Technology Transfer in the TRIPS Age: - IPRsonline.org

May 2009 l ICTSD Programme on IPRs and Sustainable DevelopmentICTSD Intellectual Property and Sustainable Development SeriesTechnology Transfer in the TRIPS Age:The Need for New Types of Partnerships betweenthe Least Developed and Most Advanced EconomiesBy Dominique ForayEcole Polytechnique Fédérale de LausanneICTSDIssue Paper No.23

May 2009 l ICTSD Programme on IPRs and Sustainable DevelopmentTechnology Transfer in the TRIPS Age:The Need for New Types of Partnerships betweenthe Least Developed and Most Advanced EconomiesBy Dominique ForayEcole Polytechnique Fédérale de LausanneICTSDIssue Paper No.23

iiDominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesPublished byInternational Centre for Trade and Sustainable Development (ICTSD)International Environnent House 27 Chemin de Balexert, 1219 Geneva, SwitzerlandTel: CH +41 22 917 8492 Fax: +41 22 917 8093E-mail: ictsd@ictsd.chInternet: www.ictsd.orgExecutive Director:Core Team:Ricardo Meléndez-OrtizChristophe Bellmann: Programmes DirectorDavid Vivas-Eugui: Deputy Programmes DirectorAhmed Abdel Latif: Programme Manager, Intellectual PropertyPedro Roffe: Senior Fellow, Intellectual PropertyAcknowledgmentsThe author is grateful for comments and suggestions received from Pedro Roffe as well as for the verystimulating discussions during presentations of the first and second drafts of the paper at various occasions(meetings of the IP Group of the Geneva Trade & Development Forum; panel discussion organised byUNCTAD and ICTSD, Palais des Nations, Geneva, 16 June 2008). The author is grateful for the support forthis research by the ICTSD Programme on Intellectual Property Rights and Sustainable Development.Support for the ICTSD Programme on Intellectual Property Rights and Sustainable Development hasgenerously been provided by the UK Department for International Development (DFID).Dominique Foray is Chair in Economics and Management of Innovation, Ecole Polytechnique Fédérale deLausanne.For more information on ICTSD’s Programme on IPRs and Sustainable Development, please visit our websiteat http://ictsd.net/programmes/ip/ICTSD welcomes feedback and comments on this document. These can be sent to Ahmed Abdel Latif ataabdellatif@ictsd.chCitation: Foray, Dominique (2009). Technology Transfer in the TRIPS Age: the Need for New Types ofPartnerships between the Least Developed and Most Advanced Economies. ICTSD Programme on IPRs andSustainable Development, Issue Paper No. 23, International Centre for Trade and Sustainable Development,Geneva, Switzerland.The views expressed in this publication are those of the authors and do not necessarily reflect the views ofICTSD or the funding institutions.Copyright (c) ICTSD, 2009. Readers are encouraged to quote and reproduce this material for educational,non-profit purposes, provided the source is acknowledged.This work is licensed under the Creative Commons Attribution-Non-commercial-No-Derivative Works 3.0License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/ or send aletter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA.ISSN 1684-9825

ICTSD Programme on IPRs and Sustainable DevelopmentiiiTABLE OF CONTENTSAcronymsList of Figures, Tables and BoxesForewordExecutive SummaryvviviiixINTRODUCTION 11. OVERVIEW OF MAIN FINDINGS 41.1 – The Economic Fundamentals of Technology Transfer to LDCs 41.2 – Current Situation and Key Messages 61.3 – Prioritising Areas where TT Should be Operated 71.4 – Towards New types of Partnerships 81.5 – Responding Effectively to TRIPS Article 66.2 92. THE NATURE OF TECHNOLOGY TRANSFER 102.1 – Information and Knowledge 102.2 – Consolidation of Technology Transfer 132.3 – TT in the Context of LDCs: Capabilities,Organisational Design and Models of Innovation 142.4 – The Choice of Technology Transfer Areasin the Case of LDCs 182.5 – Transactional Modes, Barriers and Incentives 202.6 – Brief Summary 263. TECHNOLOGY TRANSFER ANDINTELLECTUAL PROPERTY RIGHTS 283.1 – The Conditions for Technology Transferto LDCs under TRIPS 283.2 – The Impact of TRIPS on Technology Transfer 293.3 – Markets for Technology: Are They Really Efficient? 333.4 – A Fundamental Asymmetry 34

ivDominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies3.5 – Beyond IPRs: the New Knowledge Economy Paradigm 394. CURRENT TRENDS IN TECHNOLOGY TRANSFER 414.1 – The World Bank Framework 414.2 – Virtuous Circles 424.3 – Vicious (or Non Virtuous) Circles: the Case of LDCs 435. THE ECONOMICS OF PUBLIC-PRIVATE PARTNERSHIPSFOR TECHNOLOGY TRANSFERS BETWEEN DEVELOPEDCOUNTRIES AND LDCS 455.1 – The Economic Rationale of TRIPS Article 66.2:the Potential and Limitations of Incentives 455.2 – Historical Precedents:Both Incentives and Specialised Agents in PPPs Matter 475.3 – Public-Private Partnerships for Technology Transfer 515.4 – Effectiveness Conditions: Appropriate Donor Assistance 535.5 – Drawing from TRIPS Council Country Reports underArticle 66.2 50Conclusion 56ENDNOTES 57REFERENCES 59

ICTSD Programme on IPRs and Sustainable DevelopmentvABBREVIATIONS AND ACRONYMSFDI: Foreign Direct InvestmentGPT: General Purpose TechnologiesICTs: Information and Communication TechnologiesIPRs: Intellectual Property RightsLDCs: Least Developed CountriesMNCs: Multinational CorporationsODA: Overseas Development AssistancePPPs: Public-private PartnershipsTRIPS: Agreement on Trade-Related Aspects of Intellectual Property RightsTT: Technology TransferWIPO: World Intellectual Property OrganisationWTO: World Trade Organisation

viDominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesLIST OF FIGURES, TABLES AND BOXESLIST OF FIGURES1 – The Growth of Royalties and Licensing Fees 252 – Patent Applications in the LDCs, Residents vs.Non-residents, 1995-2002 363 – Positive Feedback Systems, Virtuous and Vicious Processes 424 – Functions and Role of a PPP as TT Operator 485 – Global and Local Demand and the Need for SpecialIncentives for Local Entrepreneurs 506 – Projects Targeted by an Optimal Government-IndustryTT Programme 52LIST OF TABLES1 – A Typology of TT as an Economic Operation 62 – Mapping the Areas for TT Focus 203 – Costs and Benefits of a Stronger Patent System 354 – Patent Applications by Residents and Non-residents in LDCs 365 – Mapping Technology Transfers Documented in a Country Report:A Real Example 55LIST OF BOXES1 – Knowledge as a Commodity 112 – Cleaner Production Centres as an Example of a Technology Platform 163 – Defining the Areas of Focus: Six Arguments 184 – TRIPS and the Public Domain of Knowledge 29

ICTSD Programme on IPRs and Sustainable DevelopmentviiFOREWORDTechnology Transfer in the TRIPS Age: The Need for New Types of Partnerships between theLeast Developed and Most Advanced Economies is a recent contribution of the ICTSD Programmeon Intellectual Property Rights and Sustainable Development. Technology transfer (TT) todeveloping countries has been discussed by the international community and policy makers ingeneral for more than three decades and continues to resurface in international deliberations.In recent years, ICTSD has contributed to these discussions by producing some pioneeringstudies. 1This study draws on the empirical literature addressing TT issues in an age where the trend ofreinforcing the protection and enforcement of intellectual property rights (IPRs) prevails. Ithighlights the complexity of the TT process particularly when entities at different developmentlevels are involved and the importance, in these cases, of the absorptive capacity of therecipient firms. The study also examines the arguments advanced to consider intellectualproperty (IP) protection as both a hindrance and a stimulus to TT. The paper makes practicalrecommendations, applicable, first, to least developed countries (LDCs), that want to use TTsas an effective growth engine, and, second, to developed countries that have to comply withthe provisions of Article 66.2 of the World Trade Organisation (WTO) Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS).According to the study, many countries are increasing their exposure to foreign technologiesby means of trade and foreign direct investment (FDI), while improving absorptive capacitiesto facilitate the dissemination of technologies and spillovers within the domestic economy.This represents a virtuous path that a number of developing countries – notably middle-incomecountries – follow. The experience of LDCs, however, is less promising. Firstly, FDI and traderemain at a low level and the poor quality of their absorptive capacities makes it unlikely thatthe few foreign technologies that are transferred will disseminate throughout the economy. Inthis respect, the main message is that the number, scale and domains of TTs cannot be left todepend on FDI or trade, nor can they take only the form of market transactions (licences).With respect to TRIPS Article 66.2 obligations, the study suggests that transfer of technologyshould be part of a principal economic operation and not be a joint product or by-product,contingent on other operations. The locus of decision-making regarding modes of learning andareas for focus must shift away from foreign bodies to local agents and authorities. As faras additional incentives by governments are concerned, they should consist of assistance forprojects that are socially beneficial but with low expected profitability for technology supplierfirms. This should ensure that conditions for TT involve the choice of relevant partners on boththe supply and demand sides, selection of the right area related to a clearly expressed localdemand for technology and the creation of organisational forms that favour the consolidationof the transfer (absorption, adaptation and subsequent spillover), as well as the relatedentrepreneurial dynamic.In this context, the study advances the need for greater use of public-private partnerships(PPPs) as a mechanism for ensuring both the effectiveness of the intervention and the efficiencyof the TT operation.The premise of ICTSD’s work in this field is based on the appreciation that IPRs are economicallyand politically important but controversial, and towards that end a better understanding of IP isindispensable for informed policy making at the national and international level. The relationshipbetween transfer of technology and IP stands high today in the deliberations of the Council forTRIPS, in the discussions of the UN Framework Convention on Climate Change (UNFCCC) and inthe implementation of the WIPO Development Agenda recommendations adopted in 2007.

viiiDominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesEmpirical evidence on the role of IP protection in promoting innovation and dissemination ofknowledge remains inconclusive. Diverging views persist on these matters. However, there isa consensus that IP matters and that the right policies and friendly environments are neededto gear the system to achieve its fundamental objectives: namely, to promote innovation andcreativity and facilitate the dissemination of knowledge.A central objective of the ICTSD Programme on Intellectual Property and Sustainable Developmenthas been to facilitate the emergence of a critical mass of well-informed stakeholders indeveloping countries – including decision makers and negotiators, but also within the privatesector and civil society – who will be able to define their own sustainable human developmentobjectives and effectively advance them at the national and international levels.We hope you will find this study an additional contribution to the debate on IP, innovation andsustainable development, and particularly in responding to the need for increased awarenessand better understanding of the complexities of technology transfer and, among others, itsinterface with IP. An underlying assumption of our work has been the pursuit of a proper balancebetween the different interests at stake in designing appropriate IP regimes compliant withinternational commitments and supportive of public policy objectives as well as developmentefforts.Ricardo Meléndez-OrtizChief Executive, ICTSD

ICTSD Programme on IPRs and Sustainable DevelopmentixExecutive SummaryThis study draws on the abundant empirical literature produced over recent decades that addressesthe issues of technology transfer (TT) between countries with very different development levels in anage of stronger intellectual property (IP) regimes.The key findings of this study are that, in the case of least developed countries (LDCs), the number,scale and domains of TT cannot depend alone on general economic operations, such as foreign directinvestment (FDI) or infrastructure construction; neither can they only take the form of markettransactions (licences). In all these cases, the particular circumstances and conditions that prevail inLDCs imply a suboptimal level of TT in relation to these countries’ needs.There is therefore an obvious economic rationality for specific projects in which the TT is the primaryproduct (an economic project in itself, not linked with another economic operation) but entails alow expected private profitability for the technology-owning firm. Such a prospect would involveacknowledging the existence of TT operations with far smaller commercial returns or no commercialreturn at all and finding operational mechanisms to incentivise these firms to sink costs in theseoperations. Such a strategy requires the provision of additional incentives from governments ofdeveloped countries.Incentivising foreign firms to enter such transactions is a clear opportunity for developed countrygovernments to properly fulfil their obligations contained in Article 66.2 of the World Trade Organisation(WTO) Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). The followingrecommendations are applicable to developed countries’ TRIPS obligations:• The transfer of technology should form the subject of a principal economic operation (andnot be a joint product or by-product; i.e. contingent on other operations);• The locus of decision-making regarding modes of learning and areas for focus must shiftaway from foreign bodies to local agents and authorities;• In providing additional incentives to the technology-owning firms, governments should seekeffectiveness and efficiency. To achieve this:- Governments should provide incentives in an effective way by only assisting projectsthat are socially beneficial but not very profitable for the firms that own and couldtransfer the technology; and- Conditions for the efficiency of the TT operations involve the choice of relevantpartners both on supply and demand sides, selection of the right area for focus (relatedto a clearly expressed local demand for technology) and the creation of organisationalforms that will favour the consolidation of the transfer (absorption, adaptation andsubsequent spillovers), as well as the related entrepreneurial dynamic;• Governments should make as much use as possible of public-private partnerships (PPPs) as amechanism for ensuring both the effectiveness of the intervention and the efficiency of theTT operation.Technology transfers have to be operated in many domains (including export-oriented industry). Butthey must be particularly supported in those domains that correspond to the model of innovationcentral to economic growth in LDCs: that is, entrepreneurial activities meeting needs in local marketsthat are likely to generate domestic spillovers. In other words, TTs must offer a positive supply

xDominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesresponse to a demand for technology stemming from local entrepreneurs. Two factors are relevanthere:1. These domains are potentially important for growth because the spillovers generated in thecourse of such projects are likely to be captured by the local economy; and2. These domains need additional incentives so the donor’s intervention will be effective andrespond fully to the TRIPS provision, which is not necessarily the case of export-orientedsectors in which the market incentives alone are sufficiently strong to motivate firms in richcountries to operate TTs.The other areas – for instance the export goods-oriented manufacturing and processing sectors – arealso important but they will in any case be served through TTs operating as joint products of FDI. Assuch, they should not be listed as part of the efforts made by the developed countries to comply withArticle 66.2 of the TRIPS Agreement.

ICTSD Programme on IPRs and Sustainable Development1INTRODUCTIONThe broad and rapid diffusion of new andadvanced knowledge is positive for social wellbeing.Efficiency and growth are promotedby the speed with which new knowledge andinnovative technologies are disseminated: thegreater the proportion of individuals, firms, orcountries making swift use of superior productsand processes rather than being restricted toinferior substitutes, the more widespread andsubstantial the growth benefits. If, for instance,only one firm (or one country) uses a newtechnology that increases productivity whileother firms (or countries) are obliged to retainthe old, less efficient technique, the benefitsare significantly less than if all firms were ableto adopt the new technology. This disseminationis important for the efficiency of the economicgrowth process.One of the main forms of knowledge dissemination(and probably the most valuable from the pointof view of developing productive and innovativecapabilities) is technology transfer (TT). Krugman(1979), for instance, considers the pattern oftrade and economic growth to be governedprimarily by two activities: innovation and TT.Early literature based on the standard modelof economic growth predicted convergenceacross countries and considered TT as an “easy”mechanism to achieve this process. Once produced,technology can be widely applied. Theimplicit assumption is that the cost of technologydiffusion is lower than that of its production.However, on the ground, the difficulties andcomplexity of TT operations, particularly whenthey involve two countries different levels ofdevelopment, has been recognised in thegrowing literature on international technologytransfer. Some of the landmarks towards a betterunderstanding of the difficulties and complexityof TT include:• A more adequate conceptualisation oftechnology (stressing its tacit dimension);• A careful appreciation of the transactionaldifficulties particularly related to problemsof contracting tacit knowledge and pricingtechnologies on markets that remainrelatively inefficient in this respect;• A better understanding of the importance(and non-trivial nature) of diffusion withinthe country once a technology has beentransferred there. This is a process thatinvolves particular barriers and incentives;• The growing consideration of the fact thatinnovation and technological progress inLDCs cannot be a mere reproduction ofwhat is happening in developed countries.Innovations must be tailored accordingto local opportunities, capabilities andneeds. Their consequences for economicdevelopment are strongly dependent onhow such “particularisation” materialisesin products, processes and services;• The recognition that the most importantinnovations for LDCs are probably not“technical” but reside in the process ofdiscovering a country’s strength; and• An accumulation of evidence concerningthe relative advantages and shortcomingsof the different channels through which TToccurs.These landmarks towards a better understandingof TT, the conditions for its success and how tominimise the risks of failure, have not, however,given rise to new policy opportunities and theimplementation of new programmes.In parallel, the knowledge economy has beenestablished slowly in rich countries, stronglyaffecting the relative importance of the various TTchannels. And while the market for technologiesas a prominent mechanism has grown over thelast two decades, this has possibly come at theexpense of non-market based mechanisms, suchas imitation.This study draws on the abundant empiricalliterature on TT produced over recent decades.

2 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesAfter an overview of the main findings of thepaper in chapter 1, chapter 2 expands on the ideaof the cognitive complexity of TT based on thedistinction between knowledge and information.It then examines the different phases that make upthis complexity. It discusses how this complexitybecomes even greater when the transfer involvesentities of very different development levels anddemonstrates the pivotal nature, in this case, ofthe notion of capabilities. The building of thesecapabilities is carried out within the framework ofsuitable organisational forms, such as technologyplatforms or production centres. This sectionthen elaborates on a concept of incrementalinnovation oriented towards the development oflocal applications – a concept that should help tostructure the objectives of TTs made to LDCs.Chapter 2 also raises the question of the choicesof specialised domains in which to undertakeTTs. It makes reference to the work by Enosin particular to show how choices that areirrelevant, unconsidered and guided by donorscan lead to catastrophic results. It suggests theimportance of the process of discovering goodspecialisations, as effectively demonstratedby the works of Hausman and Rodrik; indeed,technologies from industrialised countries veryoften require complex and costly adaptations,with an uncertain degree of success. Learningwhat a country is good at producing is thereforea critical issue and requires entrepreneurialcommitments. The knowledge that is likely tobe generated from these “experiments” is ofgreat social value. Finally, chapter 2 tacklesthe question of barriers and incentives offeredto firms to undertake TT operations and carrythem through successfully. These incentives andbarriers are examined for different forms of TTs(main product, joint product and by-product).Chapter 3 deals with the role of intellectualproperty in TTs. Recently, the number of transfersbased on licence purchasing has increasedconsiderably. This must be correlated with theglobalization of IP systems and the increasingpower of technology markets. This part of thetext summarises the arguments that point to anegative role of IP protection and those that,on the contrary, indicate a positive role. InChapter 3, the question of the effects and generalimpact of TRIPS on technological innovation inLDCs is also dealt with. The development of aglobal IP system, corresponding to uniform qualitystandards has, without a doubt, positive effectson rich countries. It makes extremely uncertaininvestments more secure, encourages betterknowledge dissemination via the market and givesgreater visibility to innovation efforts and theaccumulation of intellectual assets, especially insmall firms. These positive effects indisputablycompensate for the social cost imposed bymonopoly prices in a system that is functioning“reasonably” well. However, these positiveeffects can under no circumstances be predictedin the context of international TTs betweencountries with very different development levels.In this particular case, the new system takes theplace of a previous system based on imitation,without replacing it completely. A comparativelyimportant TT channel disappears; others takeits place but also stipulate different rules lessfavourable to LDCs.Chapter 4 takes stock of the most recent evidenceabout TT and technology diffusion as compiledby the World Bank (2008). It draws on this datato observe that middle-income countries areincreasingly successful in using FDI and trade tooperate TTs (TT having clearly the status of ajoint product) and that due to the improvementof their absorptive capacities, those TTs generatespillovers to the domestic economy resultingin an increase of its overall efficiency. Whenthings are going well in this way, one may standback in awe at the unfolding of the process andits ability to sustain high marginal social andprivate returns on TTs over an extended periodof time. But this is occurring far less in LDCsand evidence shows that trade and FDI remainat a low level. Thus the success of TT as jointproduct is poorly supported in LDCs, while weakabsorptive capacities prevent the few TTs thatdo take place from leading to spillovers into thedomestic economy.Finally, chapter 5 develops a TT model thatincorporates the different elements consideredto provide general practical recom-mendations,first to LDCs that want to effectively use TTs

ICTSD Programme on IPRs and Sustainable Development3as a growth engine, and second to developedcountries that have to comply with theprovisions of TRIPS Article 66.2. This segmentfirst examines the conditions for implementationof TRIPS Article 66.2, which encourages firms inrich countries to transfer technologies to LDCs.Here, the pertinence of the objective (increasingincentives) is shown and, through the use ofrecent examples, suggests that under certainconditions private firms can “respond” positivelyto an increase in incentives – in the case of orphanand neglected diseases, for instance. Success inthese areas (expressed as the commitment ofprivate firms to activities considered a prioriunprofitable) is apparently not carried over tothe transfer of technology to LDCs however.Chapter 5 thus attempts to explain this absenceand to outline some rudimentary solutions,involving the use of PPPs as a mechanism toincrease the efficiency of TT operations, as wellas the careful selection of areas to ensure theeffectiveness of aid from international donors(incentives should be provided where and whenthey are really needed).

4 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies1. OVERVIEW OF MAIN FINDINGS1.1 The Economic Fundamentals of Technology Transfer to LDCsTechnology transfer basically consists of a“transfer” of technological knowledge, includinga “transfer” of the capacity to assimilate,implement and develop a technology. Theacquisition of information concerning thetechnology is thus only a part – althoughadmittedly an important one – of the transfer.The process of learning how to use andmaintain the technology is at least as importantas its frequently required adaptation to localconditions. Ultimately, this adaptation may leadto the development of new applications basedon the transferred technology. Furthermore, atechnology is itself progressive; therefore, thechallenge of a TT is to master technologies thatare changing continuously. Finally, the perimeterfor potential adopters of the transferredtechnology is never predetermined and thetransfer’s social returns depend essentially onthe extension of this perimeter beyond theinitial target – for example, a specific industrialinstallation. In other words, the transferprocess should not stop at the first installationbut with the generalisation of the technology inthe geographical space considered.Thus a successful TT goes through the perilousphases of the assimilation and absorptionof technological knowledge: adaptation tolocal conditions, absorption of subsequentimprovements and generalisation of thetransferred knowledge. These phases are describedas the consolidation of the transfer. Thiscomplex cognitive dimension of TT demandsan equivalent level of complexity of theorganisational forms related to the transferprocess. The latter must guarantee not onlythe acquisition of information but also thelearning of the technology, its adaptation andits progression. Favourable organisationalconditions can also promote its generalisation.It is in accordance with this logic that theefficiency of the various transfer modes can beevaluated.The TT context between systems with verydifferent development levels makes theproblems of TT consolidation (absorption andlearning, adaptation, assimilation of subsequentimprovements, generalisation) even moredifficult to solve. Such heterogeneities create“capabilities’ issues.” Weak capabilities on theLDC side impose strong forms of internal andexternal organisation in the LDC to maximisethe probability of succeeding in the variousphases of TT. Weak capabilities also imply thatthe various modes of TT are not “equivalent”in terms of their potential to impact theproductivity of a wide range of sectors in thelocal economy.As far as LDCs are concerned, this studyemphasises the centrality of a particular modelof innovation for growth and development.Research and development (R&D) and othermore informal learning activities undertaken toproduce locally oriented innovations, allow thecountry to develop absorptive capacity, whileat the same time the locally generated spilloverfrom this same R&D may end up diffusing awayfrom the local economy. There are vast areas ofeconomic activity where innovation is neededto serve local needs and local demand, whereby“local” may mean a large fraction of the worldpopulation.Economically speaking, it makes sense to knowwhether the TT constitutes an economicoperation in itself or is dependent on aneconomic operation that exceeds it. In thisrespect, a distinction will be made between:• Transfers dependent on more generaleconomic operations, such as foreign directinvestment (FDI), the construction andsupply of infrastructure, or the integrationof companies from developing countries asexporters in international trade. In anyevent, the TT is a joint product or byproduct;2 the importance and quality ofTTs are therefore contingent on a largenumber of factors resulting from theprimary economic operation. It is thusthe incentives for the primary economic

ICTSD Programme on IPRs and Sustainable Development5operation to succeed that determinewhether the TT and its consolidation alsosucceed (or not);• Transfers that in themselves constitutethe main operation. This is carried outeither through the medium of the market(licence, joint ventures), or via nonmarketchannels. In these cases, the primemotivation for the operation is the successof the TT and it is the incentives directlylinked to the TT (cost and profit) thatcommand the operation.When TT is a joint product of FDI, the locusof decision-making lies in foreign firms. Foreignfirms decide the modes of learning, the amountof resources devoted to the TT operation, thepotential scope for further dissemination,and the objectives of assimilating subsequentimprovements. The choice of the domain wherethe TT will operate is pre-defined when the TTis a joint or a by-product. Drawing on Enos et al.(1998), this paper proceeds to argue that whenthe locus of decision-making for TTs lies withthe foreign firms, there is a risk of suboptimaldecision processes.When the TT is the main operation, there ismore flexibility about who will make the maindecisions: i.e., domestic bodies (entrepreneurs,governmental agencies), foreign donors or foreignfirms. For instance, when the TT is themain operation, the question of the choiceof domains (who makes the choice, based onwhat criteria, how to go about discovering the“best” domains, etc.) becomes an integralpart of the economic issue of TT. This aspectis crucial – the choice of where TTs will operatematters a great deal. TTs operate in specialiseddomains that they will strengthen in order totransform them into a growth engine for therelevant country. Enos et al. (1998) made thisargument well, stating that: “the main taskin advancing science and technology in thedeveloping countries is to identify the mostattractive direction in which to proceed. Thisproper direction is that which best representsthe interests of the developing countries andthis interest is not congruent with the interestsof banks or foreign donors.”It is therefore apparent that the technologyholder’s commitment may vary considerablydepending on the TT’s status as an economicoperation and the importance attributed to it inthe success of the primary operation (if it doesitself not constitute the primary operation).When the TT is a joint product (occurring asa consequence of a direct investment), and ifthings are going well, the phases of absorption,adaptation, and assimilation of subsequentimprovements are in a sense “embedded” inthe investment plan. However, since the TTis a joint product, there is the problem of“balancing incentives” between the need tomake the direct investment operational andprofitable in the short-term, and the need totransfer technologies and capabilities. Indeed,the risk of a strong imbalance arising betweenthese two incentives is considerable in LDCssince the cost of transferring technologies andbuilding capabilities is very high. As a result,the foreign firm is tempted to limit the scopeand depth of the transfer of technologies andcapabilities in order to increase the shorttermprofitability of the investment. In such acase, the TT is no longer a joint product butbecomes a by-product – that is, a negligibleobjective.When TT is the main operation, the problemis that incentives cannot be allowed to dependon another economic operation; the TT must besufficiently attractive in itself for the technologyholder to enter into the transaction. In thiscase, operational mechanisms must incentivisetechnology-owning firms to sink costs into theseoperations. In addition, classes of firms likely toserve as effective partners in such transactionsneed to be specifically targeted. This includes,for example, independent and specialiseddevelopers and suppliers of technologicalprocesses and solutions.The table below summarises the economiccharacteristics of TTs as main, joint- and byproducts.

8 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economieshigh for motivating foreign firms to commitresources; and• Domains where there is a clear demand fortechnology from local entrepreneurs, whoare then incentivised to develop innovativeprojects to suit local needs and markets.These projects are likely to generatespillover effects that are more easilycaptured by the local economy than thosegenerated through FDI and the productionof knowledge for global markets.In summary, prioritisation done as a responseto the TRIPS 66.2 provision should emphasisethose domains corresponding to a certainmodel of innovation that is central to LDCs.This includes, for instance, locally orientedinnovations that address local needs throughlocal entrepreneurial activities, which allowthe country to develop absorptive capacities.These domains are important for growthbecause the spillover generated in the courseof such projects is likely to be captured bythe local economy. Additionally, in thesedomains, TTs need additional incentives sothe donor’s intervention will be effective andrespond fully to the TRIPS provision, which isnot necessarily the case for export-orientedsectors in which the market incentives aloneare sufficiently strong to motivate firms inrich countries to operate TTs. Other areas,such as export-oriented goods processing andmanufacturing, are also important but do notneed additional incentives. Also, TTs in thesedomains are likely to generate less spilloverbecause of asymmetric capabilities betweenmultinational corporations (MNCs) and therest of the economy. As such, they should notbe listed as part of the efforts made by thedeveloped countries to comply with the 66.2provision.1.4 Towards New Types of PartnershipsThe complexity and difficulties of TT operationscall for new types of public-private partnerships(PPP). The essence of this kind of arrangementresides in the involvement of a third party, whichis specialised in linking public donors, privatefirms, and local entrepreneurial activities toensure the effectiveness of the TT operation.As such, this process will compensate for theshortcomings of existing mechanisms and whichare essential to address problems arising fromTT management as a main operation.As stated above, in the circumstances of TT asa main operation, the locus of decision-makingshould not be kept in foreign assistance bodiesbut rather transferred to local governmentinitiatives and entrepreneurs. Shifting thelocus of decision-making to local actors andauthorities is part of PPPs’ menu of tasks.The PPP has to work on the supply and demandside. In both cases, the incentive issue (i.e. tomotivate local entrepreneurs and technologyholders) has to be addressed.On the demand side, the centrality ofinnovations targeting local needs and potentiallygenerating spillover that is capturedby the local economy has been previouslyemphasised in this text. In addition, the areasin which TTs must be primarily carried outare those of goods and services that addressdomestic needs through local entrepreneurialactivity. Projects in these domains are sociallybeneficial and extraordinary advances mustbe achieved mainly in traditional sectorsthat generate local spillover. Therefore,the PPP has to target the local demand fortechnologies.On the supply side, the PPPs must accountfor the existence of a certain profile oftechnology holders in developed countriesthat can be motivated and proceed efficientlywhen beginning an LDC TT transaction. As aconsequence of a certain stage of verticaldisintegration in industries, the emergence ofspecialised segments focusing on the inventionand development of technologies, while notcompeting on the downstream market, ispositive for TT to LDCs. This is a favourablecontext to find capable and motivatedsuppliers that are likely to undertake TT in anefficient way.

ICTSD Programme on IPRs and Sustainable Development91.5 Responding Properly and Effectively to TRIPS Article 66.2The role and functions of developed countrygovernments must be examined and evaluatedwith regard to the different arguments previouslydeveloped. In short, TRIPS Article 66.2, which invague terms calls for the provision of additionalincentives to the firms and other organisations ofdeveloped countries to undertake TTs to LDCs,should be made more explicit. Practical meansto achieve this include:• Relying on the joint product and by-productslogics only to ensure a satisfactory flow ofconsolidated technologies towards LDCsdoes not suffice. Above all, TTs as mainoperations must be developed. According tothe new logic, the locus of decision-makingregarding modes of learning and areas forfocus will likely shift away from foreignbodies to local agents and authorities.• In providing additional incentives to technology-owningfirms, governments are seekingeffectiveness. To achieve this, they should:- Provide effective incentives by offeringassistance to projects that are sociallybeneficial but not profitable for thefirms that own and could transfer thetechnology;- Ensure that conditions for efficientTT operations involve the choiceof relevant partners on supply anddemand sides, the selection of theright areas for focus (related toa clearly expressed local demandfor technology) and the creationof organisational forms that willfavour the consolidation of thetransfer (absorption, adaptation andsubsequent spillover), as well as therelated entrepreneurial dynamic;• Make use, as much as possible, of PPPsas a mechanism for ensuring both theeffectiveness of the intervention and theefficiency of the TT operation.

10 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies2. THE NATURE OF TECHNOLOGY TRANSFERChapter 2 expands on the idea of the conceptualcomplexity of TT based on the distinctionbetween knowledge and information. It thenexamines the different phases that make upthis complexity. It goes on to discuss the factthat this complexity becomes even greaterwhen the transfer involves entities of verydifferent development levels and demonstratesthe pivotal nature, in the case of LDCs, ofthe notion of capabilities. The building ofthese capabilities is carried out within theframework of suitable organisational forms,such as technology platforms or productioncentres. LDCs are also distinctive with regardto the type of innovation that is importantto support as an engine for growth. Finally,chapter 2 tackles the question of barriersand incentives offered to firms to undertakeTT operations and carry them throughsuccessfully.2.1 Information and KnowledgeDiscussion of TT in this paper refers to thetransmission of technological knowledge. Ofcourse, not only product or process-relatedknowledge, but also organisational knowledgecan be transferred.The channels of transmission may involve tradeand transfers of materials, designs, blueprints,scientific papers and patents, formulae, manuals,databases, instruments and machinery, as wellas the physical relocation of people possessingspecialised technical abilities and skills. In allcases, the transferring of knowledge is different,more complex and costlier than the disseminationof information concerning a technology.Disseminating information is simpler andless costly than the activity of transferringtechnological capabilities to individuals,organisations and, on a wider scale, social aggregationsthat were previously lacking them.While a TT can be considered a success ifthe technology in question has been put intooperation and resulted in increased productivityin a certain economic activity, it is only thecreation of technological capabilities in thehost country that will ensure the long-standingefficient use of that technology in the countryconsidered.This section elucidates the content and implicationsof the distinction between knowledgeand information. Box 1 below systematicallyexplains all the properties and characteristicsof knowledge that make its transfer andmanagement difficult.As conceived here, knowledge entails somethingmore than information (Foray, 2004).Knowledge, in whatever field, empowers itspossessors with the capacity for intellectualor physical action. The meaning of knowledgehere is fundamentally as a matter of cognitivecapability. Information, on the other hand,assumes the form of structured and formatteddata that remains passive and inert until used bythose with the necessary knowledge to interpretand process them.The full meaning of this distinction becomesclear when the conditions governing thereproduction of knowledge and informationare examined. While the cost of replicatinginformation amounts for no more than the cost ofmaking copies (next to nothing thanks to moderntechnology), reproducing knowledge is a far moreexpensive process since cognitive capabilitiesare not easy to articulate explicitly or transferto others: “we can know more than we cantell” (Polanyi, 1966). Knowledge reproductionhas therefore long hinged on the “masterapprentice”system (where a young person’scapacity is moulded by watching, listening andimitating) or on interpersonal transactions amongmembers of the same profession or communityof practice. These means of reproducingknowledge may remain at the heart of manyprofessions and traditions, but they can easily

ICTSD Programme on IPRs and Sustainable Development11fail to operate when social ties unravel, contactis lost between older and younger generations,and professional communities lose their capacityto act in stabilising, preserving and transmittingknowledge. In such cases, reproduction grindsto a halt and the knowledge in question is inimminent danger of being lost and forgotten.Therefore, the reproduction of knowledge andthe reproduction of information are clearlydifferent phenomena. While one occurs throughlearning, the other occurs simply throughduplication. Mobilisation of a cognitive resourceis always necessary for the reproduction ofknowledge, while information can be reproducedby a photocopy machine.A further complication is the fact that knowledgecan be codified: that is, articulated and clarifiedin such a way that it can be expressed in aparticular language and recorded via a particularmedium. Codification involves the exteriorisationof memory. It hinges on a range of increasinglycomplex actions such as using language to writea cooking recipe, applying industrial designtechniques to draft a scale drawing of a pieceof machinery, creating an expert system fromthe formalised rules of inference underlying thesequence of stages geared to problem solving, andso on. As such, knowledge is detached from theindividual, and the memory and communicationcapacity created are made independent ofhuman beings (as long as the medium upon whichthe knowledge is stored is safeguarded and thelanguage in which it is expressed remembered).Learning programmes are then produced thatpartially replace the person who possesses andteaches knowledge.When knowledge is differentiated from information,economic problems relating to the twocan be distinguished. Where knowledge isconcerned, the main economic problem is itsreproduction and transfer (problem of learning),while the reproduction of information poses noreal problem (the marginal cost of reproductionis close to zero). The economic problem ofinformation concerns essentially its protectionand disclosure: that is, a problem of publicgoods. However, the codification of knowledgecreates an ambiguous good. This good hascertain properties of information (public good)but its reproduction, as knowledge requires themobilisation of cognitive resources.Box 1. Knowledge as a CommodityKnowledge is Sticky and TacitThere is codified technological knowledge that is quite easy to transfer since it mainlyinvolves the “costless travel” of a paper, software, or database. But new knowledge andexpertise have a broad tacit dimension, meaning that they are neither articulated norcodified. Tacit knowledge resides in people, institutions, or routines. Tacitness makesknowledge difficult to transport, memorise, recombine and learn – all operations ofhigh importance in TT activities. For instance, the knowledge involved in the effectiveproduction of advanced information technology products entails far more than thetransfer of a set of blueprints. Given tacitness, knowledge is, therefore, costly to transferfrom one site to another in a useable form. As von Hippel (1994) puts it, knowledge issticky. Knowledge stickiness refers to the incremental expenditure required to transfera unit of knowledge to a specified locus in a form usable by a given knowledge seeker.When this cost is high, stickiness is high. Stickiness raises a number of issues in terms ofthe organisation of knowledge production, product design and system integration.Industrial and technological operations draw crucially upon sets of human skills andtechniques that have been acquired experientially and are transferred between peopleby demonstration, informal personal instruction, advice and consultations, rather thanbeing reduced to explicit and codified methods and procedures. Even in industries wheremature technologies are transferred, training costs can be substantial and greater thanadaptive R&D costs.

12 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesMany production processes that are implemented by people trying to follow the codifiedinstructions without having access to the complementary tacit understanding of expertsfail to deliver outputs of the expected quality immediately, and for a long time producebelow the rates anticipated.With regard to TTs, the tacitness and stickiness of knowledge create transactionaldifficulties. The difficulties created by the complementary role of tacit knowledge insuccessfully transferring incompletely codified technological knowledge from advancedto developing countries suggests that involving firms from the former countries viacooperative ventures, technology support and training contracts will be in the interestsof firms in the latter countries.Knowledge is Partially LocalisedTechnological knowledge is often not of general value for the economy because it hasbeen produced in a local context for particular purposes. A large body of literatureargues that the production of knowledge is at least partially localised: learning thatimproves one technology may have little or no effect on other technologies. Howeverthe degree of standardisation and maturation of technology and knowledge can mitigatethese difficulties.The other facet of this property is that local environments are subject to strong variations:even stringent requirements for equipment duplication cannot eliminate all significantdifferences between local environments (Appleyard et al., 1996). As explained by Ho(1997), because a technology is developed for specific conditions, transferring it toan environment different from that for which it is designed often requires adaptationbefore it can be successfully absorbed. In fact, a technology is rarely introduced into anew environment without some adaptations.Knowledge is Weakly PersistentEvidence in psychological literature shows that people forget: if the practice of a task isinterrupted, forgetting occurs. Hirsch (1952) found that when performance was resumedafter interruption it was at a lower level than that achieved prior to the interruption.Furthermore, knowledge can be depreciated (through deterioration and obsolescence).Communities that are in possession of it can fall apart, resulting in the disintegration oftheir collective knowledge.Knowledge is often a Joint productKnowledge is produced in a context of activities in which other motivations (themanufacturing of a good or provision of a service) are predominant. People learn bydoing or by using. There is learning-by-doing because knowledge is not absolute but mustbe defined in relation to a specific physical context. Such a characteristic gives manyactivities important potential value in terms of knowledge production and innovation:those activities related, for instance, to the introduction of a novel type of equipment,organisation, or method.Knowledge is Progressive and CumulativeIn the field of science and technology, knowledge is more often than not cumulativeand progressive: technological improvement is therefore a central issue and a criticalchallenge for workers and engineers who need to recognise and exploit the latestimprovement. An organisation must thus be far more technically competent to advancewith the times (to advance with the state of the art) than to stand still.

ICTSD Programme on IPRs and Sustainable Development132.2 Consolidation of Technology Transfer2.2.1 Evidence on CostsTeece’s empirical works about TTs bymultinational companies (1977) showed thatthe transmission and absorption of the knowhowrequired to actually put the technologyinto operation involved substantial costs.These averaged roughly 20 per cent of the totalproject costs and were supplementary to thecost of transmitting knowledge in the form ofcapital goods, blueprints, specifications andspecial materials. Also included in the costwere pre-engineering technological exchanges,engineering costs associated with transferringprocess or product design, and the associatedprocess – or product –engineering, as well asR&D necessary for tailoring technology to suitlocal conditions. They also included pre-start-uptraining costs and outlays made during the initial‘shakedown’ and ‘debugging’ phases of learningthat was required for the plant to achieve itsdesigned performance specifications.Cost variations are clearly related to thefundamental objective of the TT, whether toacquire the capacity to produce a productaccording to the design and specificationsof others (material transfer), to acquire thecapacity to produce a product and make minordesign and technical changes (design transfer),or, to lessen dependence on foreign technology(i.e. to acquire the capacity to develop one’sown technology).Evidence regarding costs is a relatively goodindication of the fact that a TT involves morethan the mere acquisition of information abouttechnology. Not only must the knowledge besuccessfully transferred for the technologyto be put into operation, but persistence andmemory have to be ensured and new knowledgegenerated must adapt the technology to thenew environment and application. This newknowledge has, in-turn, to be managed,memorised, adop-ted and shared. Finally, theassimilation of subsequent knowledge improvementsis a critical requirement.Moreover, the skills necessary for puttinga technology into operation appear to beorganisational, rather than exclusively individual.Recurrent difficulties in absorbing newtechnologies during the 1960s and 1970s ledpolicymakers to define technology more broadlyfor it to include organisational and managementskills and know-how, rather than just engineeringin a narrow sense. This enabled the developmentof a greater appreciation of the importance ofacquiring “software” (disembodied informationin the form of codified instructions and directionsfor use) and above all “wetware” (what is in thememories of individuals), as well as technologyembodied in physical capital (Ho, 1997). Onefirm hiring another firm’s most skilled engineersmay be necessary but not sufficient to “reverseengineer” and successfully transfer the newtechnology (Appleyard et al., 1996).The learning capacity of the recipient is thereforevital: countries that spend relatively largeamounts on R&D in the relevant industry tendto be quick to begin producing a new producteven if they are not its inventor. Similarly, firmsthat spend relatively large amounts on R&D tendto quickly adopt new technology developed byothers (Mansfield, 1982).2.2.2 ConsolidationConceptually, the consolidation of TT involvesdifferent phases and steps: absorption andlearning, adaptation to the local environmentand needs (which is likely to include trueinnovative work), assimilation of subsequentimprovements and, finally, generalisation.The absorption phase basically correspondsto initial activities dedicated to thelearning and application of the importedtechnological knowledge, plus the maintenanceof equipment and its long-termperformance.

14 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesThe phase of adaptation to local conditions maycomprise development and innovation tasks,not entailing major discontinuity, but allowingthe imported technology to be adjusted to localspecificities, on both the supply and demand sides.The assimilation of subsequent improvementsphase signifies that the improvements of thetechnological knowledge do not of coursecome to an end once the transfer has beenmade. Sometimes, these improvements are apure extrapolation of existing phenomena andtherefore their mastery only requires an extensionbeyond the present boundaries; at other times,the improvements involve the opening up of newareas of knowledge, in which case their masteryconfronts the operating personnel with newphenomena.Finally the generalisation phase correspondsto the dynamics of adoption of the transferredtechnology by all the potential adopters. Thisgeneralisation also includes the dissemination ofimported technological knowledge.2.3. TT in the Context of LDCs: Capabilities, Organisational Design and Modelsof InnovationTechnology transfers between developedeconomies and LDCs involve very specific issuesand must be treated as an entirely differenttopic, as opposed to TTs occurring betweentwo firms with the same level of technologicalcapacities and management capabilities, forexample.TT between developed economies and LDCsinvolves the transfer, implementation andabsorption of a technology from a maturetechnological structure to an entirely disarticulatedproduction and knowledge system.Each consolidation phase – absorption oftechnology, adaptation to local conditions,assimilation of subsequent improvements, andgeneralisation – causes significant difficulties.As opposed to a TT involving two entities of the2.3.1 Capabilitiessame level of development, greater attentionmust generally be paid to:• Learning and training services;• What is imported in terms of equipment(some sources of failure are importation ofthe wrong equipment, incomplete sets ofequipment, or inappropriate equipment);• The sufficient availability of high qualitylocal raw material and components (amajor source of failure is the inadequatesupply of local raw material andcomponents (see Ho, 1997));• The local demand structure (productsunsuited to the local market constituteanother source of failure).The term “capabilities” here draws on Enos(1996) to introduce an issue relevant for theleast technologically advanced firms. The mosttechnologically advanced firms can profitablyabsorb new knowledge and subsequent improvements,and undertake development to adapt thetechnology to specific conditions. They employthe skilled persons needed to appreciate andassimilate advanced technologies, and can drawupon their previous experience in carrying outeach successive task. Technology transfer is infact a decreasing cost activity. The more extensivethe experience previously acquired by thetransferring organisation in supplying subsidiarieswith the technology in question, the narrowerthe gap between the technical capabilities of thetwo participating organisations, and the lowerthe transfer costs in relation to total project size(Mansfield, 1995).Less advanced firms lack these prerequisites fortechnological progress: even if they draw uponoutside suppliers for the tasks of planning, design,engineering, construction and initial operation,they are likely to find themselves incapable ofoperating the plant in a way that exploits its fullpotential, let alone securing the mundane dayto-dayimprovements that so markedly increase

ICTSD Programme on IPRs and Sustainable Development15its performance. It may take all the technicaland managerial resources of the less advancedfirms to master the transferred technologyand implement the necessary adaptations anddevelopments. Mastering improvements as theycome along may prove too great a challenge.Mastering a given state of the art is not enough;what is critical is to master a progressive state ofthe art (Enos, 1996). In the knowledge economy,these tasks are never-ending. No sooner haveworkers mastered one state of the art that theymust begin to shift their attention to its successor.Improvement can occur so rapidly that workerscan never relax thinking they have absorbed thecurrent set of knowledge, as the next phase ofimprovements is already upon them.Building capabilities to increase a TT’s chancesof success is thus a crucial matter. It pointstowards economic models of developmentthat emphasise the accumulation of skills andlearning capacities, rather than fixed assetsor capital, in facilitating the TT process. This,in turn, calls for certain proper organisationalstructures.2.3.2 Organisational designOrganisational structures are critical for thesuccessful management of the whole TT process.What is at stake here is an idea that goes back toA. Marshall’s concept of the internal and externalorganisation of firms. The internal organisationof the receiving entity is central to the processof technology absorption and its adaptation tolocal conditions, as well as the assimilation ofsubsequent improvements. External organisationis critical to the process of broader disseminationand spillover from one particular entry point ofthe technology into the country.These two organisational dimensions are likelyto be “weak” in the case of LDCs. It is thereforecritical to establish organisational structuresthat are dedicated to improve both internal andexternal dimensions in order to maximise theprobability of TT success.The terms “technology platform” or “productioncentre” are used to designate forms of organisationexplicitly aimed at facilitating the learningof the technology, its adaptation to localconditions, the assimilation of subsequentimprovements and its generalisation. Theseessentially involve technology developmentcentres devoted to a specific domain and partlyfinanced by public development assistance.These centres provide a certain number oftechnological services to assure the developmentof appropriate innovations. They pinpoint andstructure demand for technology from localentrepreneurs. They also ensure the updatingof technological knowledge and its diffusion.Crucially, they facilitate access to the financingof innovation by local banks, either by simplysupporting the project in question, or creatingcredit lines from developed countries. Box 2describes this type of project using the exampleof the Cleaner Production Centers set up aspart of collaboration between Switzerland’sState Secretariat for Economic Affairs (SECO),the United Nations Industrial DevelopmentOrganisation (UNIDO) and several developingcountries.Technology platforms constitute an attractiveorganisational innovation when the TTinvolves countries with different levels ofdevelopment. They represent a method ofcoordinating and adapting resources whoseassembly is, by definition, problematic.They provide a better understanding oflocal technology demands. Finally – andperhaps most importantly – they “anchor”the technological development in the localeconomy, endeavouring to attach it to anindustrial dynamic.A TT linking entities with very different developmentlevels calls for sophisticated organisationalforms if the success of the differentconsolidation phases is to be guaranteed.However, these organisational forms arethemselves heavily dependent on the chosentransaction modes, as discussed further below.

ICTSD Programme on IPRs and Sustainable Development172.3.3.1 The centrality of local innovation and local spilloverInnovation should be widely distributed overthe whole spectrum of economic activitiesacross sectors (not just high tech) and typesof innovations (not just formal R&D). In LDCs,this means incremental, cumulative, andmostly informal (without R&D) innovations,developed mainly in “traditional” sectors or inservices not qualifying as “high tech.” Althoughmostly dealing with low-tech activities, theseinnovations are generating local spillover andwill ultimately impact the productivity of awide range of sectors in the local economy.If we regard information and communicationtechnologies (ICTs) as the major generalpurpose technology of our time, ever expandingsegments of LDC economies should adopt and“invent” new applications for ICTs in waysthat increase productivity. General PurposeTechnologies (GPT) foster economy-widegrowth not simply or principally by innovationin the GPT itself, but rather when a wide andexpanding range of other sectors adopt theadvancing technology, generating new usefulapplications of the GPT. Therefore, the keyissue for “secondary countries” (countriesthat are not at the frontier of the GPT) is howto allocate R&D and other innovative inputsso as to level the growth potential of theprevalent GPT. The main point is that it is notICTs alone that cause growth, but rather thatadopting sectors ought to establish innovationcomplementarities for economy-wide growthto take place. These types of innovationcomplementarities (adoption, local innovationsin traditional sectors) may be less overtlyinnovative and therefore not be deemed asworthy of support or encouragement. Yet,ultimately, they constitute the key to economicgrowth.Any innovation policy in the case of LDCsshould, therefore, pay attention to theseissues. It should not aim just at increasingtotal R&D, but do so in a way that incentiviseslocal innovation and spillover rather thanglobal R&D and external leakages. Such apolicy should develop absorptive capacity andultimately impact the productivity of a widerange of sectors in the local economy.This is the “model” of innovation that TThas to foster in an LDC context. From thisperspective, government and donors shouldpay more attention to local demand fortechnology. Much of the discussion in TTliterature has focused on the supply side – thewillingness of technology holders to transfertechnology (Arora, 2007). Very little attentionhas been paid to providing better knowledgeof the structure of the demand for technologyin a given country, region, or industry.2.3.3.2 Discovering the next areas for focusAccording to Hausman and Rodrick (2002), thereis a key role for entrepreneurs in LDCs: to learnwhat the country is good at producing. For anLDC, there is great social value in discovering therelevant specialisation since this knowledge canorient the investments of other entrepreneurs.It is also a misperception of LDCs’ realities toassume that the production functions of allextant goods are common knowledge.But the entrepreneur who makes the initial“discovery” can capture only a small part ofthe social value generated by this knowledge;other entrepreneurs can quickly emulate suchdiscovery. Consequently, entrepreneurship ofthis type – generating learning on what can beproduced – will be undersupplied.If learning what a country is good at producingrequires an investment and the return on thisinvestment cannot be fully appropriated, this isa problem that is unlikely to be solved with legalprotection. Indeed, entrepreneurs in LDCs aretrying out technologies that already exist abroad.The discovery may be that an existing good canbe produced profitably in the country. Suchdiscovery does not normally get legal protectionno matter how high the social return. There istherefore a role for government policy, probablynot fulfilled by the IP protection system.

18 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies2.4 The Choice of Technology Transfer Areas in the Case of LDCsIn a note-worthy case study of science andtechnology policy in four sub-Saharan countries(Ghana, Kenya, Tanzania and Uganda), Enos(1998) clearly demonstrates that the areas ofscience and technology most vigorously pursuedin these countries are increasingly determined byforeigners. Tanzania’s Agricultural Master Plan,for example, assigns the first R&D priorities tocoffee, tea, rice, animal health, soil and watermanagement, and ‘farm system’ research,in that order. Of course, setting priorities,planning, and allocating resources to undertakeresearch according to these priorities are highlydesirable policy actions. However, the questionis: on what bases are the priorities set? Who doesthe setting? Is there any impact?2.4.1 Enos’ main argument: the process used to determine priorities is dubiousAs described in Box 3, Enos has six strongarguments to describe the vicious circle forthose economies in which the decision locus hasshifted to foreigners and where the re-focus ison scientific and technological progress in thearea of export-oriented commodities facingdeteriorating terms of trade instead of areas ofdomestic expansion.Box 3. Defining the Areas of Focus: Six Arguments1. Enos argues that there is a shift in the locus of decision-making concerning the futuredirection of economies from local authorities to foreign assistance bodies. This is likelyto continue for a long time since local resources devoted to the pursuit of scienceand technology are scarce and it may well be decades before the countries will havesufficient resources to finance scientific education and R&D without foreign support.Let us go back to Tanzania’s selection of target areas in science and technology tounderstand what is going on. Obviously the criterion used to select R&D in coffee, cotton,and tea in that order is based on the country’s export specialisations that provide thebiggest export earnings for financing national development. The argument is as follows:the countries are short of foreign earnings with which they finance their development.An increase in the capability of these countries to provide the goods and services desiredabroad will help grow their foreign earnings. And, the main foreign markets for thesegoods and services have always been, and will remain, in the developed countries. Theargument is thus that these economies will best serve their own members by conformingto the existing pattern of world trade.2. Decisions made by foreigners regarding the direction taken by science and technologyare primarily based on the principle of comparative advantage. A country should investin activities that can potentially be carried out at a relatively lower cost than inother countries. So some activities, such as production of textiles and processing oftropical foodstuffs, are excluded from the menu of activities appropriate for developingcountries (they are reserved for producers in developed countries). As a result, themenu of activities available is very restricted. It essentially includes those commoditiesthat are consumed but not produced in developed countries: primary commodities suchas beverages (cocoa, coffee, tea) and, to a lesser extent, fibres (cotton and wood).Scientific research is undertaken to increase the production and lower the costs ofproducing these commodities. Education is directed towards training people who willorient their careers accordingly.3. The choice of science and technology areas to be pursued is primarily based uponthe effects of such choices on developed countries. This means that among the variousareas where Tanzania has a comparative advantage, the commodities accorded highestpriorities are those that are not produced by the countries sponsoring the master plan,which includes the EU, Germany, the Netherlands, and the UK. Not only do developed

ICTSD Programme on IPRs and Sustainable Development19countries support R&D in those commodities that they import but they also tend toensure that these are the very commodities that receive the most R&D.Tanzania certainly has a comparative advantage in coffee or tea. As such, this is ageneral argument to allocate resources for technological development to these areas.However, Tanzania may also have a comparative advantage in the production of maize,roots and tubers, oil seeds, and etc., which are considered as a medium priority in themaster plan. R&D in these areas will thus receive far less attention.Why is this happening? From the donor point of view, there is the need not to offendpowerful groups of producers in the developed countries, the universal aim of exportpromotion, and the reliance of the country considered on competitive advantagecriteria.4. The implication of arguments one, two, and three above on the shift in decisionmakings,comparative advantage rationale, and the dominance of effects on developedcountries, is the change in the aim of economic development; the shift is from domesticexpansion to export promotion. Export promotion involves focusing on world markets,discovering those products most in demand, and mobilising domestic resources for thatpurpose. Between domestic expansion and export promotion, there is an obvious clash,at least in the short-term.5. Focusing on technological progress in a given area of export-oriented commodity is acorrect decision for a single country (in the absence of other countries adopting the samedecision). But it is likely to be a wrong decision when many countries adopt it. The chainof events is well known for commodities for which demand is price inelastic: the totalvalue of the marketed commodity falls. Second, the country has reallocated resourcesto an activity that yields fewer returns as time passes. The cost of this reallocation maybe high and is sunk. Third, the country is left with a monoculture.6. It is unfortunate that these countries are allocating most of their resources for scienceand technology to areas facing deteriorating terms of trade. Devoting scarce R&D totheir improvement will merely accelerate this deterioration. R&D and its concomitantactivities should therefore be directed elsewhere.The severe deterioration of terms of trade is thus an emerging property of demandinelasticity and the successful transfer of technologies carried out to improve similaractivities in too many LDCs.2.4.2 The question then is “where else?”In principle, the answer is quite simple and wasprovided in the preceding section (2.3.3.1).That is, greater advances should be sought inthose activities that provide employment for arapidly expanding population and generate theproduction of goods and services for domesticconsumption. Thus the aims of advancing scienceand technology are: modest gains for exportcommodities characterised by deterioratingterms of trade, and immodest gains for nontradedgoods and activities like tourism, whichuse relatively abundant resources to earn foreignexchange. This last area could be referred toas “stable-terms-of-trade goods” in order todistinguish it from exportable commodities likecoffee whose terms of trade will continue todeteriorate in the future.To summarise, the aims of science and technologypolicy in LDCs are to make reasonable advancesin the area of exportable commodities andextraordinary advances in the areas of goodsfor domestic consumption and stable-terms-oftradegoods.

20 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesTable 2. Mapping the Areas for TT focusProduct typeExamplesCurrentproductionpotentialR&DcontributionDemandelasticitySupplyelasticityTerms oftrade (trend)TraditionalexportablecommodityCoffee Excellent Moderate Low High DeterioratingTraditionallocallyconsumedprimarycommodityCottonExcellentModerate tosubstantialwithsufficientR&DMedium High StableLocallyprocessedperishablegoodsSoap(manufacturing)cookingoils (refining)GoodModerate,only if R&DdirectedtowardsappropriatetechnologyMedium High StableSimpletraditionalcapital goodsand servicesAgriculturalmechanicalFairSubstantialonly if R&DdirectedcorrectlyMedium High StableSource: Enos, 1998TTs have to be operated in many domains,including export-oriented industry. But theymust be particularly supported in those domainscorresponding to the model of innovationthat is central to economic growth in LDCs.This includes entrepreneurial activities thatmeet the needs of local markets likely togenerate domestic spillover. In other words,TTs must offer a positive supply responseto a demand for technology stemming fromlocal entrepreneurs. Two reasons for this areprovided:• First, these domains are potentiallyimportant for growth because the spillovergenerated in the course of such projects arelikely to be captured by the local economy;• Second, these domains need additionalincentives so that the donor’s interventionwill be effective and will respond fully tothe TRIPS provision. This is not necessarilythe case of export-oriented sectors in whichthe market incentives alone are sufficientlystrong to motivate firms in rich countries tooperate TTs (see chapter 5 below).2.5 Transactional Modes, Barriers and Incentives 4The major sources of transfer, as described in theliterature (see e.g. Maskus, 2004, World Bank,2008) include trade, FDI (i.e. MNC), licensing,joint ventures, and the movement of people.Among these various forms of TTs, there aretwo logics to distinguish that make sense whenorganisational issues, as well as incentivesand barriers are considered: packaged andunpackaged forms of TT (Enos et al., 1998).The packaged form essentially means that thetechnology is transferred through FDI, importof goods, or the building of infrastructures byforeign firms. In other words, the transfer oftechnology is a joint product or a by-productof another economic operation. The importanceand quality of TTs are therefore contingent on alarge number of factors resulting from the primaryeconomic operation. It is thus the incentives forsuccess of the primary economic operation that

ICTSD Programme on IPRs and Sustainable Development21determine the motivation for success of the TTand its consolidation (or not).Unpackaged is defined as the transfer of atechnology not channelled through directinvestment, trade, or infrastructure development.Joint ventures, licensing, andarrangements involving technical assistance,collaboration contracts, informal transfer ofknow-how, and consultancy are classified asunpackaged modes of TT. The technologicalknowledge is “disembodied” and transferredthrough software and wetware, not hardware.In this case, the TT in itself constitutes theprimary economic operation. Here, the primemotivation for the operation is the success of theTT and the incentives directly linked to the TT(cost and profit) that control the operation. Themain advantage of this mode is that it gives thehost country control over technology selection,management decisions, and development oflocal skills.Discussion here of TT as a joint product,by-product or main operation refers to thedefinition of these concepts in accounting.Joint products are two products that aresimultaneously yielded from one shared costand they have comparably high (sales) value.By-products for their part are produced alongwith a main product. The latter constitutesthe major portion of the total (sales) value.By-products have a considerably lower (sales)value than these main products. These conceptscan be applied to TTs with the substitution of“perceived value to technology holders” for“sales value.”It is therefore apparent that the technologyholder’s commitment may vary considerablydepending on the TT’s economic operationstatus and the importance attributed to itin the success of the primary operation,if it does itself not constitute the primaryoperation.In the case of TT that is a by-product, thetechnology holder is interested in the successof the main economic operation. Here, realisingthe TT is of secondary importance. In the caseof a joint product, the issue is to balanceincentives between the success of the othereconomic operation (e.g. a direct investment)and the success of the TT. Finally, in the case ofTT as a main operation, the economic incentivescannot be allowed to depend on anothereconomic operation. The TT must be sufficientlyattractive in itself for the technology holder toenter the transaction.These three forms of TT (joint product, byproductand main operation) are examinedbelow, along with the various channels implementedfor each. The incentive structure, barriers,limitations, and opportunities associatedwith each type are also examined.2.5.1 TT as a joint producr or by-product: the need for balancing incentivesThis section briefly examines TT mechanismsas related to trade and direct investment.In both cases, TT is a joint product and assuch creates both opportunities for knowledgetransfer and limitations.2.5.1.1 TradeTrade as a channel for TT involves imports ofgoods and services — especially capital goodsand high tech products – and export by firmsfrom lesser developed countries. The role ofhigh tech products and capital goods import asa TT mechanism is clear: it directly impacts thedomestic economy by improving productivity andit can foster local innovation through reverseengineering of imported high tech products.Exporting goods creates an indirect mechanismfor TT. By participating in the global value supplychain, LDC firms will benefit from numeroustraining and technological spillover effects fromtheir customers.One interesting form of trade-related TT issubcontracting, whereby the subcontractormanufactures the final product under the

22 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesprincipal’s brand name (Enos et al., 1998). Thisallows foreign involvement without the transferof ownership. Such an arrangement often involvesthe foreign partner in selecting capital equipment,training managers, engineers and technicians,and advising on production, financing, andmanagement. Since the quality, delivery, andprice of the final product are critical for theforeign investor, this is likely to generate longtermtechnical relationships for capacity buildingin the host country. Here incentives seem to bealigned. Moreover, learning and constructingtechnological capabilities often occurs (Hobday,1995). For example, many electronic systemspurchased under this form of TT (subcontracting)were designed, specified, and manufactured bythe local firm rather than the foreign company(see the case study on the electronics industryfrom the Asian new industrial economies).The transfer of technology through thisarrangement nevertheless has certainlimitations. It is difficult for the company in thedeveloping country to establish internationalbrand images. Dependence on foreign companiesfor technologies and components can persist fora long time. On the one hand, evidence showsthat substantial learning can take place throughthis mode. Yet, on the other, it seems difficultto overcome the limitations arising from thedependence of firms in developing economieson this mode of technology acquisition.2.5.1.2 Direct investmentsA lot of TT occurs through foreign directinvestment. FDI is an efficient transfermechanism because it provides the necessaryincentive and long-term framework to makeTT effective. When FDI is involved, theforeign participant has a direct interest in thesuccessful TT (Ho, 1997). Consequently, it islikely to participate actively in the transferprocess, supplying not only equipment andcapital to its foreign affiliate but also highqualitydisembodied knowledge, particularlytacit knowledge such as general technical,business, and managerial know-how. In otherwords, the foreign investor has an incentiveto supply much of the missing technical andmanagerial expertise and know-how neededto start up the project promptly and help itto rapidly reach its designed production rate.Incentives are aligned.There is abundant empirical literature thatfocuses on the dissemination effects ortechnology spillover from FDI (Blomström andKokko, 1998). Spillover occurs when localfirms benefit from the MNC affiliate’s superiorknowledge of technologies without incurring acost that exhausts gains from improvements.They operate via different mechanisms aselaborated on below:• Backward linkages: This arises from the MNCaffiliate’s relationships with local suppliers.Complementary activities developed throughthese linkages are likely to createspillover effects and several case studieshighlight such positive outcomes. However,they also stress that the local content ofMNC production is a strong determinant ofstrength of linkages.• Forward linkages: This stems from contactswith customers. However, there is muchless evidence of forward than backwardlinkages, so that the argument pointingtoward the existence of FDI spillover towardcustomers is suspicious.• Training of local employees in MNCaffiliates: Although it is obvious that FDIinvolves some supply of training for localemployees ranging from on-the-job trainingto seminars, and more formal schooling tooverseas education, perhaps at the parentcompany, the evidence on spillover fromMNC affiliate training of local employeesis far from complete. There is certainly anaccumulation of human capital skills in theMNC’s employee stock. Some of these skillscan be appropriated by local firms whenemployees move to new jobs, but howmuch is an open question.• Demonstration effect: A few case studiessuggest that demonstration (when affiliates’

ICTSD Programme on IPRs and Sustainable Development23technology imports have induced localcompetitors to imitate their behaviour)may be an important channel for spillovers.However, there are too few studies to revealhow important the simple demonstrationeffects are.To summarise this literature, potentials forknowledge and technology dissemination fromFDI are obviously significant. But many conditionsare necessary to realise it fully. Some of theseconditions relate to the host country: marketsize, local content regulations, and the sizeand technological capability of local firms arecountry characteristics that will clearly influencethe extent of spillovers.FDI as a vehicle for TT also exhibits someshortcomings that create challenges for theimporting country. First, the decision concerningwhich technology to transfer and how toorganise the production of the knowledgerequired for development and local adaptationremains in the hands of the foreign investor.The transferred technology could thusconceivably be inappropriate for the hostcountry – for instance, too modern for its needsand too capital intensive given the resourceendowments of most developing countries. Itis also likely that foreign investors will do mostof their research and development in theirhome countries, preventing the developmentof core technologies in host countries (Enoset al., 1998). Will the developing economiesget the right technologies on the right terms?If the main vehicle for transfer is foreigninvestments, the main TT aspects will notbe decided publicly, via the intervention ofgovernment agencies, but privately, by thefirm. There is therefore the risk that the shiftfrom external and collective to internal andindividual choice of technology may removethe transfer of technology from the publicarena (Enos et al., 1998).2.5.1.3 Balancing incentives: from joint product to by-productAre incentives really aligned (the advantagemost frequently advanced)? Foreign investorsprimarily want to succeed in putting the plantinto operation and keeping it running for acertain period of time. The success of the TT islikely to become of secondary importance.If incentives are not properly balanced betweenthe need to make the industrial facilities operateefficiently and the need to transfer trainingand knowledge to the future local workers andengineers, it is likely that the foreign investorwill devote insufficient resources and time to thelearning process. The whole range of capacitiesand capabilities (including tacit knowledge)has to be absorbed by the nationals of theimporting country. But what matters most forforeign investors is the success of the industrialoperation and not the success of the transferin itself. For example, Choii et al. (1994) arguethat foreign investors have little incentive totake the initiative in shifting responsibility fortechnological adaptations to local suppliersor staff. If the replacement of expatriates isunnecessarily delayed, this prevents the learningprocess from fully taking place. This is a clearcase of unbalanced incentives between the needto make the investment operational in the shorttermand the need to transfer the technology.In this case, TT becomes more like a sort of byproduct(as such a very uneven and unplannedevent) rather than a joint product.2.5.2 Technology Transfer as Main operationJoint ventures, licensing, and arrangements forTT as technical assistance, such as collaborationcontracts and the informal transfer of knowhowhave been the most popular alternativesto wholly foreign owned MNC subsidiaries assources of technology (Enos et al., 1998). Jointventures are contractual arrangements betweentwo (or more) firms in which each provides someadvantage that should reduce the costs of jointoperations. For instance, the MNC will make thenew technology available while the domesticfirm provides its knowledge of the market, theregulatory and business environment, and someother local advantages.

24 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesIn this case, TT is the main operation andtherefore the incentives are shaped by thecost and benefits of the TT only. In otherwords, incentives cannot be leaning againstanother economic operation. The TT must beattractive enough for the technology holderto enter into the transaction. As a mainoperation, the TT provides an opportunity toshift the locus of decision-making away fromthe foreign bodies to local agents both in termsof the learning process and the areas in whichTTs will operate. TTs as a main operation givethe host country control over managementdecisions and the development of local skills.This is why many countries have expresseda preference for joint ventures, with theforeign partner in a minority position, overwholly FDI.2.5.2.1 LicensingLicensing involves the purchase of productionrights, protected by IPRs, and in many cases,the provision of technical assistance and knowhow,which are needed to adopt and adapt thetechnology. The transfer of tacit knowledge andthe provision of technical services are centralto ensure that the licensor will secure theproper capabilities so as to use the technologyin an effective way. For instance, studies oflicensing in India suggest that Indian firmsimporting technology tend not to receivesufficient amounts of technological know-how.As a result, their ability to assimilate, utilise,and improve the technology is limited. (Arora,1996). Theoretical works (notably by Arora,1995, 1996) demonstrate the conditions underwhich know-how can be transferred throughcontracts. One important condition is tobundle know-how transfer with the provision ofcomplementary inputs.Licensing is becoming central among thevarious market-based channels as marketsfor technology become thicker and more“efficient.” Athreye and Cantwell (2007) findthat international patent licensing and royaltyreceipts have surged since the mid-1980s. Fromaround USD 10 billion in 1984, internationalpatent licensing and technology receipts grewto more than USD 80 billion in 2002 (at currentprices). Over 120 countries reported receivingsuch royalties and more than 130 countriesreported making such payments in 2002. Therehas therefore obviously been a significantincrease in international TT through licensingoperations (see Arora, 2007, for an up-todateoverview). What is more, this appearsto coincide with an overall strengthening ofIP regimes, first in rich countries but later inpoorer countries too.Some countries have pursued a licensingbasedstrategy of technology acquisition in thebelief that it is more favourable to TT and thedevelopment of learning capacities by domesticfirms than FDI. Some sparse evidence (Enos etal., 1998) tends to show that joint venturesprovide greater training than licensing, while thelatter involves greater indigenous technologicaleffort.However, one should not forget that marketsfor patents or licenses are not markets fortechnology. Most firms want to purchasetechnologies, not mere patents. For this reason,patents are sold or licensed as part of the bundleof goods and services needed to implement atechnology (see Arora’s argument below). Butthis sort of transaction is more complicated andcostly than a mere patent license or assignment.In addition, in view of tacit knowledge, andthe importance of collaborative development,it is unrealistic to expect technology marketsfor today’s complex inventions to function aswell as the robust markets during the heydayof the individual inventor (Bessen and Meurer,2008). Therefore, the various phases of a TT,notably absorption, learning, adaptation, andassimilation of the subsequent improvementsare complex to predict and to plan. Theyare therefore difficult to draft as detailedobjectives and milestones in the contractualarrangement.

ICTSD Programme on IPRs and Sustainable Development25Figure 1. The Growth of Royalties and Licensing FeesSource: Athreye and Cantwell, 20072.5.2.2 TT as a main operation: cost and benefitsIn TT as a main operation, there are no othereconomic operations to “help” and the prospectof returns on the TT operation must onlybe sufficiently attractive to incentivise thetechnology holder to enter the transaction.This means that very often additional incentivesprovided by governments will be needed(chapter 5, below). But TT as a main operationoffers opportunities to shift the locus for decisionmakingto the local agents and domestic firms.When TT is a joint product of FDI, the locus ofdecision-making lies in foreign firms that decideabout the domain, the modes of learning, theamount of resources devoted to the TT operation,the potential scope of further dissemination,the objectives of assimilating subsequentimprovements, and etc. Clearly, when the locusof decision-making regarding areas for TTs lieswith foreign firms, there is a risk of a suboptimaldecision process (see the argument above in thecase of FDI). 5When the TT is the main operation, there ismuch more flexibility about who will make themain decisions: domestic bodies (entrepreneurs,governmental agencies), foreign donors or foreignfirms. The argument is that in the circumstancesof TT as the main product, the locus of decisionmakingshould not be kept in foreign assistancebodies but transferred to local governmentinitiatives and entrepreneurs. This is particularlyimportant for decisions that concern:• The mode of learning and the “quality” ofTT. To what extent conditions for effectivelearning and transfer of capabilities will beorganised by the technology holder? and;• The choice of areas to be concentrated on.TTs address specific knowledge areas andtechnological trajectories. The issue of selectioninvolving both priority-setting andtechnology forecasting must be addressedsince it is a critical one: a successful TT, ifnot done in a relevant area for the countryconsidered, will produce very little benefit.It is thus imperative to critically examinethe conditions and procedures of prioritysettingand technology selection. Section2.4 above already clarified in what domainsTTs should be undertaken first. Thisargument will be fully developed below (inchapter 5).

26 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies2.5.3 Technology transfer without compensation: imitation and reverse engineeringWhat distinguishes this mode of operation fromlicensing and joint ventures is that imitation doesnot entail any compensation to the technologyowner. This does not mean that it is a costlessprocess, however. Imitation can be very complex– it takes time and resources, which makes itperfectly tolerable from a competition pointof view. The initial inventor can use the “leadtime” (the gap between the invention and theentry of imitators) to capture a large fraction ofthe benefits and thereby cover R&D fixed costs.Being the first is an asset that can command apositive price under competitive conditions. Thisprice reflects the present value of the futureflow of marginal utilities that subsequent copieswill yield to impatient consumers because theprocess of imitation takes time.It is because imitation is time consuming andcostly that Samuelson and Scotchmer (2001)argue that reverse engineering has generallybeen considered a positive element for fosteringinnovation and the transfer of knowledge. Reverseengineering is fundamentally directed todiscovery and innovation. Engineers learn thestate of the art not only by reading publications,going to conferences, and working on projects,but also by reverse engineering others’ products.Learning what has been done before often leadsto new products and advances in know-how. Ifreverse engineering is costly and takes time, asis usually the case, innovators will be protectedlong enough to recoup R&D expenses. As such, itis a competitively healthy way for second comersto access and discern the know-how embeddedin an innovator’s product.Samuelson and Scotchmer (ibid.) also arguethat the very act of reverse engineering rarely(if ever) has destructive effects on the market.Harmful effects are far more likely to resultfrom post-reverse engineering activities, suchas selling a competing product made with knowhowfrom an innovator’s product. Because ofthis differentiation, policy has at least to ignorethe act of reverse engineering while focusing onregulatory controls on market-destructive postreverse-engineering activities.2.6 Brief SummaryThis section has put a framework in place basedon two key taxonomies:• The phases of TTs: absorption, adaptation,assimilation of subsequent improvements,and dissemination (spillovers); and• The economic nature of TTs: by-product,joint product, or main operation.When the TT is a joint product (occurring as aconsequence of a direct investment), the phasesof absorption, adaptation, and assimilationof subsequent improvements are in a senseembedded in the investment project. However,since the TT is a joint product, there is theproblem of “balancing incentives” between theneed to make the direct investment operationaland profitable in the short-term, and the need totransfer technologies and capabilities. Indeed,the risk of a strong imbalance arising betweenthese two incentives is considerable in LDCssince the cost of transferring technologies andbuilding capabilities is high. As a result, foreignfirms are tempted to limit the scope and depthof the transfer of technologies and capabilitiesin order to increase the short-term profitabilityof the investment. In such a case, the TT is nolonger a joint product but becomes a by-product– that is to say a negligible objective.When TT is the main operation, the problem isthat incentives cannot be allowed to dependon another economic operation. The TT, inthat case, does not offer other benefits to thetechnology holder than what is generated throughthe transaction. On the other hand, it providesopportunities to shift the locus of decisionmakingtoward local authorities and agents bothin terms of the choice of areas for focus and themode of learning to be implemented.In both cases (joint product and main operation),the broad dissemination of the technology andthe spillovers stemming from a given TT locusare not guaranteed. The successful achievement

ICTSD Programme on IPRs and Sustainable Development27of this final phase of the TT requires particularorganisational forms.This also underlines the peculiar nature ofinnovation in LDCs that TTs have to serve(demand side). The centrality of a particularmodel of innovation for growth and developmenthas already been emphasised. R&D and othermore informal learning activities undertaken toproduce locally oriented innovations allow thecountry to develop absorptive capacity, while atthe same time the locally generated spilloversfrom this same R&D may end up diffusing awayfrom the local economy. TT as a main operationprovides the opportunity to target those areasin which such innovative activities can beundertaken.Chapter 4 uses recent evidence to observethe mechanisms (in particular FDI, trade andlicensing) that are effectively operating in theLDC context and their relative importance andimpact on economic growth and development inthese countries.

28 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies3. TECHNOLOGY TRANSFER AND INTELLECTUAL PROPERTY RIGHTS3.1 The Conditions for Technology Transfer to LDCs under TRIPSAs a new international system of IP protectionhas slowly been implemented as part of WTOAgreements to guarantee the enforceabilityof IPRs globally, policy discussions have beenreinvigorated. This places significant pressureon countries that decided one or more decadesago to limit the scope of IP protection tofacilitate knowledge access for both immediateconsumption (drugs) and learning (technologicalknowledge).3.1.1 An overview of TRIPSThe TRIPS Agreement establishes global mandatoryminimum standards for the grantingand protection of IP rights in several areas,particularly copyrights and patents. It alsoprovides for strong enforcement measures anda settlement of disputes mechanism. Countriesare free to decide how to implement theseprovisions according to their own legal systemand practices. The application of TRIPS indeveloping countries has been mandatory since2000, except for technologies not previouslyprotected by-product patents. However, LDCsare benefiting from a transition period until2013, with a further extension until 2016 forpharmaceutical products.Both the US and EU benefit economically fromstronger foreign IP rights, as this strengtheningwill lead to an increased flow of royalties orprofits in the entertainment, pharmaceutical,and other industries. The introduction of aglobal system to enforce IPRs is a positivedevelopment for multinational companies aswell, which expect an increase in the profitabilityof their R&D investments. These industriesare therefore placing strong pressure on thenegotiators of the developed nations. Hencethere is a sense that if stronger IPRs cannot benegotiated at the World Intellectual PropertyOrganisation (WIPO) or the WTO, they shouldbe obtained via bilateral agreements. Whilethe TRIPS Agreement established minimumstandards of IP protection for WTO Members,substantial efforts are being made, particularlyby the United States, to negotiate bilateralarrangements to limit part of the autonomy leftto developing countries by TRIPS to go beyondits required levels of IP protection, therebystrengthening IP rights.On the other hand, the official TRIPS messageis that developing countries should considerthis legal obligation as a positive arrangement.In a knowledge economy, strong IPRs arepresented as beneficial for developing countriesthemselves since this evolution will createincentives for the endogenous development ofdomestic entrepreneurial capacities and helpthese countries to attract more FDI and absorband acquire more foreign technologies throughlicensing and other market-based transactions.It is time to revisit this economic discussionwithin the new context of a global IPR systemunder construction. Regarding TTs, the questionis how IP protection effects the different transfermodes as described above. There are two relatedquestions here: first, how does patent protectionaffect TT within a given mode? Second, how doespatent protection affect the choice betweenthese modes? Arora (2007) suggests that therelative importance of the different sources mayvary over time, and that IP protection may affectthese sources differently.3.1.2 Transformation of the institutional conditions for TTs under TRIPSMaskus (2004) proposes a distinction between“market-based forms of TT” (including FDI,technologies embodied in imported goods,licensing, exports by firms in the recipientcountry) and “non-market-based forms of TT”(including imitation and reverse engineering,learning from patents filed by foreign inventors,learning from scientific publications, movementof workers and students, and other types ofspillovers). Maskus (2004) shows that certain“market-based” channels have increased inimportance over time.

ICTSD Programme on IPRs and Sustainable Development29Simple reasoning, plus some sparse empiricalstudies, would allow the following developmentsto be identified. The strengthening of IPprotection can be expected to result in anincrease in the relative importance of licencesand direct investments, and in the quasidisappearanceof imitation. It can also be anticipatedthat within a specific mode – suchas, direct investment – the TT will be moresuccessful insofar as a strengthening of IPprotection will increase the technology holder’sincentives to transfer in an efficient manner.Some transfer modes would thus practicallydisappear, while others would be strengthened.The TRIPS philosophy is basically to reduce therange of transfer modes, whilst consolidatinginstitutions that facilitate a small number ofmodes, at the risk of eliminating others (seeBox 4).Box 4. TRIPS and the Public Domain of KnowledgeTRIPS has considerable potential to increase the “excludability” of R&D results andreduce knowledge diffusion and informational spillovers (Foray, 2004).By focusing on licensing as the main means of transferring technology, TRIPS isconceptually based on a narrow view of the channels through which knowledge candiffuse. In reality, these channels are multiple and all contribute to the transfer ofknowledge, while the incentives created by TRIPS promote only one channel (patentingand licences), entailing the risk of blocking the others. Focusing exclusively on providingbetter IP protection is likely to cause serious collateral damage to other complementaryinstitutions that support TTs in different ways.The fact that the diversity of institutional arrangements is threatened constitutes acause for concern. Traditionally, IPRs are considered as one of the incentive structuressociety employs to encourage innovative effort. They co-exist with other incentivestructures, each of which involves costs and benefits as well as a certain degree ofcomplementarities. The new view is that IPRs are the preferred means of commodifyingthe intangible capital represented by knowledge, and should therefore be seen as acommon currency or “yardstick” for measuring the output of activities devoted toknowledge generation and the foundation for markets in knowledge exchange.The space for public research and knowledge sharing is shrinking and functions thatwere assumed by the public domain are no longer assumed at the same level. In short,recent decades have seen the emergence of a pronounced worldwide trend towards thecommoditisation of publicly funded research outputs, including underlying data andinformation resources.It is also advisable to examine to what extentTT modes are entirely interchangeable fromthe point of view of their effectiveness,especially for LDCs. If this were the case, thenthe strengthening of IP protection would notradically affect the capacity of these countriesto assimilate and absorb new technologies. If,on the other hand, this is not the case then thestrengthening of IP protection could seriouslyaffect the capacity of certain countries toabsorb the technologies they need. Countriesextensively using imitation as TT mode are lesswell placed with regard to direct investmentsor licensing, for example.3.2 The Impact of TRIPS on Technology TransferIt has been argued for a long time by developingcountries (and many economists) that theinternational IP system was biased against them.This argument was essentially a distributional orequity argument: poor countries, at their stage ofdevelopment, ought not to be obliged to pay forknowledge that would ultimately become part ofmankind’s universal heritage. Along these lines,India and Brazil, among other countries, passedlaws restricting the scope of IP protection. 7

30 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesA rather different rationale is based on theidea that since, in most cases, the markets ofdeveloping countries were not those for whichpatent holders had targeted, their innovationsand profits deriving from sales represent pureeconomic rents. So that while it might be debatedas a matter of equity that as beneficiaries theyshould pay, or that being poor they should notbe asked to pay, the important fact remainsthat it would not make much difference from anefficiency standpoint: the world’s supply of suchinnovations would not be much diminished – ifat all – by these countries adopting a free ridingpolicy. The more recent growth in consciousnessregarding global marketing possibilities hastended to vitiate this line of rationalisation.The pros and cons of the impact of TRIPS on thecapacity of developing countries to perform TTsare systematically discussed below.3.2.1 TRIPS barriers and obstacles to LDCs performing technology transfersMany TTs occur through involuntary disseminationvia copying and reverse engineering.Such mechanisms are perfectly acceptablefrom a competition policy point of view whenoperations are complex and demand resourcesand time (see below). The technology holderdoes not participate in this transfer, and in manycases, seeks to restrict it. During the periodof no – or moderate – IPR in the LDCs, copyingwas certainly a major channel for TT. One canargue that it is plausible that today stronger IPRsmay retard such transfer while strengthening“market-based” channels. 8The fact that copying and reverse engineeringrequire more complex and sophisticated legalconditions might be a significant problemfor developing countries; a problem whosemagnitude can be realised just by recalling thatmost rich countries have used this mechanism asa main strategy for technological improvementsand growth. For a while Switzerland, forinstance, excluded all inventions in the chemicalfield from patentability. Such a decision wasexplained by the difficult position of the youngSwiss chemical industry, unable to competewith German firms that had a large-scaleadvantage. The Swiss chemical industry adopteda strategy based on two pillars: innovation andimitation/variation. It focused on products withhigh-added value, especially medicines. Thisstrategy was backed by a policy of imitation.The absence of regulations concerning patentsfor the Swiss chemical industry allowed theBasel firms to concentrate their resources onimitating procedures developed abroad. It wasonly in 1907 that a patent law worthy of thename came into being as a consequence of theindustry’s changing attitude. The development ofthe Swiss chemical firms made them increasinglydependent on innovation through their own R&Dactivities and less on imitation and learning bydoing. Under these conditions, a patent lawbecame important for the industry. Within a fewdecades, the accusation of piracy was forgottenand the Swiss chemical industry became knownfor the quality of its products.What kind of catching-up and knowledgeaccess mechanisms does the internationalgovernance system leave LDCs when copyingis becoming legally more difficult? What is thealternative? And, what are the mechanismsthat could replace copying? These are importantquestions since copying and reverse engineeringproved historically to have two virtues fordevelopment.First, it allows immediate and free access toessential knowledge. Second, it facilitates thebuilding of productive capacities and industriallearning, as already explained in the Swiss case.Another example is the Indian Patent Act in 1970that made pharmaceutical product innovationsunpatentable, allowing innovations to be copiedand marketed in India. Regarding pharmaceuticalprocess patents, the statutory term was shortenedto seven years and automatic licensing wasput in place. As a result, Indian industry learnedvery fast: Indian firms accounted for 70 percent of the bulk drug market. Of the top tenfirms, based on 1996 pharmaceutical sales, sixwere Indian rather than subsidiaries of foreignmultinationals (Lanjouw, 1998). Cassier andCorrea (2005) made a careful empirical studyof the anti-retroviral copying project in Brazil,

ICTSD Programme on IPRs and Sustainable Development31showing evidence of technological learning as aresult of copying.Now that TRIPS makes copying and reverseengineering more difficult, there is still theproblem of devising new mechanisms to addressthese issues: how to preserve free and immediateaccess to essential knowledge for passiveconsumption, and how to ensure free access totechnological knowledge for active contributionto incremental improvements, local innovations,and capacity building?New evidence tends to suggest that if strongerIP protection slows down imitation in the South,this is offset by an increase of multinationalactivities (such as FDI), suggesting an overallenhancement of industrial development indeveloping countries (Branstetter et al., 2007).But while it might be true for a certain categoryof developing countries, it is certainly not truefor the poorest. It is therefore necessary toreview the pros and cons with regard to TRIPSimpact on TT and to qualify these arguments forthe particular case of LDCs.3.2.2 Stronger IPRs are needed to support technology transferThere is a well-known set of arguments basedon the theoretical view that IP protection isan incentive mechanism that rewards andmotivates innovative activities and as such isan indispensable component of a TT-friendlysystem. With some nuances, this argument isperfectly plausible and acceptable for the caseof developing countries that are clearly catchingup(see Arora, 2007, for the most recent paperon this subject). It is likely, however, that sucharguments do not apply to the reality of a numberof developing countries and particularly LDCs.Export and parallel trade issues would alsoarise with respect to goods produced undersuspension of industrial design protection.Because countries protect industrial designwith different legal mechanisms and becausea wide spectrum of goods may be protected byindustrial design, it is difficult to generalisewith respect to the specific type of exhaustionissues that would be presented. If exportswere to be undertaken, this would in anyevent be included within the calculation ofthe level of suspension.3.2.2.1 Attracting foreign direct investments and promoting domesticentrepreneurial activitiesIt is claimed that where the IP protection regimeis weak, FDI is discouraged, and that wheninvestments do occur, they are more likely tobe confined to wholly-owned subsidiaries or thetransfer of older technologies. But, again, thesearguments need qualifications.First, many factors influence FDI decisions andthere is little empirical evidence to supportthe assertion that the insecurity of intellectualproperty exerts powerful adverse effects onthem. The fact that the type of R&D undertakenin developing countries, being less innovative,is not sensitive to patent protection must beconsidered here. There are exceptions, notablythe chemical and pharmaceutical industries,in which the host country’s IP regime doessignificantly effect FDI decisions and levelsin R&D facilities (Mansfield, 1995). But theseindustries are not many. They are those wherethe centrality of patent as a mechanism toreward innovators has been observed for a longtime. 9Second, as already noted, FDI by multinationalsin wholly-owned subsidiaries is considered tobe, at best, a somewhat uncertain channelfor transferring technological knowledge andempowering the indigenous industries withtechnological capabilities (see above chapter 2,2.5.1.2). Therefore, it seems more relevant toconsider how the efficacy of other TT modes,such as joint venture and arms-length licensingagreements, are affected by the nature of theprevailing IP regime.Another possible positive influence of strongerIPRs concerns entrepreneurial initiatives inthe LDC itself. Entrepreneurs, like any othereconomic agents, respond to incentives. They will

32 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesstart businesses and develop new applications ofthe technology if the expected private returnsassociated with these creative activities aresufficiently high, i.e. higher than those promisedby other (less productive) kinds of business.Clearly, stronger patent laws contribute positivelyto private returns on innovation. Forinstance, securing mechanisms to capture theeconomic rents created by innovation will changethe pay-off structure of the economy to makeinvestments in innovation a highly profitableactivity, so that entrepreneurial activities –previously developed in non-productive areas– will be reallocated to productive areas suchas technology development. Alas, while thistheory may apply in rich and middle-advancedcountries, it will not in the case of LDCs.Many other problems have to be solved beforestronger patents will start to have an impacton entrepreneurial incentives in the domesticeconomy. Without good infrastructure, the ruleof law, and well educated people, patents alonewill not do the job.In other words, if an LDC is seeking to attractmore FDI and promoting entrepreneurial activitiesat home, it needs to solve many difficultproblems related to investment climate, efficientgovernance, market size and infrastructurebefore dealing with the patent issue. Thus, therelevant policy question is to ask at what stageof development economic and market-basedincentives (such as patents) become importantto incite productive entrepreneurial activitiesand attract more FDI.Hall (2005) summarises this discussion asfollows:i) Throughout history, a stronger patentsystem has tended to be the result oftechnological development and the creationof firms capable of taking advantage ofthese systems, and not a precondition.ii)Stronger patent rights are likely to increasepayments from developing to developedcountries for technology rights.iii) International trade flows and FDI respondpositively to strengthened patent rightsin middle-income and large developingcountries, but not in the poorest ones(Branstetter et al., 2005).Again, the argument here concerns the leastdeveloped countries. Regarding a few majorcatching-up economies, the current alignmentof national patent and regulatory systemswith those of developed countries is basedon the belief that such alignment is usefulfor attracting foreign capital and, to a lesserextent, that it might stimulate local innovationand entrepreneurships (see chapter 4 below andAbbott, 2003).3.2.2.2 Structuring complex transaction on knowledgeOne argument that has attracted attention,initially voiced by Arora (1995) involvesthe role of IP protection in the creation oftechnological capabilities associated with atransfer of technologies. The argument goesas follows: to be successful, a TT must takeinto account the complementary role of tacitknowledge, which is absolutely necessary to“activate” the technology in an economicallysustainable way. The association of the transferof this tacit knowledge (through technologysupport and training services) with the transferof the technology itself requires the design ofspecific contracts. While information asymmetriesand monitoring difficulties make itvirtually impossible to draft efficient contractsspecifying the transfer of tacit knowledge, it ispossible to design contracts for the successfulimplementation of technologies by bundlingthe provision of assistance together with thelicensing of the use of codified information suchas patent and copyrights (Arora, 1995).In such cases, patents play a significant rolein helping to structure a complex transactionincluding unpatented knowledge.If the protection for such property is weak inthe borrowing country however, the originatingfirm is unlikely to enter into such contracts.The implication of this is clear: the wouldbeborrowers have an interest in a stronger IP

ICTSD Programme on IPRs and Sustainable Development33protection regime. This argument tells us thatthe South will gain from extending IP protectioninto its own markets by focusing on the conditionsfor the successful transfer of codified and tacitknowledge regarding innovations.The argument developed by Arora (1995)is a sophisticated one. It is based on thetransactional difficulties created by the factthat codified information and tacit knowledgeare complementary and must be transferredtogether. However, this argument overlooks theproblem of legal and technical capacities ofthe recipient country that needs, in particular,highly skilled people able to deal with complexcontract negotiations.3.2.2.3 Structuring and providing free technical information, while not impedingaccess and transfer “to certain countries”Finally, “copying” as a TT mechanism is notnecessarily impeded or prohibited in a strongIP environment but can, on the contrary, beenhanced in certain special circumstances.First, it is useful to recall that in exchange forpatent rights, the inventor must publicly divulgetechnical details of the new technologicalknowledge. Technical description is an essentialact. It is intended to provide sufficient “instructions”for a specialist in that particular field tobe able to reproduce the invention. In this sense,the patent system generates a huge repositoryof technical information in any technologicalarea that can be freely used by anyone lookingfor information about a given technology. It maythen happen that the patents providing the usefulinformation are not valid in certain countries sothat not only can the information be freely used,but also the knowledge can be reproduced andused commercially in these countries.This is exactly what happened in Ethiopia wherecertain technologies were needed to develophearing aids powered by solar energy. A searchwas carried out for patent documents relatedto hearing aids, which enabled the retrieval ofa number of interesting documents. Detailedanalysis revealed that all these documents relatedto patents that did not protect the disclosedinventions in any African countries. Therefore,the published patent documents containedtechnical information that could be freely usedin Africa for solving technical problems, and thedisclosed devices could be freely manufacturedand distributed in Africa (Corsi, 2007).This case exemplifies an ideal situationin which the IP system provides a strongmechanism to structure information andincreases the productivity of informationsearch, while not impeding access to theknowledge once retrieved. It is important toknow whether this case represents a generalpattern – i.e. there are many granted patentsfor which the applicants have no interest in theAfrican markets and consequently will neverapply for corresponding protection – or whetherit is quite exceptional. 10There is therefore an obvious need forinternational collaboration to establish IPinformation systems and clearing houses. Such asystem could greatly reduce the cost of patentsearches by developing countries. There arealready a few examples of internet-based patentdatabases that enable a user to easily access andanalyse published patents and patent applicationsfrom many countries (Byerlee and Fischer, 2002).This mechanism, sometimes called “unilateralaccess” has a number of limitations however,even when it is strictly legal (ibid.).3.3 Markets For Technology: Are They Really Efficient?In an ideal world where all entities are at thesame level of technological advancement, it canbe hypothesised that the technology market,based on a strong and respected IPRs system,offers a very effective mode of TT. This is Baumol’stheory, supported by empirical evidence fromArora et al. (2001): IP creates transferable rightsand can also help to structure complex markettransactions involving technologies.Baumol (2002) argues that firms are remarkablyquick to transfer their technologies, which can

34 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesbe explained by capitalist incentives as wellas the efficiency of markets for technology.Under (theoretically) simple conditions, anequilibrium price will emerge where both firms– buyer and seller – will be better off afterhaving carried out the transfer; markets work,knowledge transfer occurs. The very simplecondition is that the buyer is a more efficientuser of the technology than the seller. Thebuyer’s superior efficiency means that it canearn more from its use of the technology thanthe owner. Thus the buyer can still make ahigher profit than the owner would be able toearn by using the technology itself, even if thebuyer pays a somewhat costlier licence fee.Since TT through market transactions providesa means of obtaining a sufficient reward frominnovation, if the price is right, it will pay thefirm to permit others to use its technology.Such conditions are not likely to occur inthe case of TTs involving very heterogeneoussystems. No equilibrium price will emerge andit is not easy for a profit-seeking organisationto engage in such an operation if the profitpotential is very low, or even zero.And even in a homogeneous world in which allfirms are equally technologically advanced,markets for technology as a vehicle for TT donot work that well. In a convincing paper basedon recently collected evidence, Cockburn(2007) argues that it is difficult to think of anyother market that has as many failures. Thereare many obstacles to licensing, includingthat:• Knowledge-based transactions are costlyand complex;• There are many “missing” markets forspecific technologies (difficulty in findingappropriate licensees);• Reaching agreements regarding price andother conditions is difficult; and• Operating managers are reluctant to selltechnologies.As a result, a large fraction of the total IPinventory appears to be un-licensable under anycircumstances; an average of over one-third ofa firm’s total IP inventory is considered unlikelyto be licensed, despite the firm’s willingnessto license its technologies. These obstaclesimpose two major types of disadvantages:• “Unrealised deals,” which mean anunderutilisation of IP on the markets;and• Wrong prices for technologies, leading topoor outcomes.The markets for technology are thereforeinefficient. Such inefficiency is even greaterwhen the market is placed upstream of theinnovation value chain (i.e. when transactionsinvolve research tools, scientific information,etc.). Baumol’s ideal world (in which markets fortechnology work well and knowledge disseminateswidely and quickly through this mechanism) doesyet not exist – and perhaps never will.3.4 A Fundamental AsymmetryStronger IP protection systems create afundamental asymmetry when a proportion ofagents are “IP-users-only.” In a rich countryor region, this situation is not likely tohappen (or only marginally so) and a strongerIPR protection system will not generateasymmetry. Each entity is a potential IPproducer and IP consumer so that both willreceive some further benefits and bearadditional costs. The only exception concernsthe final consumers who, by definition,are not IP producers or holders. 11 This nonasymmetricworld is not the case, however,when the global world is considered: a largeproportion of potential patent users have nocapacity to produce IP so that the costs andbenefits of the strengthening of the IP systemwill be unequally distributed.

ICTSD Programme on IPRs and Sustainable Development353.4.1 TRIPS in an asymmetric worldIn a multi-country world, the cost to one countryof introducing patent protection depends notonly on the size of the deadweight loss butalso on who is doing the inventing. If the newlyavailable patent rights for drugs in an LDC areentirely assigned to inventors elsewhere, theloss of consumer surplus is a net cost to the LDC.All the profits accrue to foreign nationals in theform of royalties. Thus, static costs to a countryintroducing patent protection in a multi-countryworld may be higher than the standard onecountrymodel would suggest (Lanjouw, 1998).The table below summarises the static anddynamic costs and benefits resulting from astronger IPR system.Table 3. Costs and Benefits of a Stronger Patent SystemCost of strongerIPRs for country X**Benefits of strongerIPRs for country X**StaticKnowledge is bought at monopolyprice (above marginal cost)leading to a loss of welfare(which can be dramatic in LDCs)*Knowledge is sold at monopolyprices and companies in X cancapture the economic rentsDynamicBarriers to access modifiabletechnologies (as base for learning)are increasingThe pay-off structure (incentives)of the economy changes so thatentrepreneurial activities becomemore profitable.Foreign capital (R&D) is attracted* The cost of building a legal system is not considered here, although this can be very high in acountry where such a system does not yet exist (or is performing badly)** This table applies indifferently to developed and developing countriesAs far as static cost and benefit are concerned,an LDC that has “nothing to sell” will not benefitfrom the strengthening of the IP system, whileit will bear the high costs of making certainessential knowledge inaccessible to large fractionsof the population.As far as dynamic cost and benefit are concerned,the new system impedes learning by increasingbarriers to access technological componentsand modules for those who cannot afford to buytechnological licences. The benefits are clearlysignificant for catching-up countries that alreadyhave a class of entrepreneurs and innovativecompanies likely to respond positively (in termsof investments) to a stronger IP system. Theyare negligible for LDCs, however. Indeed, it isan illusion to think that the mere manipulationof incentives (such as creating an IP system)will suffice to motivate the development ofentrepreneurship and innovative activities in thepoorest countries.Using a general equilibrium model frameworkwith two regions (the North and South), Angeles(2005) shows that the effect of strengtheningIP protection in the South has a positive effecton the increase of world innovation, but thisdoes not come without costs. Higher prices forconsumers in the South are the negative side ofthis policy. The South may suffer a net welfareloss if its productivity is very low in relationto the North, which is clearly the case of theLDCs.Table 4 below presents absolute numbers ofpatent applications by residents and nonresidentsin LDCs and shows that theseasymmetric relations are likely to be amplifiedin the near future.

36 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesTable 4. Patent Applications by Residents and Non-residents in LDCsPatent applications by non- residentsPatent applications by residents1995 172 731996 195 116 271997 261 141 61998 449 616 701999 570 676 182000 978 409 182001 1 352 635 232002 1 753 699 6Source: WIPO, 2006Resident applications are those for which the first name applicant or assignee is a resident of the state or regionconcerned; Non-resident applications are those from applicants outside the relevant state or region.The table above not only illustrates the strongasymmetric distribution of costs and benefitsas produced in a global patent system, butit also shows very worrying trends: while thegrowth of patent filings by non-residents hasbeen continuous and significant since 1995,patent filings by residents show no suchtrend. No growth (even at a very low rate) isperceptible, meaning that the impact of TRIPSon incentives and the ability of domestic andlocal entrepreneurs to patent, are negligible.As a result, the asymmetric distribution ofcosts and benefits is likely to intensify in thenear future.Figure 2. Patent Applications in the LDCs, Residents vs. Non-residents,1995-2002Source WIPO: 2006

ICTSD Programme on IPRs and Sustainable Development37TRIPS works like an electric conductor. When itis set up, any failures occurring somewhere (inthe rich countries) will place the whole systemat risk. In other words, TRIPS means that LDCsare not immune to – and in fact are extremelyexposed to – the negative excesses of the patentsystem in rich countries. In others words,such excesses, which are the manifestationof a problem of pure bad governance of thesystem in the rich countries, impose costs andblockages not only in the countries responsiblefor these excesses, but also in any country thatis signatory to the Agreement.Secondly, in an era of global integration,countries are not immune to repercussionsfrom the (IPR strengthening) policies of othercountries (Hall, 2001). The strengthening ofIPRs in rapidly catching-up economies createsa negative externality for LDCs (reducing theincentive for innovative activity in LDCs byboth attracting R&D to move within its bordersand raising the costs of follow-on inventionelsewhere).So TRIPS is not a magic solution for promotingTTs in LDCs. For LDCs, as far as IPRs and patentsare concerned, the correct policy orientationshould be to place more emphasis on protectingthese countries from the potential collateraldamage likely to be caused by IP protectionstrengthening, rather than using IP protectionas a positive market incentive to supportentrepreneurship and attract FDI. Indeed, notonly does TRIPS not solve development issues,it also creates new problems for LDCs.3.4.2 One size does not fit all and the need for flexibility in patent and copyright systemsThe fact that the costs and benefits of astronger IPR system are unequally distributedin a multi-country system and that the lessadvanced countries will bear high static anddynamic costs while not enjoying any benefit(at least in the mid-term) creates a strong casefor adapting the system to particular socioeconomiccontexts. A one-size-fits-all principlewould be suboptimal where countries’heterogeneities are concerned. As discussedin the 2002 Report of the UK Commission onIPRs, one size does not fit all. The poorestnations clearly need some flexibility as well asad hoc mechanisms to solve their access andknowledge production challenges.Strong arguments against this one-size-doesnot-fit-allposition have been developed andare worth considering. The most relevantarguments against this principle, however, donot deal with differences between countriesbut differences between technologies. Indeed,Jaffe (2005) argues that while it is relativelyeasy for economic theory to demonstrate thatoptimal patent design should differ significantlyacross technological areas and industries andthat optimisation is achievable in a purelytheoretical world, a strong case can be madefor not opening this particular Pandora’s Box.Efforts toward some kind of fine-tuning accordingto technological heterogeneities will ultimatelyfail and are likely to weaken the patent system.Indeed, the theory identifies certain features ofa technology that make strong or long patentsless desirable (for example, the cumulativenature of some technological knowledge, orits importance and generality). However, it isextremely difficult to identify these featuresin particular technologies (this is a matter ofdegree and not clear categories). Even if inprinciple the applicability of patents basedon such analysis of intrinsic technologicalprinciples could be restricted, it is likely that– in practice – such efforts would fail. Draftersof patent applications will always prove moreingenious than writers of patent rules. Thus,prescribing patent protection for certain classesof technology will simply force applications tobe written in such a way that they appear inother classes. These strong arguments againstthe principle of one-size-does-not-fit-all do notapply, however, when heterogeneity concernsnot technologies but “socio-economic” contexts.The argument is well expressed by David in anunpublished paper, in which he states:Historical studies reveal that althoughpatents, copyrights, and legal protection

38 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesof trade secrets are recognisableinstitutions familiar in western societiesfor centuries, policies bearing uponthe protection accorded to intellectualproperty and the juridical-institutionalarrangements used to implement themhave been mutable things, adaptingover time and across societies to theperceived needs and advantages ofinterested parties. The adaptations ofthe IPR systems have occurred withinthe historical context of other, related,institutional arrangements affecting thecosts and benefits of maintaining specificIPRs. An implication of this observation isthat externally-dictated efforts to achievenational compliance with a uniforminternational regime of IPR protectionare almost bound to occasion conflictand controversy. Even where it is possibleto argue that a new IPR regime couldbe constructed that would be Paretoimprovingfor the countries involved, theneed to align domestic and internationallaws introduces additional constraintsthat tend to render such solutionsimpractical. As a result discussions onthe ‘correct’ international system toprotect intellectual property are morelikely than not to degenerate intorhetorical efforts to impose institutionalarrangements that may well be adaptedto the national purposes and thedomestic legal contexts of one country(or several similar countries) uponsocieties that are quite different inthose respects.Two types of flexibility should be particularlyscrutinised: the internal flexibility offeredto countries by TRIPS to limit (or extend)exclusion rights, and an external flexibility,which mainly consists of using the power oflegal institutions to reconstruct researchand information commons and support opensourceinitiatives as a way of mitigating theadverse effects of a highly protectionistIPR environment and promoting low cost TTmodels in LDCs (see section 3.5, below).Fully exploiting the scope of TRIPS flexibilityin one sense or another (limitations to orextension of exclusion rights) is a crucialissue. 12 But exploiting the autonomy left opento LDCs by TRIPS not only raises questionsabout the legal rights to do so, but moreimportantly about the technical capabilitiesto use the opportunities offered by the systemin the best interests of the knowledge ecologyof LDCs.3.4.3 Learning to use flexibility: a role for national patent officesTRIPS clearly offers some degree of autonomyto LDCs. However, having these provisionsavailable is a different matter than being ableto apply them effectively.Firstly, as already mentioned, there arecases where LDCs formally renounce theiruse as a condition for obtaining further tradeadvantages through bilateral agreements.Secondly, these mechanisms are difficultto implement and sophisticated knowledgeand skills concerning law and internationalagreements are needed. This is why a TRIPSprovision involves the obligation for thedeveloped countries to provide bilateraltechnical assistance to LDCs that requestit (see Article 67). Recent experienceshave clearly demonstrated however thatdeveloped countries provide valuable helpin establishing the appropriate measuresto strengthen IPR protection in the countryconsidered, but do not extend that helpto the use of mechanisms like compulsorylicensing (Kostecki, 2006). For example, theUS IPR training coordination group, whichexemplifies the application of Article 67 bythe US, is dominated by private firms andonly focuses on strengthening IPR systemsin the less advanced countries. “NGOs andacademics with the knowledge and expertiseto redress the balance by highlighting thescope for TRIPS flexibilities alongside issuesof protection and enforcement of intellectualproperty rights are excluded from the US

ICTSD Programme on IPRs and Sustainable Development39IPR training coordination group altogether”(Matthew, 2005).There is certainly a role for national patentoffices to create political awareness thatthese mechanisms are critically important forperforming TTs in LDCs and, therefore, shouldnot be regarded as bargaining chips in tradeagreements. Another role for national patentoffices is to build and maintain the legalskills essential for using these mechanismseffectively.3.5 Beyond IPRs: the New Knowledge Economy ParadigmAny analysis of the relationship between IPpolicies and TTs in the twenty-first centuryshould not neglect the fact that the innovationprocess itself has changed: there is a degreeof innovation within the process of innovation(Ghosh and Soete, 2006).The old model of innovation was based on arelatively clear dichotomy between those withthe capacities to discover new technologicalprinciples and those without such capacities.In such a model, TTs were dedicated to helpingthe latter upgrade their technologies, whilethe former did not expect any retro-spilloversfrom TTs and the incentives to promote TTswere designed accordingly. The new model isone in which rapid collaborative processes forincremental innovations based on free accesscreate multiple sources of new knowledge,leading to more efficient and faster technicalinnovation, with the entrepreneurial risks ofinnovation widely spread.The idea is to use open-source licensing tokeep discoveries freely available to researchersand eventually manufacturing. What seemedunfeasible a decade ago appears possible todaythanks to the greater size and variety of chemical,biological and medical databases, new softwareand more powerful computers (Maurer et al.,2004). All these features have strongly enhancedthe productivity of collaborative efforts torapidly develop knowledge and products incertain fields.Today, technological changes make it possible toextend open-source mechanisms far beyond thefield of software where this model has provedto be remarkably successful in generating a highrate of innovation and reliable products at verylow costs. This is the case of drug research,for example. The rise of in silico biology hasdramatically lowered the cost of conductinguseful drug research. It blurs traditionaldistinctions between drug discovery, academicdatabase production, and open-source. Commonfeatures include (Maurer et al., 2004):• Community-wide collaborations that producescale and network effects (the scale ofthe problems requires many contributors);• Loose, non-hierarchical groups workingtogether to perform complex tasks andcreate specific products;• On-line collaboration, which dramaticallyincreases the productivity of collaborativeresearch; and• Unpaid volunteers who contribute to suchprojects for a host of reasons, includingidealism, learning new skills, gainingreputations, and impressing potentialemployers.Open-source discovery beyond software wouldoperate like open-source software projects.There is no reason not to expect similarlypositive results in terms of cost effectivenessand innovation performance.The main advantage of open-source is thatit is likely to reduce the total life cycle costrequired to get the job done. Cost effectivenessis based on the fact that such projects do notoffer financial incentives but create properincentives for voluntary contributions. A secondreason for cost containment concerns theabsence of patents and above marginal costpricing.Innovative performance is a second advantage.It is related to the expected productivity ofsuch projects, which in turn is related to rich

40 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesspillovers that are created de facto in an opensourceenvironment.Last but not least, the most important enablingfeature of this model is access. TTs are interactiveand reciprocal. In this paradigm, providingaccess to technologies and knowledge shouldnot be seen as charity or aid, but as a necessarystep to enlarge the resource base of potentialinnovators, i.e. a way to increase efficiency atsystem level (Ghosh and Soete, 2006).As argued by Maurer et al. (2004), no knownscientific or economic barrier bars the way foropen-source discovery and collaborative modelsof TT.

ICTSD Programme on IPRs and Sustainable Development414. CURRENT TRENDS IN TECHNOLOGY TRANSFERThis section analyses recent data about technologydiffusion and TT toward developingcountries to determine whether the variousTT mechanisms (shown in chapter 2) are usedeffectively by countries; the extent to whichthose mechanisms play a role as an engine forgrowth; and if there is any need to reconsiderthe general dominant policy that is based onthe premise of the centrality of FDI and trade asTT mechanisms. The recent World Bank reportoffers a complete set of data that will be usedbelow, with a focus on technology diffusion.4.1 The World Bank FrameworkThere are two fundamental determinants oftechnology diffusion in the LDCs that the WorldBank places at the centre of its framework ofanalysis. The first determinant involves the threemain channels by which developing countries areexposed to external technologies. This includestrade, FDI (and licensing that can be used as asubstitute for FDI), and highly-skilled diaspora.The other important determinant of technologydiffusion is the absorptive capacity ortechnological adaptive capacity of the country.Proper policy actions that create improvementsin governance, the business climate, humancapital (increase of basic technology literacy),technological capacities of firms, and access tocredit on capital markets (innovation policy)increase this capacity.These two determinants are interrelated andcreate externalities to each other, therebyforming a dynamic system with positivefeedbacks; they generate multiple equilibriumsthrough virtuous (or vicious) circles (see e.g.Stiglitz, 1991). Such systems are well knownand well studied in the literature of economicdevelopment.For example, the fact that a country becomesmore exposed to foreign technologies (throughthe increase of FDI) may actually increasereturns to additional improvements of absorptivecapacity. And the fact that the absorptivecapacity is improving in turn increases theprobability of spillovers being diffused withinthe domestic economy. As a consequence, theoverall efficiency of the economy is increased,which positively influences FDI in the country.As for any positive feedback system, there arevirtuous circles (taking the structural form ofwhat has been just described) but also viciouscircles (Figure 3).

42 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesFigure 3. Positive Feedback Systems, Virtuous and Vicious Processes4.2 Virtuous CirclesThe basic message of the World Bank reportis that many developing countries – notablymiddle-income countries – have entered avirtuous circle: the basic components of thefeedback system described above are mutuallyimproving each other. The increased exposureto foreign technology (through FDI and trade)is co-evolving with the growth in disseminationand spillovers of these technologies within thedomestic economy. As such, by many measures,these countries have made outstanding progressin technology diffusion. The main indicators ofsuch trends involve:• The rising share of high tech and capitalgoods imports;• The expansion of exports of technologicalgoods;• The increase of FDI as a percentage ofGDP, as well as the increase of FDI as apercentage of fixed capital formation; and• All these trends leading to the overallincrease in exposure to external technologiesindex.Therefore, it appears that for these countries,trade and FDI are the main channels foraccessing foreign technologies; TTs as a jointproduct work well, while absorptive capacitiesare sufficiently positive to allow spilloversfrom the technology to be transferred tothe rest of the economy. Good policies andgovernance remain of course central. Thisinvolves keeping FDI and trade at a high leveland continuously improving the absorptivecapacity.

ICTSD Programme on IPRs and Sustainable Development43This trend is consistent with evidence onthe positive relation between IPR reformand the stimulation of TTs in middle-incomecountries. The recent study by Park andLippoldt (2008) finds that stronger patentsystems tend to be positively associated withinward FDI and trade, and the strength of thepatent system is positively and significantlyassociated with TT (i.e. the inflows of hightech products, like pharmaceutical goods,chemicals, aerospace, computer services,information, and office and telecomequipment). It is also consistent with recentempirical evidence produced by Branstetteret al. (2007), which suggests that due toIPR reform, increased multinational activityin developing countries is sufficiently highto offset potential declines in imitativeactivity, suggesting an overall enhancementof Southern industrial development.4.3 Vicious (or Non Virtuous) Circles: the Case of LDCsUnfortunately, in the case of low-incomecountries, things do not happen as describedabove. Empirical evidence suggests that thevarious channels by which countries are exposedto foreign technologies are far less powerful.FDI remains at a very low level (less than 1 percent of GDP) and the share of FDI in low-incomecountries capital formation is low as well. Inaddition, the ratio high tech product import/GDP is also low – LDCs remain marginal playersin the world market for high tech goods. Finally,having a licensing-based strategy to acquiretechnology to complement or substitute FDIis not proven to be efficient due to the lackof capabilities both in technological and legalterms and the inefficiency of technologicalmarkets, particularly in case of transactionsbetween heterogeneous players.Least developed countries have not succeededin improving their absorptive capacities, whichlimits the potential for the improvement offoreign technologies to strengthen the domesticeconomy. How these countries make use offoreign technologies is qualified by the WorldBank as a “passive approach and limited effortto leverage the technology imported by foreignfirms operating on their soil” (World Bank,2008).So not only has the exposure to foreign technologynot increased that much through FDI and trade,but the extent to which these economies havebenefited from this exposure is limited by weakcapabilities. As a result, the gap between middleincomeand low income countries is widening.This is seen, for instance, in terms of the shareof capital goods in GDP.What works in middle-income countries – TT asa joint product of main economic operationsinvolving FDI and trade – does not work well inLDCs. This argument leads to one of the mainmessages of this paper:In the case of LDCs, the number, scale, anddomains of TTs cannot be allowed to dependon general economic operations such as FDI orinfrastructure construction; neither can theytake the form of market transactions alone(licences). In all these cases, the particularcircumstances and conditions prevailing inLDCs imply a suboptimal level of TT in relationto the needs of these countries.Therefore, augmenting FDI to emulate thesuccessful model of some middle-incomecountries cannot be the only policy response.First, it is a long-term issue.Second, as already mentioned, TT as a jointproduct of FDI raises a “balancing incentivesproblem;” a problem that is likely to be properlysolved in the case of middle-income countriesbut not in the case of LDCs. Let me discuss anexample to illustrate the last point. Chapter 2discussed both the advantages and drawbacks ofFDI as a TT channel. The drawbacks involve thefact that decisions concerning most aspects ofTT remain in the hands of the foreign investorsand as a consequence there are doubts aboutforeign investors devoting sufficient resourcesand time to the learning process. For example,foreign investors are likely to have littleincentives to initiate a shift in responsibilitiesfor technological adaptations to local suppliersor staff, preferring instead to delay the

44 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesreplacement of expatriates. This prevents thelearning process from taking place. Now thesedrawbacks are clearly mitigated in the case ofcountries with stronger absorptive capacitiesand thus these countries rightly rely on FDIto operate TTs. But as absorptive capacitiesare low, the FDI’s drawbacks are not reduced.In the example above, the cost of increasingcapabilities so that at some point local staffcan replace expatriates can be so high thatthe foreign investor will not consider thisreplacement as an economically viable option.In the case of LDCs, TTs cannot only be a byproductor joint product of another kind ofeconomic operations. Incentives are likely tobe strongly imbalanced between making theinvestment operational and effective, andsucceeding in TT and learning. The other policyresponse is to provide additional incentivesto undertake projects in which TT is the mainoperation.There is therefore an economic rationalityfor specific projects in which the TT is theprimary operation (an economic project initself, not linked with another economicoperation), but entails a very low expectedprivate profitability for the technologyowningfirm. Such a prospect would involveacknowledging the existence of TT operationswith far smaller commercial returns orno commercial return at all and findingoperational mechanisms to incentivise thesefirms to sink costs in these operations. Sucha strategy obviously requires the provision ofadditional incentives.

ICTSD Programme on IPRs and Sustainable Development455. THE ECONOMICS OF PUBLIC-PRIVATE PARTNERSHIPS FOR TECHNOLOGYTRANSFERS BETWEEN DEVELOPED COUNTRIES AND LDCSThis study has shown that there is an obviouseconomic rationale for specific projectsin which TT is the primary operation (aneconomic project in itself not linked withanother economic operation). TT as a mainoperation provides opportunities for shiftingthe locus of decision to “local agents.”This would affect the modes and quality oflearning, technology transfer and the choiceof the area to be concentrated on. In contrastwith TT as a joint product, TT as a mainoperation offers an opportunity to prioritisesome domains.But TT as a main operation is costly whenthe recipient has weak capabilities. Thisis due to the cost of building capabilitiesand supporting absorption, adaptation,assimilation of subsequent improvements,and the generalisation of the technologyin a system exhibiting low absorptivecapacities.Given these arguments, private firms that holdthe technology cannot realistically be expectedto find sufficient market incentives to investin this activity at the socially desirable levelany more than governments of rich countriescan be expected to force companies to do so.Without additional incentives, TTs carried outwith LDCs would mainly be the result of moregeneral economic operations, such as FDI orthe construction of infrastructures. In all thesecases, TT is at best a joint product (or evenby-product), which is therefore dependent onthe main operation.Incentivising foreign firms to enter suchtransactions is an opportunity for developedcountry governments to properly fulfil theirobligations as contained in Article 66.2 of TRIPS.In somewhat vague terms, Article 66.2 callsfor the provision of additional incentives fordeveloped country firms and other organisationsto undertake TTs to LDCs.This section provides some recommendationson the best approach for developedcountries to fulfil their obligations in thisdomain. It also discusses the centralityof “specialised agents” (PPPs) to ensureboth the effectiveness of governments’interventions and the efficiency of the TToperation.5.1 The Economic Rationale of TRIPS Article 66.2: the Potential andLimitations of IncentivesArticle 66.2 of the TRIPS Agreement asksdeveloped country members to provide incentivesto their enterprises and institutions forpromoting and encouraging transfer to LDCs.This is indeed an economically valid prescriptionas argued above. However, many obstacles andlimitations impede the ability of governmentsto influence firms’ decisions regarding theundertaking of TTs with LDCs. In its report, theEuropean Community (2007) acknowledges thedifficulty of putting Article 66.2 into practice.Some issues that it notes are:• The private and commercial sector is clearlythe main source of technologies, and in thiscontext, TT is often one component of amore complex project, rather than a standaloneactivity;• In their efforts to encourage and promoteTTs, governments are limited by two factors:they do not own the vast majority of thetechnologies and they cannot force theprivate sector to transfer its technologies;• Government incentives to undertake TTsto LDCs are only one of many factorsrelevant to companies’ decisions onwhere to direct FDI and TTs. Other factorslike macroeconomic stability, a safelegal framework, and a well-functioninggovernment administration are alsoimportant drivers;• Incentives can only take the form ofencouragement, promotion, and facilitationof the most fruitful projects.

46 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesAgainst this background, the report considersthat the following incentives are relevant andmust be activated by governments:• Promotion of projects such as FDI,subcontracting, licensing, franchising;• Improvement of access to techniques andindustrial processes;• Support of joint projects;• Provision of training in management oftechnology and production methods;• Improvement of absorptive capacities; and• Encouragement of trade in technologicalgoods.5.2 Historical Precedents: Both Incentives and Specialised Agents in PPPsMatterThere are historical precedents showing thatthe manipulation of incentives in an area oflow private profitability can profoundly changethe strategic behaviours of economic agents.The reader must understand that the followingexample does not directly concern incentivesto undertake TTs. Rather, it shows that it ispossible to create sufficient incentives andcontain costs in an effective way so that privatecompanies commit resources to projects initiallycharacterised by low profitability expectations.While the usual approach had been to estimatethe minimum level of commercial profits and thensupplement low developing country purchasingpower with large market pull incentives(advanced purchase commitments), the newerapproach is to recognise the existence of R&Doperations with far smaller commercial returns(for small business) or no commercial return atall (but no loss) for large companies and to findoperational mechanisms to incentivise thesefirms to sink costs in these operations. But sucha strategy requires strong cost containmentmechanisms. As argued by Maurer et al.(2004), the failure of Western governments andpharmaceutical companies to cure developingcountry diseases is almost entirely about cost.It is therefore reasonable to assume that themechanisms presented below can break theimpasse and help to support R&D investmentsaddressing neglected needs.The evidence is striking. Over the last four years,the number of neglected-disease drug projectshas increased significantly. As Moran (2005)put it, there were 63 neglected-disease drugprojects under way at the end of 2004, includingtwo new drugs at the registration stage andeighteen new products in clinical trials. Assumingstandard attrition rates, these projects would beexpected to deliver eight to nine new neglecteddisease drugs within the next five years, even ifno further projects were commenced after thistime. But new projects have been launched sincethe end of 2004, amplifying a trend that leads todeep-seated structural change in the economicsand organisation of neglected disease R&D.Commitment to such R&D by for-profit companiesthat are constrained by shareholders’ valuesshould not exist (according to economic theory),but it does. As Galileo is said to have murmuredafter officially recanting his statement thatthe Earth moves around the sun: “and yet itmoves!”.Moran (2005) argues that such an increase in“non-profitable R&D” addressing the neglectedneeds of LDCs indicates deep-seated structuralchanges. Three main characteristics areobservable:- Multinational companies work on a noncommercialbasis – that is, they are notmotivated by commercial returns for this kind ofproject and agree to provide the final productsto poor countries at not-for-profit prices. Longtermbusiness considerations include reputationeffects, corporate social responsibility, ethicalconcerns, and strategic considerations, suchas the growth potential of LDC markets in thelong run. One important change that helpedthese companies undertake these strategies wasmoving upstream to the less expensive and more

ICTSD Programme on IPRs and Sustainable Development47innovative drug discovery stages, allowing themto maintain costs and resource inputs at levelsmore acceptable to shareholders.- Smaller scale commercial firms do not abandonthe idea of profit-making from their projects butare motivated by far smaller commercial returnsthan large companies. They see LDC markets assufficiently attractive to warrant some positivereturns, although not at a rate that would berequired by larger firms or external investors.- Public-private partnerships. None of thecommitments from the large MNCs and smallerbusinesses would be possible without thecontribution of PPPs. In the case of largecompanies, PPPs facilitate further developmentby subsidising clinical trial costs and etc. Theintervention of PPPs is critical to sustain this“no profit no loss model,” which allows largecompanies to participate in neglected diseaseresearch while still protecting shareholder valueand manufacturing. It also enables them todistribute final products to developing countrypatients at no mark-up. In the case of smallerbusinesses, these companies need substantialPPP support, including full cost coverage andsignificant skills input. Thus PPPs seem to be thecritical institutional innovation for new modelsof research, something that seemed impossible(because it was unprofitable) some years ago.But the operational role of PPPs goes far beyondtargeting funding at ad hoc R&D projects. PPPsplay a role in coordinating and assemblingdispersed resources that gain value through theirassociation.The very economics of R&D in certain fields tellsus that R&D investments constantly generateexcess capacity. This is due to indivisibilitiesof R&D investments (you cannot invest belowa certain threshold and this applies to bothphysical and human resources); and patentraces phenomena that may lead either tooverinvestment in R&D or the proliferation ofIPRs – not all appreciated by firms in the contextof the patent race considered.The context of excess capacity is a structuralfeature of the economic organisation of R&Dbased on certain characteristics, such as minimumefficient scale, lumpiness, and indivisibility, aswell as incentive structures, like patent races.While some of these resources can be sold on asecond-hand market, the role of PPPs is to createa feasibility space for social sharing rather thanrequiring a model of second best pricing, andreallocate these resources efficiently (i.e. tosocially useful projects). While indivisibilities,overinvestment, and the generation of profusionof IP (through patent races) generate challengesfor efficient pricing, they also create conditionsin which new institutional machinery is likely toprovide a more efficient framework to supplyand exchange those goods than would the pricesystem.5.3 Public-Private Partnerships for Technology TransferAs pointed earlier in the paper, the complexityand difficulties of TT operations as supported andincentivised by governments of rich countriesmake it necessary to use “specialised agents”that have accumulated experiences in TT’soperations. 14 The essence of a public-privatepartnership – to put this institutional mechanismin contrast with the more usual “public funding ofprivate initiative” kind of arrangement – residesin the involvement of a third party, which isspecialised in linking public donors, privatefirms, and local entrepreneurial activities toensure the effectiveness of the operation.The third party is a not-for-profit organisationwhose objective is to efficiently managepublic or philanthropic funding. As a generalstatement, the PPPs will establish the link andmake it operational and effective between thepublic donor, a private company that holds thetechnology, and local demand.

48 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies5.3.1 The essence of a PPP: the functional role of a third partyAs a specialised entity, the PPP will compensatefor the critical deficit of institutional mechanismsboth in the LDC and the developedcountry, which are needed to address problemsarising from the management of a TT as a mainoperation (see Figure 4). The PPP will:• Make sure that the whole process is“demand pull” (i.e. an area or field of TTthat clearly matches a local entrepreneurialneed) and create the proper incentivestructure for the local entrepreneurs toengage resources in the project;• Target the appropriate partner (technologyholder) and generate the incentive toinvolve it into the transaction;• Create the proper conditions to containcosts – a crucial condition for increasingthe anticipated private profitability ofTTs;• Generate the proper organisational forms(internal and external, i.e. platforms) toensure the TT will successfully meet thedifferent phases (absorption, adaptation,assimilation of subsequent improvements,generalisation), since there is no “superior”economic operation that will providethe organisational structure, such aswhen the TT is a joint product of a directinvestment;• Supervise the whole process so that thevarious phases of the TT are successfullymanaged;• Create mutual contractual obligations sothat no party can leave the project beforeits completion; and• Last but not least, manage the IPR side.As already argued (chapter 2, 2.5) the locusof decision-making with regard to undertakingthe TT in a certain field and under specificconditions is also an important variable whenTT is the main operation. In circumstanceswhere TT is the main operation, the locus ofdecision-making should not be kept in foreignassistance bodies but transferred to localgovernment initiatives and entrepreneurs. PPPswill be central in managing such a shift.Figure 4. Functions and Role of a PPP as TT Operator

ICTSD Programme on IPRs and Sustainable Development495.3.2 Supply side: “right” technology holders?Are there any “model of firms” in developedcountries that might be more capable than othersto enter such operations of TT toward LDCs, giventhe constraints mentioned above (disarticulatedsystem, weak capabilities, low affordability ofthe potential clients”)? If the answer is yes, onerole of government policy is to identify and enrolthose firms into TT’s projects.Arora et al. (2001) developed interesting casestudies of specialised technology suppliers inthe chemical industry. They studied how thedevelopment of specialised upstream technologysuppliers in developed countries improvestechnology access and lowers investment costsfor downstream firms in developing countries.Testing this idea they showed that investmentsin chemical plants are greater in developingcountries where the number of technologysuppliers that operate in developed countriesis greater. According to the authors, whatmatters is therefore the vertical organisationof the industry in the developed world. Thatis, investment is taking place earlier and morerapidly than if developing countries had to relysolely upon chemical producers in the developedworld to transfer the technology. The mechanismis quite simple: specialised suppliers developtechnological capabilities that are then sold todownstream firms. Because the expertise andthe technologies developed are process (and notlocation) specific, they can be made available todownstream firms in other countries. Moreover,competition between suppliers implies thatthe expertise and the technology will be madeavailable to prices close to the marginal costof transfer. The economic logic of this story istherefore that the fixed cost of developing thetechnology is paid by the industries or countriesthat emerge earlier, while the industries orcountries that come later pay only the marginalcost.As a consequence of a certain stage of verticaldisintegration of the industry, the presence ofindependent suppliers that do not produce thedownstream product is important. Downstreamproducers (chemical firms) are less likely to selltechnology to potential competitors (located inless developed economies). Thus, specialisationand division of labour can have a benefit forindustrial growth – that is, the ability and interestof independent suppliers to operate TTs whilenot undermining their competitive position.Given this result, governments should notrandomly screen the entire firm’s pool of richcountries to identify potentially reliable partnersable and willing to take part in TT’s operations.Rather, there is the critical task of identifyingthe right firms that are naturally and logicallyadapted to this kind of operation.5.3.3 PrioritisingDomains to be prioritised by governmentsin their effort to comply with Article 66.2obligations involve two basic criteria:1. Domains where additional incentivesare actually needed. There are domainswhere market incentives are sufficientlyhigh for motivating foreign firms tocommit resources; and2. Domains where there is clear demandfor technology from local entrepreneurs,who then are incentivised to developinnovation projects to suit local needsand markets and are likely to generatespillovers more easily captured by the localeconomy (more easily than those spilloversgenerated through FDI and the productionof knowledge for global markets).To summarise, prioritisation undertaken as aresponse to the 66.2 provision should emphasisedomains that correspond to a certain model ofinnovation, which is central to LDCs: locallyorientedinnovations addressing local needsthrough local entrepreneurial activities thatallow the country to develop absorptivecapacities. These domains are important for

50 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiesgrowth because the spillovers generated inthe course of such projects are likely to becaptured by the local economy. Moreover,these domains need additional incentives sothe donor’s intervention will be effective andrespond fully to the TRIPS provision. This isnot necessarily the case for export-orientedsectors in which market incentives alone aresufficiently strong to motivate firms in richcountries to operate TTs.The other domains – for instance, the exportgoods-oriented manufacturing and processingsectors – are also important but they will, inany case, be served through TTs operating as ajoint product of FDI. TTs in these domains donot require additional incentives provided bythe government. They are the natural resultof rationale and profit-maximising strategiesof foreign firms.Having made the point that local innovationsare socially useful and are likely to generate avirtuous process of diffusion and informationspillovers within the domestic economy doesnot mean that market incentives would beenough to initiate local entrepreneurialinitiative towards these local needs.The figure below illustrates why the issueof incentives needs to be addressed in thisparticular case. DG denotes the demand fromadvanced countries (the global demand); DL isthe local demand; and AC is the average costcurve facing local innovators (with a shapedriven by a fixed cost like R&D).Figure 5. Global and Local Demand and the Need for Special Incentivesfor Local EntrepreneursPACDGD LSource: Trajtenberg, 2008QThe DL slope shows buyers for whom quantitiespurchased are extremely price sensitive,as compared with buyers whose demandsare more inelastic (DG). In this case, localentrepreneurs – if they have the capacity –will likely develop an innovation to serve theglobal demand because doing so would resultin positive profits, whereas, as it stands,serving the local market would not cover thefixed cost.A small R&D subsidy may tip the balance andmake it profitable to innovate for the localmarket, and the local surplus generated maybe significantly larger than the subsidy. Thesocial gains of serving the local market with

ICTSD Programme on IPRs and Sustainable Development51regard to consumer surplus may be significant,as is likely to be the case in the area of medicalcare. Moreover, local spillovers may, in somecases, be more significant and widespread ifinnovating for the local market, if only becauseof the demonstration effects.This case shows clearly that even if thetechnology is transferred for free – due tothe additional incentives received by thetechnology holder – it might not be enough toincentivise local entrepreneurial activities.Subsidies allocated to the local side might beneeded to operate the various tasks that arenecessary (absorption, adaptation, assimilationof improvements) to make the TT a successfuloperation. Investing in innovation orientedtowards local needs may, therefore, requireadditional support from the government to thelocal entrepreneurial activities.5.3.4 Special IPRs for public-private partnershipsThere is a growing awareness of the potentialrole of PPPs to address the issues of accessto and management of IPRs in cases of TTs(see for instance de Boer, 2008, about sucha role in the area of environmentally-soundtechnology transfers).A first issue to be addressed is related tothe fact that most local entrepreneursand research institutions in developingcountries lack even minimal capacity inIPR management and negotiations. The PPPshould, therefore, contribute to clarifyinginstitutional roles, identifying proprietarytechnologies, negotiating ownerships of IPfrom the TT, and guiding the management ofIPRs.A second issue, well addressed by Byerleeand Fischer (2002), is the contribution ofthe high transaction costs of patent searchesto the market failure in the internationaltransfer of proprietary technology. The PPPshould there-fore assume the role of reducingthe cost to developing countries of patentsearches and offer a “one-stop” brokerageservice for buying and selling IP.Finally, as argued by Karapinar and Temmerman(2007), IP-related accessibility issues could beresolved on a purely contractual or free willbasis. However, it is likely that PPPs simply basedon contractual arrangements do not reducethe risk of disagreements between partnerson commercialisation and royalty sharing.Moreover the same authors observe that dealingwith complex licensing and contractual issuesin each PPP results in high transaction costs(ibid.). It is important for PPPs to create a spacefor technology transferring at low transactioncosts. There is thus a need for devising a specialIP-regime tailored for PPPs that could providea more effective environment ensuring bothincentives for technology holders to be partof the transaction and easier access for localentrepreneurs.Depending on the capacity and capability levelof the recipient country, the PPP should eitherfocus on negotiating commercial licensingagreements directly with private companies foraccessing tools and technologies for commercialand emerging markets, or bargaining for accessto products under royalty-free license (Byerleeand Fischer, 2002).5.4 Effectiveness Conditions: Appropriate Donor AssistanceThe use of public money to undertake TT (inorder to create more incentives for the privatepartner) will be considered as efficient not onlybecause the TT meets some measure of successat the end of the day (enhancing efficiency andgrowth in the host country) but also because theoperation would not have been possible withoutthis money. This section discusses the lattercondition, which is obvious but harder for anypublic agency to control and meet.

52 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies5.4.1 The basic modelFigure 6 below shows when and why publicincentives have to be provided to get the privatesector to carry out more TTs toward LDCs. Thisfigure can help us evaluate the effectiveness ofsubsidising TT. The figure shows the expectedreturns on TT in two dimensions, social andprivate (private for the firm which masters thetechnology). The expected total returns on aTT project are the sum of the expected socialand private returns.Government support has several possibleoutcomes. In some cases, government supportgoes to infra-marginal projects, which arehigh return projects that would have beenundertaken anyway (projects in Area C). In thisinstance, government support is just a transferof payment. In our example, a TT aimed atimproving the manufacturing of a good facingdeteriorating terms of trade is a case in point.There is no need for further incentives for acompany in a developed country to transfertechnologies in such domains.Some of the money may go to marginalprojects, such as projects in Areas A and B,which firms will not undertake on their own.If the government cannot really estimate thepotential social return on these projects, thesubsidy causes firms to undertake some TTprojects that are (at the margin) unproductiveprojects (in Area A), and also some projectsthat (at the margin) have a total return inexcess of the opportunity cost of capital thatwould not otherwise have been undertaken(Area B).Figure 6. Projects Targeted by an Optimal Government-IndustryTT ProgrammeSource: Stiglitz and Wallsten, 1999If the government is effective in identifyingprojects with social returns not captured bythe private sector (in Area B), then the supportunambiguously increases economic efficiency.TTs of course have to be operated in manydomains, including export-oriented industry.But, they must be particularly supported inthe domains corresponding to the model ofinnovation, which, as identified above, areentrepreneurial activities dealing with needson local markets that are likely to generatedomestic spillovers. Two reasons have beengiven:

ICTSD Programme on IPRs and Sustainable Development531. These domains are potentially verysignificant for growth because of thespillovers that are likely to be captured bythe local economy;2. These domains need additional incentivesso that the donor’s intervention will beeffective and respond fully to the TRIPSprovision, which is not necessarily the caseof export-oriented sectors in which marketincentives are only sufficiently strong tomotivate the firms of the rich countries tooperate TTs.5.4.2 Implementation issuesStiglitz and Wallsten (1999) argue that thebetter a government is at funding projectsin Area B (projects with relatively high socialbenefits and low private returns), the moreeffective TT programmes will be. However,these authors also argue that it is unclearhow this goal can best be accomplished.Most of these programmes require industryto propose projects and the government todecide whether to subsidise them. Often,government subsidies are less expensive thancapital from other sources, meaning that firmsmay be tempted to look to government beforelooking to other sources for financial support.In other words, there is no reason to believethat firms will provide TT projects only inArea B. Firms may be tempted to first proposeprojects that fall within Area C which, beingsubsidised, will be privately super-profitable.It is up to the government to determine whichof these research projects would benefit LDCsbut would not be privately profitable withouta subsidy. This means that government shouldnot simply fund the best proposals it receives.Instead, it should fund the best amongthose that could not be funded elsewhere.In other words, programme managers whomake funding decisions must reject not onlyprojects of dubious technical merit, but alsotechnologically-sound proposals that are verylikely to yield commercial successes and couldtherefore be funded elsewhere. In practice, itis difficult to follow such allocation principlesand the next section details that many projectsreported as a response of developed countriesto the 66.2 provision should not be subsidisedby government since they clearly fall withinArea C.5.5 Drawing from TRIPS Council Country Reports under Article 66.2The TRIPS Council at the WTO decided in 2002that developed countries should submit reportsabout their policy of incentives and regulatorymeasures that they are implementing inorder to support and promote TTs from thebusiness sector to LDCs. It was specified thatthe information provided should show that theparticular incentive regimes are not simply partof the general overseas development assistance(ODA) that benefits all developing countries butspecifically address the LDCs’ target to promoteTT. It was also requested that developedcountries should provide detailed reports onthe practical functioning of these incentivesand update the Council regularly regarding theresults of these policies.At the TRIPS Council of 2007, however, theBangladesh representative, while appreciatingthe fact that several countries “did the job”and regularly update the Council concerningtheir activities in this area, complained aboutthe lack of information on how the technicalassistance featured in these reports hadactually led to TTs and how the private sector isresponding to these incentives (June 5, 2007).In general, LDCs expressed frustration about thefact that rich countries are not really fulfillingtheir obligations in relation to Article 66.2. 15There are roughly four main domains whereincentives of some kind are provided to encourageTTs with LDCs:• Domain 1: TT to improve the manufacturingand processing of export goods (cacao,coffee, teas) that are facing deterioratingterms of trade;• Domain 2: TT to meet vital needs (drugs andvaccines). This is of course an importantmission but relates less to TT than to R&D

54 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economiescarried out in developed countries, theresults of which are then distributed inLDCs for passive consumption;• Domain 3: TT to build infrastructures. It isnot clear however whether firms contractingwith governments to build an infrastructurehave strong TT duties; and• Domain 4: TT in the area of simple andtraditional capital goods and servicesoriented toward local entrepreneurialactivities in relation to domestic needs.Studying country reports leads to the conclusionthat governments report a high concentration ofprojects in Domain 1, moderate concentrationin Domains 2 and 3, and low concentration inDomain 4 (see Table 5 below as an example).This does not represent a very strong resultin terms of government effectiveness sinceDomain 1 clearly corresponds to Area Cin Figure 6, i.e. projects that would havebeen undertaken anyway. Domain 2 clearlycorresponds to Area B in Figure 6, i.e. projectsthat need to be subsidised but are not pure TTprojects. Domain 4 also corresponds to Area Band does involve true TT operations but veryfew projects are actually undertaken.Some government interventions entail therisk of super-inefficiency. This is the caseof subsidies to projects in Domain 1. Theseprojects do not need further incentives sincethe positive effects of efficiency improvementswill be primarily beneficial to firms fromdeveloped countries importing these goods.As already explained (chapter 4), since thesegoods are characterised by deteriorating termsof trade, successful TTs are likely to worsen theeconomic situation of the considered sector inthe LDCs. In helping such projects, governmentsare therefore being super-inefficient. Theyare helping projects that would be carried outanyway and these projects are not likely toimprove the economic situation of the poorestcountries.To conclude, it appears that the locus ofdecision-making concerning areas and domainsfor TTs are kept in foreign assistance bodies,leading to suboptimal decisions. For example,many projects are actually serving tradepolicy needs rather than innovation policy. Inaddition, governments of developed countriesare not very effective in their allocationdecisions. Many subsidised projects do notin fact require further support to be carriedout. And finally, country reports should bemore transparent and informative regardingwhich domains are supported and helpedthrough TTs and the provision of incentives.The development of a standard typology ofareas for TTs that developed countries woulduse systematically to report on their activitiescould go a long way towards making thesereports more informative (see table below).

ICTSD Programme on IPRs and Sustainable Development55Table 5. Mapping Technology Transfers: An Example Based On CountryReports to the TRIPS Council under Article 66.2TT domainsInstitutionalsourcesPrivateExportorientedgoodsDeterioratingterms of tradeVital needs:education,healthInfrastructureSimple capital goodsand services to serveentrepreneurialactivitiesOrganic cotton(Mali)Cashew nuts,fruits, cotton(Mozambique)Coffee, cashewnuts (Tanzania)Mango (Burkina)Coffee (Zambia)Efficiency in thebrick sector (Nepal)PublicProductionReform ofhealth system(Tanzania)Malaria research(Tanzania)Watermanagement(Burkina)PublicProcurementBridge and roadbuilding(Tanzania)Source: This table is based on a country report submitted to the Council for TRIPS, WTO, 2006.The recent Policy Brief by S. Moon (2008), thatsystematically analyses the reports submitted bycountries to the TRIPS Council, provides similarconclusions. It concludes that the evidenceproduced by her study “does not paint a rosypicture of compliance with Article 66.2.” Moonpoints to the lack of definition of the basicterms (technology transfer, developed country),the absence of data and indicators requirementto measure efforts, and the vagueness of thedescription of the general obligations. A majorityof projects listed by countries as part of theirefforts to comply with 66.2 just miss the target: asignificant proportion of programmes actually donot encourage TT while a majority of them do notspecifically target LDCs. Given this poor activitiesreporting, which creates a very opaque situationand makes it difficult to evaluate efforts andobserve progress (or stagnation) over time andbetween countries, short-term recommendationsare obviously very practical, involving thestandardisation of the reporting procedure, thedefinition of common format and categories, theemphasis on data collection, and presentation.However these needed efforts should not hide thelong-term objective, which has been articulatedand clarified in this paper: making TT an effectivetool for innovation and growth in LDCs.

56 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesConclusionTechnology transfer is a difficult and complexprocess, particularly for LDCs which suffer froma limited exposure to foreign technologies and aweak absorptive capacity. These countries cannotonly rely on FDI and trade to access foreigntechnologies and TTs which are likely to spilloverinto the domestic economy as a result.Effective TTs to LDCs call for new types of PPPswith the involvement of a third party, which isspecialised in linking public donors, private firms,and local entrepreneurial activities to ensure theeffectiveness of the TT operation. The PPP willwork to ensure that the locus of decision-makingon TT is shifted to local actors and authorities. ThePPP would also contribute to clarifying institutionalroles, identifying proprietary technologies andguiding the management of IPRs.Practical means to make the implementation ofTRIPS Article 66.2 more effective should includedeveloped countries providing more effectiveincentives, by offering assistance to projectsthat are socially beneficial but not profitablefor the firms that own and could transfer thetechnology.In this connection, developed countries shouldalso ensure that conditions for efficient TToperations involve the choice of relevant partnerson supply and demand sides, the selection ofthe right areas for focus (related to a clearlyexpressed local demand for technology) and thecreation of organisational forms that will favourthe consolidation of the transfer (absorption,adaptation and subsequent spillover), as well asthe related entrepreneurial dynamic.

ICTSD Programme on IPRs and Sustainable Development57ENDNOTES1 See Linsu Kim (2003), Technology Transfer and Intellectual Property Rights: Lessons fromKorea’s Experience, ICTSD Issue Paper No. 2; Sanjaya Lall, with the collaboration of ManuelAlbaladejo (2003), Indicators of the Relative Importance of IPRs in Developing Countries,ICTSD Issue Paper No.3; UNCTAD-ICTSD Policy Discussion Paper, Intellectual Property Rights:Implications for Development (2003); John Barton (2004), Nutrition and Technology TransferPolicies, ICTSD Issue Paper No. 6; Keith Maskus (2004), Encouraging International TechnologyTransfer, UNCTAD-ICTSD Issue Paper No.7; UNCTAD-ICTSD Resource Book on TRIPS andDevelopment, Cambridge University Press; John Barton (2007), New Trends in TechnologyTransfer: Implications for National and International Policy, ICTSD Issue Paper No. 18.2 These concepts are defined as they are employed in accounting: joint products are twoproducts that are simultaneously yielded from one shared cost and they have comparably high(sales) value. By-products for their part are produced along with a main product. The latterconstitutes the major portion of the total (sales) value. By-products have a considerably lower(sales) value than these main products. We can apply these concepts to TTs, substituting“perceived value to technology holders” for “sales value.”3 The discussion in this section draws heavily on my scholarly exchange and many discussionswith M.Trajtenberg whose synthetic view on these issues for LDCs can be found in Trajtenberg(2007).4 This section draws on the literature that focuses on TT channels, their respective advantagesand shortcomings. It refers especially to Maskus, 2004, Barton, 2007, and the World Bank,2008.5 This is for instance a fundamental weakness of the provisions made under Article 66.2 of theTRIPS, which are intended to encourage firms in developed countries to transfer technologiestowards LDCs. Nothing in this article specifies how (and where) decisions should be made.These provisions are obviously not providing the anticipated results.6 The transitional period expired in this case on 1 January 2005.7 It is also useful to recall that countries like Switzerland or Japan did the same thing in the19th and early 20th centuries to build their own competitive industries.8 Reverse engineering of software is acceptable within the TRIPS framework but legal conditionsare implemented differently at the national level, making the legal framework for reverseengineering difficult to understand and to cope with (see UNCTAD-ICTSD, 2005).9 These industries share the following characteristics that obviously make patents a quiteeffective incentive mechanism: high R&D cost, reverse engineering, and other means ofinnovation reproduction allow competitors for rapid and inexpensive imitations, and lowcosts of manufacturing the final product.10 The author warmly thanks K.Karachalios from the EPO for calling attention to this argumentand providing some very interesting documentation on the Ethiopian case.11 Open-source and community-based user innovation projects can be interpreted as a mechanismto mitigate the formation of asymmetry between final user and commercial producer during aperiod of IP protection strengthening.12 See Roffe et al. (2007) for an overview of the various types of flexibility in TRIPS.

58 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced Economies13 PPPs are not-for-profit organisations that support and coordinate R&D for neglected needs.They are financed by philanthropic and public donors, who have collectively pledged morethan one billion US dollars to PPPs. Almost all PPPs have been created within the last tenyears.14 Recall that TT is a decreasing cost activity (chapter 2, section 3): the more extensive theexperience previously acquired by the organisations involved in the process, the lower thetransfer costs in relation to total project size.15 See Correa (2007) for a careful analysis and discussion of the country reports and moregenerally of the problems raised by the Article 66.2.

ICTSD Programme on IPRs and Sustainable Development59REFERENCESAbbott, F. 2003. The Competition Provisions in the TRIPS Agreement: Implications for TechnologyTransfer. Presented at a joint WIPO-WTO workshop.Angeles, L. 2005. Should Developing Countries Strengthen Their Intellectual Property Rights?Topics in Macroeconomics Vol.5, Issue 1.Appleyard M., Hatch N. and Mowery D., 1996. Managing the Development and Transfer of ProcessTechnologies in the Semiconductor Manufacturing Industry. IIASA Working Paper.Arora, A., 1995. Licensing Tacit Knowledge: Intellectual Property Rights and the Market for KnowHow. Economics of Innovation and New Technology Vol.4, Issue 1.Arora, A., 1996. Contracting for Tacit Knowledge: The Provision of Technical Services in TechnologyLicensing Contracts. Journal of Development Economics, Vol. 50.Arora, A., 2007. Intellectual Property Rights and the International Transfer of Technology: SettingOut an Agenda for Empirical Research in Developing Countries. WIPO Conference.Arora, A., Fosfuri, A. and Gambardella A., 2001. Specialized Technology Suppliers, InternationalSpillovers and Investment: Evidence From the Chemical Industry. Journal of DevelopmentEconomics Vol. 65.Arora A., Fosfuri, A. and Gambardella, A., 2002. Markets for Technology: Economics of Innovationand Corporate Strategy. Cambridge: MIT Press.Athreye, S and Cantwell, J., 2007. Creating competition? Globalisation and the Emergence of NewTechnology Producers. Research Policy Vol. 36, 2.Barton J., 2007. New Trends in Technology Transfer: Implications for National and InternationalPolicy. ICTSD Programme on IPRs and Sustainable Development, Issue Paper No.18,International Centre for Trade and Sustainable Development, Geneva, Switzerland.Baumol, W., 2002. The Free Market Innovation Machine. Princeton, New Jersey: Princeton UniversityPress.Bessen, J. and Meurer, M., 2008. Patent Failure. Princeton, New Jersey: Princeton UniversityPress.Blomström, M. and Kokko, A., 1998. Foreign Investment as a Vehicle for International TechnologyTransfer. In Creation and Transfer of Knowledge, ed. Navaretti, Dasgupta, Mäler andSiniscalco. New York: Springer.Branstetter, L., Fisman R. and Fritz Foley C., 2005. Do Stronger IPR Increase TT? Empirical Evidencefrom US Firm Level Panel Data. New York: Columbia University.Branstetter, L., Fisman, R., Fritz Foley, C. and Saggi K., 2007. Intellectual Property Rights, Imitation,and Foreign Direct Investment: Theory and Evidence. NBER Working paper 13033.Byerlee, D. and Fischer K., 2002. Accessing Modern Science: Policy and Institutional Options forAgricultural Biotechnology in Developing Countries. World Development Vol.30, 6.

60 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesCassier, M. and Correa, M., 2005. Intellectual Property and Public Health: Copying of HIV/AIDS Drugsby Brasilian and Private Pharmaceutical Laboratories. Draft paper.Chooi, C., Chang, L. and Zhang, Y., 1994. Technology transfer and International Organizations: TheQuestion of Localisation. Science, Technology and Development 12(1/2).Cockurn, I., 2007. Is the Market for Technology Working? Obstacles to Licensing Inventions, andWays to Reduce Them. Paper presented at the Monte Verita Conference The Economics ofTechnology Policy in Ascona, Switzerland.Correa, C., 2007. Intellectual Property in LDCs: Strategies for Enhancing Technology Transfer andDissemination. Background Paper No.4: UNCTAD: The Least Developed Country Report2007.Corsi, F., 2007. An Example of Use of Appropriate Technology: Setting Up of Hearing Aid Facilitiesin Ethiopia. Draft Paper.David P.A. Institutional Lock In: The Path Dependent Evolution of Intellectual Property Rights and theRelentless Encroachment Upon the Public Knowledge ‘Commons’ in Science and Technology.Unpublished Draft.David P.A., 1996. Re-thinking Technology Transfer in the 1990s: Incentives, Institutions andKnowledge-Based Industrial Development. In Chinese Technology Transfer in the 1990s:Current Experience, Historical Problems and International Perspectives, ed. Feinstein andHowe. United Kingdom: Edward Elgar.De Boer, Y. Address, 2008. European Patent Forum, Lubjlana, 2008Enos J., 1996. The Adoption of Innovations and the Assimilation of Improvements. In ChineseTechnology Transfer in the 1990s: Current Experience, Historical Problems and InternationalPerspectives, ed. Feinstein and Howe. United Kingdom: Edward Elgar.Enos J., 1998. In Pursuit of Science and Technology in Sub-Saharan Africa: The Impact of StructuralAdjustment Programs. UNU/INTECH Studies in New Technology and Development.Enos, J., Lall, S. and Yun M., 1998. Transfer of Technology: An Update. Asian-Pacific EconomicLiterature.European Community, 2007. Implementation of Article 66.2 of the TRIPS Agreement. Presented atthe WTO Council for TRIPS on February 12, IP/C/W/480/Add.5Foray D., 2004. The Economics of Knowledge. Cambridge: MIT Press.Ghosh R. and Soete L., 2006. Information and Intellectual Property. Working paper UNU-MERIT,Number 29.Hall, B., 2004. Government Policy for Innovation in Latin America. Regional Studies Program of theChief Economist for Latin America, the World Bank.Hall, B., 2001. The Global Nature of IP. Paper presented at the Toronto IP Conference in March.Hausmann, R. and Rodrik, D., 2002. Economic Development as Self-discovery. NBER Working Paper8952.

ICTSD Programme on IPRs and Sustainable Development61Hirsch, W., 1952. Manufacturing Progress Functions. Review of Economics of Statistics, 34, 2.Ho, S., 1997. Technology Transfer to China During the 1980s- How effective? Some Evidence fromJiangsu. Pacific Affairs Vol.70, Number 1.Hobday M., 1995. East Asian Latecomer Firms: Learning the Technology of Electronics. WorldDevelopment 23(7).ICTSD, 2007. Views on the Future of the Intellectual Property System. ICTSD Programme onIntellectual Property Rights and Sustainable Development, Selected Issue Brief No.1Karapinar B. and Temmerman, M., 2007. Benefiting From Biotechnology: Pro-poor IPRs and PublicprivatePartnerships. NCCR trade regulation (Swiss National Centre of Competence). Workingpaper 2007/35.Kostecki, M., 2006. Intellectual Property and Economic Development: What Technical Assistanceto Redress the Balance in Favour of Developing Countries. ICTSD Programme on IPR andSustainable Development, Issue Paper number 14, International Centre for Trade andSustainable Development, Geneva, Switzerland.Krugman P., 1979. Increasing Returns, Monopolistic Competition and International Trade. Journal ofInternational Economics 9.Lanjouw, J., 1998. The Introduction of Pharmaceutical Product Patents in India: Heartless Exploitationof the Poor and Suffering? NBER Working Paper series 6366.Mansfield E., 1982. Technology Transfer, Innovation and Public Policy. In The Transfer and Utilizationof Technical Knowledge, Sahal. Lanham: Lexington Books.Mansfield E., 1995. Intellectual Property Protection, Direct Investment and Technology Transfer.International Finance Corporation Discussion Paper 27.Maskus, K., 2004. Encouraging International Technology Transfer. UNCTAD-ICTSD Programme onIPRs and Sustainable Development, Issue Paper No.7, International Centre for Trade andSustainable Development, Geneva, Switzerland.Matthew, D., 2005. TRIPS Flexibilities and Access To Medicines in Developing Countries: The ProblemWith Technical Assistance and Free Trade Agreement. EIPR.Maurer, S., Sali, A. and Rai, A., 2004. Finding Cures for Tropical Disease: Is Open Source the Answer?Public Library of Science: Medicine.Mendi, P., 2007. Trade in Disembodied Technology and Total Factor Productivity in OECD Countries.Research Policy, Vol. 36, No,1, 121-133.Moon, S., 2008. Does TRIPS Art.66.2 Encourage Technology Transfer to the LDCs? UNCTAD-ICTSDPolicy Brief No.2, Geneva.Moran, M., 2005. A Breakthrough in R&D for Neglected Diseases: New Ways to Get the Drugs WeNeed. PLoS Medicine, 2(9).Polanyi, K., 1996. The Tacit Dimension. New York: Doubleday.

62 Dominique Foray — Technology Transfer in the TRIPS Age: The Need for New Types of Partnershipsbetween the Least Developed and Most Advanced EconomiesRoffe, P. and Vivas D. (with G.Vea), 2007. Maintaining Policy Space for Development: A Case Studyon IP Technical Assistance in FTAs. ICTSD Programme on IPR and Sustainable Development,Issue Paper No.19, International Centre for Trade and Sustainable Development, Geneva,Switzerland.Samuelson P. and Scotchmer S., 2001. The Law and Economics of Reverse Engineering. University ofCalifornia at Berkeley.South Centre, 2005. The Potential Impacts of Nano-scale Technologies on Commodity Markets: TheImplications for Commodity Dependent Developing Countries. T.R.A.D.E Research Paper 4.State Secretariat for Economic Affairs (SECO), 2005. Cleaner Production Centres: Swiss Support andExperience. Paper presented at the OECD workshop, “Multilateral environmental agreementsand private investment,” in Helsinki, Finland.Stiglitz, J., 1991. Social Absorption Capability and Innovation. Stanford University CEPR Publicationnumber 292.Stiglitz, J. and Wallsten, S., 1999. Public-private Partnerships. American Behavioral Scientist,Vol.43, 1.Teece D., 1996. The Multinational Corporation and the Resource Cost of International TechnologyTransfer. Cambridge: Balinger Books.Teece D., 1997. Technology Transfer by Multinational Firms: The Resource Cost of TransferringTechnological Know How. Economic Journal 87.Trajtenberg M., 2007. Innovation Policy for Development. Paper presented at the Monte VeritaConference The Economics of Technology Policy in Ascona, Switzerland.UNCTAD-ICTSD, 2005. Resource Book on TRIPS and Development. Cambridge: Cambridge UniversityPress.Von Hippel, E., 1994. Sticky Information and the Locus of Problem Solving: Implications for Innovation.Management Science, Vol.40, 4.World Bank, 2008. Global Economic Prospects. Washington D.C.: The World Bank Group.World Intellectual Property Organization, 2006 and 2007. Patent Report Geneva: World IntellectualProperty Organization.

SELECTED ICTSD ISSUE PAPERSAgricultural Trade and Sustainable DevelopmentValue Chains and Tropical Products in a Changing Global Trade Regime. By Charles Mather. Issue Paper No. 13. May 2008Trade Effects of SPS and TBT Measures on Tropical and Diversification Products. By Anne-Celia Disdier, Belay Fekadu, Carlos Murillo, Sara A. Wong.Issue Paper No. 12. May 2008Tropical and Diversification Products Strategic Options for Developing Countries. By Santiago Perry. Issue Paper No. 11.March 2008Implications of proposed modalities for the Special Safeguard Mechanism: a simulation exercise. Issue Paper No.10 Raul Montemayor, November 2007Trade and Sustainable Land Management in Drylands. Selected Issue Briefs, 2007.A Comparison of the Barriers Faced by Latin American and ACP Countries’ Exports of Tropical Products.Issue Paper No. 9 by Jean-Christophe Bureau, Anne-Celia Disdier and Priscila Ramos, 2007.South-South Trade in Special Products. Issue Paper No.8 by Christopher Stevens, Jane Kennan and Mareike Meyn, 2007.Competitiveness and Sustainable DevelopmentLooking for a meaningful Duty Free Quota Free Market Access Initiative in the Doha Development Agenda. By David Laborde. Issue Paper No.4, 2007Impact of Origin Rules for Textiles and Clothing on Developing Countries. By Munir Ahmad. Issue Paper No.3, 2007Special and Differential Treatment for Small and Vulnerable Countries Based on the Situational Approach.Issue Paper No. 2 by Werner Corrales-Leal, Felipe Baritto, and Sarah A. Mohan, 2007.Basic Concepts and Proposals on the use of Policy Spaces in Trade-supported Strategies for Sustainable Development.Issue Paper No. 1 by Werner Corrales-Leal, 2007.Dispute Settlement and Legal Aspects of International TradeBurden of Proof in WTO Dispute Settlement: Contemplating Preponderance of Evidence Issue Paper No. 9 by James Headen Pfitzer and Sheila Sabune, 2009.Cross-Retaliation in TRIPS: Options for Developing Countries. Issue Paper No. 8 by Frederick. M Abbott, 2009.Suspension of Concessions in the Services Sector: Legal, Technical and Economic Problems.Issue Paper No. 7 by Arthur E. Appelton, JD, 2009.Trading Profiles and Developing Country Participation in the WTO Dispute Settlement System.Issue Paper No. 6 by Joseph Francois and Henrik Horn, 2008.Developing Countries, Countermeasures and WTO Law: Reinterpreting the DSU against the Background of International LawIssue Paper No. 5 by Andrea Bianchi and Lorenzo Gradoni, 2008.Does Legal Capacity Matter? Explaining Dispute Initiation and Antidumping actions in the WTO.Issue Paper No. 4 by Marc L. Busch, Eric Reinhardt and Gregory Shaffer, 2008.Fisheries, International Trade and Sustainable DevelopmentFisheries, Aspects of ACP-EU Interim Economic Partnership Agreements: Trade and Sustainable Development ImplicationsIssue Paper No.6, by Liam Campling, 2008Fisheries, International Trade and Sustainable Development. Policy Discussion Paper, by ICTSD, 2006.Aquaculture: Issues and Opportunities for Sustainable Production and Trade. Issue Paper No. 5 by Frank Asche and Fahmida Khatun, 2006.Market Access and Trade Liberalisation in Fisheries. Issue Paper No. 4 by Mahfuz Ahmed, 2006.Trade and Marketplace Measures to Promote Sustainable Fishing Practices. Issue Paper No. 3 by Cathy Roheim and Jon G. Sutinen, 2006.Fisheries Access Agreements: Trade and Development Issues. Issue Paper No. 2 by Stephen Mbithi Mwikya, 2006.Intellectual Property Rights and Sustainable DevelopmentThe Global Debate on the Enforcement of Intellectual Property Rights and Developing Countries.Issue Paper No.22, 2009.Intellectual Property and Competition Law Issue Paper No.21, by Carlos M. Correa, 2007.Intellectual Property Provisions in European Union Trade Agreements: Implications for Developing Countries.Issue Paper No. 20 by Maximiliano Santa Cruz S., 2007.Maintaining Policy Space for Development: A Case Study on IP Technical Assistance in FTAs.Issue Paper No. 19 by Pedro Roffe and David Vivas with Gina Vea, 2007.New Trends in Technology Transfer: Implications for National and International Policy. Issue Paper No. 18 by John H. Barton, 2007.Trade in Services and Sustainable DevelopmentMaritime Transport and Related Logistics Services in Egypt, by Ahmed F. Ghoneim, and Omneia A. Helmy, Issue Paper No 8, December 2007Opportunities and Risks of Liberalising Trade in Services in Pakistan, By Abid A. Burki, Issue Paper No 7, December 2007Regulatory Principles for Environmental Services and the General Agreement on Trade in Services, by Massimo Geloso Grosso, Issue Paper No 6,December 2007Opportunities and Risks of Liberalising Trade in Services in Mozambique, by Alberto Teodoro Bila, Eduardo Mondlane, Hélder Chambal and ViriatoTamele, Issue Paper No 5, December 2007Opportunities and Risks of Liberalizing Trade in Services in Tanzania, by Daima Associates Limited, National Consultant, Issue Paper No.4, December2007Trade and Sustainable EnergyClimate Change and Trade on the Road to Copenhagen Policy Discussion Paper, 2009.Trade, Climate Change and Global Competitiveness: Opportunities and Challenge for Sustainable Development in China and Beyond,Selected Issue Briefs No.3 by ICTSD, 2008.Intellectual Property and Access to Clean Energy Technologies in Developing Countries: An Analysis of Solar Photovoltaic, Biofuel and WindTechnologies. Issue Paper No. 2 by John H. Barton, 2007.Climate, Equity, and Global Trade. Selected Issue Briefs No. 2, 2007.The WTO and Energy: WTO Rules and Agreements of Relevance to the Energy Sector. Issue Paper No. 1 by Julia Selivanova, 2007.Linking Trade, Climate and Sustainable Energy. Selected Issue Briefs, 2006.These and other ICTSD resources are available at http://www.ictsd.org

www.ictsd.orgThe International Centre for Trade and Sustainable Development (ICTSD) has been active in the field of intellectual propertysince 1997, amongst others, through its Programme on Intellectual Property Rights (IPRs) and Sustainable Development. Onecentral objective of the Programme has been to facilitate the emergence of a critical mass of well informed stakeholdersin developing countries including decision makers, negotiators as also the private sector and civil society who will be ableto define their own sustainable human development objectives in the field of IPRs and effectively advance them at thenational and international levels.The Programme has generated an Issue Paper Series on Intellectual Property Rights and Sustainable Development with theintention of offering a clear, jargon-free synthesis of the main issues to help policy makers, stakeholders and the publicin developing and developed countries to understand the varying perspectives surrounding different IPRs, their known orpossible impact on sustainable livelihoods and development, and different policy positions over the TRIPS Agreement andother relevant international intellectual property arrangements.This Issue Paper Series is the consequence of a participatory process involving trade negotiators, national policy makers,as well as eminent experts in the field, the media, NGOs, international organizations, and institutions in the North and theSouth dealing with IPRs and development. Previous publications under this Series include:• The Global Debate on the Enforcement of Intellectual Property Rights and Developing Countries.Issue Paper No.22, 2009.• Intellectual Property and Competition Law. Issue Paper No. 21 by Carlos M. Correa, 2007.• Intellectual Property Provisions in European Union Trade Agreements: Implications for Developing Countries.IssuePaper No. 20 by Maximiliano Santa Cruz S., 2007.• Maintaining Policy Space for Development: A Case Study on IP Technical Assistance in FTAs.Issue Paper No. 19 by Pedro Roffe and David Vivas with Gina Vea, 2007.• New Trends in Technology Transfer: Implications for National and International Policy.Issue paper No. 18 by John H. Barton, 2007.• Exceptions to Patent Rights in Developing Countries.Issue paper No. 17 by Christopher Garrison, 2006.• Protecting Traditional Knowledge: Pathways to the Future.Issue paper No. 16 by Graham Dutfield, 2006.• The International Copyright System: Limitations, Exceptions and Public Interest Considerations for DevelopingCountries.Issue Paper No. 15 by Ruth L. Okediji, 2006.• Intellectual Property and Economic Development: What Technical Assistance to Redress the Balance in Favour ofDeveloping Nations?Issue Paper No. 14 by Michel Kosteki, 2006.• Utility Models and Innovation in Developing Countries.Issue Paper No.13 by Uma Suthersanen, 2006.• Intellectual Property Provisions of Bilateral and Regional Trade Agreements in Light of US Federal Law.Issue Paper No.12 by Frederick M. Abbott, 2006.• Technical Assistance for the Formulation and Implementation of Intellectual Property Policy in DevelopingCountries and Transition Economies.Issue Paper No.11 by Tom Pengelly, 2005.ABOUT ICTSDFounded in 1996, the International Centre for Trade and Sustainable Development (ICTSD) is an independent non-profitand non-governmental organisation based in Geneva. By empowering stakeholders in trade policy through information,networking, dialogue, well targeted research and capacity building, the Centre aims to influence the international tradesystem such that it advances the goal of sustainable development.