance Cambridge Cambridge University Press

Bartlett CA and Ghoshal S (1990) lsquoInnovation processes in multinational corpo-rationsrsquo Strategic Management Journal 11 499ndash518

Bell M and Pavitt K (1997) lsquoTechnological Accumulation and Industrial GrowthContrasts between Developed and Developing Countriesrsquo in D Archibugiand J Michie (eds) (1997b)

Cantwell JA (1995) lsquoThe Globalisation of Technology What Remains of theProduct Cycle Modelrsquo Cambridge Journal of Economics 19 1 155ndash74

Cantwell JA and Iammarino S (1998) lsquoMNCs Technological Innovation andRegional Systems in the EU Some Evidence in the Italian Casersquo InternationalJournal of the Economics of Business Special Issue 5 (3) 383ndash408

Cantwell JA and Iammarino S (2002) Multinational Enterprises and RegionalSystems of Innovation in Europe London Routledge

Cantwell JA and Kosmopoulou E (2000) lsquoWhat determines the international-isation of corporate technologyrsquo University of Reading mimeo

Dalton DH and Serapio MG (1995) Globalising Industrial Research andDevelopment Springeld VA National Technical Information Service USDepartment of Commerce

Daniels PL (1997) National technology gaps and trade ndash An empirical studyof the inuence of globalisationrsquo Research Policy 25 1189ndash1207

Dodgson M (1993) Technological Collaboration in Industry London RoutledgeDunning JH (1993) Multinational Enterprises and the Global Economy Wokingham

Addison-WesleyFlorida R (1997) lsquoThe globalisation of RampD results of a survey of foreign-

afliates RampD laboratories in the USArsquo Research Policy 26 1 85ndash103Freeman C (1994) lsquoThe economics of technical changersquo Cambridge Journal of

Economics 18 463ndash514Giddens A (1990) Consequences of Modernity Cambridge Polity PressGranstrand O Haringkanson L and Sjoumllander S (eds) (1992) Technology Management

and International Business Internationalisation of RampD and TechnologyChichester Wiley

Guerrieri P and Milana C (1995) lsquoChanges and Trends in the World Trade inHigh-Technology Productsrsquo Cambridge Journal of Economics 19 1 225ndash42

Guerrieri P (1999) lsquoLa competitivitagrave tecnologica dei paesi europeirsquo in S Ferrariet al (eds) LrsquoItalia nella competizione tecnologica internazionale Milan FrancoAngeli

Hagedoorn J (1996) lsquoTrends and Patterns in Strategic Technology PartneringSince the early Seventiesrsquo Review of Industrial Organisation 11 601ndash16

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 120

Hagedoorn J and Schakenraad J (1990) lsquoInter-rm Partnerships and Co-operative Strategies in Core Technologiesrsquo in C Freeman and L Soete (eds)New Explorations in the Economics of Technical Change London PinterPublisher

Hagedoorn J and Schakenraad J (1993) lsquoStrategic Technology Partnering andInternational Corporate Strategiesrsquo in K Hughes (ed) European Com-petitiveness Cambridge Cambridge University Press

Haringkanson L (1992) lsquoLocational Determinants of Foreign RampD in SwedishMultinationalsrsquo in O Granstrand L Haringkanson and S Sjoumllander (eds)Technology Management and International Business Internationalisation of R amp Dand Technology Chichester Wiley

Howells J (1997) lsquoThe Globalisation of Research and Technological InnovationA New Agendarsquo in J Howells and J Michie (eds)

Howells J and Michie J (eds) (1997) Technology Innovation and CompetitivenessAldershot Edward Elgar

Hu YS (1992) lsquoGlobal or Stateless Corporations are National Firms withInternational Operationsrsquo California Management Review 34 107ndash26

Iammarino S and Michie J (1998) lsquoThe Scope of Technological GlobalisationrsquoInternational Journal of the Economics of Business Special Issue 5 3 335ndash53

Malerba F Morawets A and Pasqui G (1991) The Nascent Globalisation ofUniversities and Public and Quasi-Public Research Organisations BrusselsEuropean Commission FAST

Molero J (1995) Technological Innovation Multinational Corporations and NewInternational Competitiveness The Case of Intermediate Countries LondonHarwood Academic Publishers

Mowery D (1992) lsquoInternational Collaborative Ventures and US FirmsrsquoTechnology Strategyrsquo in O Granstrand L Haringkanson and S Sjoumllander (eds)Technology Management and International Business Internationalisation of R amp D and Technology Chichester Wiley

Narula R and Hagedoorn J (1999) lsquoInnovating through strategic alliances mov-ing towards international partnerships and contractual agreementsrsquoTechnovation 19 283ndash94

National Science Foundation (1996 and 2000) Science and Engineering IndicatorsWashington DC US Government Printing Ofce

OECD (1996a) STAN Database 1975ndash1994 Paris OECDOECD (1996b) Research and Development Expenditure in Industry 1973ndash1993 Paris

OECDOECD (1997) Internationalisation of Industrial RampD Patterns and Trends Groups

of National Experts on Science and Technology Indicators Paris OECDOECD (1999a) STAN Database Paris OECDOECD (1999b) Measuring Globalisation The Role of Multinationals in the OECD

Economies Paris OECDOECD (2000a) Main Science and Technology Indicators Paris OECDOECD (2000b) Basic Science and Technology Statistics Paris OECDPatel P (1995) lsquoLocalised Production of Technology for Global Marketsrsquo

Cambridge Journal of Economics 191 141ndash53Patel P and Vega M (1997) lsquoTechnological Strategies of Large European Firms

Report for ldquoStrategic Analysis for European SampT Policy IntelligencerdquorsquoTargeted Socio-Economic Research European Commission

Pavitt K and Patel P (1998) lsquoGlobal Corporations and National Systems ofInnovation Who Dominates Whomrsquo in D Archibugi J Howells and JMichie (eds)

Pearce RD and Singh S (1992) Globalizing Research and Development LondonMacmillan

ARCHIBUGI AND IAMMARINO TECHNOLOGICAL INNOVATION

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 121

Pianta M (1988) New Technologies Across the Atlantic Hemel HempsteadHarvester amp Wheatsheaf

Pietrobelli C (1996) Emerging Forms of Technological Cooperation The Case forTechnology Partnership New York and Geneva United Nations Conferenceon Trade and Development

Pietrobelli C and Samper J (1997) lsquoMeasurement of EuropendashAsia TechnologyExchanges Asymmetry and Distancersquo Science and Public Policy XXIV (4)August

Scherer FM (1992) International High-Technology Competition London HarvardUniversity Press

Streeten P (1996) lsquoGovernance of the Global Economy paper presented at theInternational Conference on ldquoGlobalisation and Citizenshiprdquo lsquo GenevaUnited Nations Research Institute for Social Development December

Tyson LD (1992) Whorsquos Bashing Whom Trade Conict in High-TechnologyIndustries Washington DC Institute for International Economics

UNESCO (1996) Statistical Yearbook GenevaZander I (1999) lsquoHow do you mean ldquoglobalrdquo A taxonomy of innovation

networks in the multinational corporationrsquo Research Policy 28 2ndash3 195ndash214

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

111123456789

11110111213141516171819

11120212223242526272829

11130313233343536373839

11140414243

11144

Folio 122