Transfers

Transfers of proprietary research inputs to Centres

Proprietary research inputs come in bundles composed of four elements: (i) codifiable information; (ii) materials embodying the information; (iii) IPR – rights to use and benefit from the information; and (iv) human capital, especially tacit knowledge. Transfers of proprietary technology, be they gifts or exchanges, are complex affairs, and these four elements need to be distinguished carefully. Licences and material transfer agreements (MTAs) may be used for transferring IP and materials, respectively.

Many of the problems relating to Centres’ use of research inputs owned by others can be summarized in the question: What must be provided in return? The quid pro quo may be money, or perhaps a restriction beyond the input’s immediate use. For example, a research licence restricts use rights to research only. Such restrictive clauses hamper technology transfer. It may be better for the Centre to pay royalties for a licence rather than obtain it free if less restrictive conditions can be obtained (Nottenburg et al. , 2002).

According to a 1998 survey, Centres frequently obtained permission through MTAs, licences, or sublicences to use proprietary inputs. However, almost as frequently, permission to use proprietary inputs was absent or unknown (Cohen et al ., 1999). The primary risk of unauthorized use is not legal action against Centres, but rather IP owners’ reluctance to share their properties with Centres in the future (CGIAR, 1998, p. 6). The outright purchase of IP by Centres is much less common than the use of proprietary inputs. We know of only one such case, the purchase of the rights to a Bt gene by a public-sector consortium led by IRRI (Byerlee and Fischer, 2001, p. 13). In apparent contrast to the CG Centres, Latin American NARS do frequently purchase IP rather than license it (Cohen et al ., 1999). A variant of this approach would be to contract with the private or public supplier, perhaps through competitive bidding, to develop a specific tool, while retaining ownership of the product (Byerlee and Fischer, 2001, p. 10).

Motivations for a firm to ‘donate’ proprietary inputs include: (i) limited commercial viability of many applications that benefit the poor; (ii) public relations; (iii) connections with a network of non-profit organizations; (iv) the crossing and testing of crops in different environments, generating valuable data for subsequent crop improvement; (v) an opportunity to demonstrate the benefits of the technology; (vi) encouragement of governments to put in place regulations on safety and intellectual property; (vii) strengthening market presence; and (viii) philanthropic motives (CGIAR, 1998, pp. 9–10).

When will companies be un willing to license? Companies’ considerations here revolve around control of technology:

[C]ompanies are unwilling to license if it leads to their losing control over the licensed technology. Lack of control may cause technical problems: for example, in the case of Bt maize it could result in the companies being unable to ensure a suitable management regime that would minimize the build-up of insect resistance. It will also cause major commercial problems if the licensed technology is used to compete with the licensor in profitable markets (CGIAR, 1998, p. 10)

Because licences and MTAs may contain restrictions, it makes sense for Centres to negotiate terms rather than simply accept a gift. Awareness of the jurisdictional extent of the IP is a prerequisite for such negotiations (Binenbaum et al ., 2003).

MTAs and licences may sometimes be used in the same R&D trajectory at different points in time. The MTA may, in effect, be an incomplete contract, to be followed up later by licensing negotiations (Byerlee and Fischer, 2001, p. 10). As is typical of incomplete contracting, the party renegotiation has a bargaining disadvantage. This is known as the ‘hold-up problem’.

Problems with proprietary inputs can arise in relations with non-profit as well as for-profit organizations. For example, countries may be tempted to stake out claims to ‘their’ genetic resources. The Convention on Biological Diversity, which entered into force in 1993, provided a framework for such claims. CG Centres signed an agreement with the Food and AgricultureOrganization (FAO) in 1994 whereby most of the materials (so called ‘designated material’) in the Centres’ genebanks are held ‘in trust’ as a common property resource. Under this in-trust agreement, Centres may not seek IPR over designated materials, and are required to ensure that subsequent recipients will not do so either. These provisions aim to reassure nations that materials provided to the CG will not be appropriated by anyone, thus providing them with incentives to keep sharing genetic resources. However, these incentives may not be especially strong, in part because NARS are not party to the agreement. To provide a firmer basis for continued germplasm exchange, many countries were involved in 23 years of negotiations dubbed ‘the International Undertaking on Plant Genetic Resources for Agriculture’. In November 2001, a draft International Treaty on Plant Genetic Resources was adopted by 116 nations. Japan and the USA abstained from voting on the treaty although the USA did sign the treaty in late 2002. The treaty, which comes in to force when ratified by 40 countries, ‘establishes a multilateral system of access and benefit sharing for 64 crops and plants [including maize, wheat and rice but excluding soybeans, tomatoes, groundnuts and tropical grasses] that are fundamental to food security (FAO, 2002)’. The intent is to ensure the pool of genetic resources encompassed by the treaty will be freely available to plant breeders in countries that adopt the treaty, in exchange for royalties if the seeds are used to develop commercial varieties. Determining these royalties implies keeping track of breeding pedigrees, an issue yet to be resolved in the context of this treaty. Problems might also arise with technologies developed by non-profit organizations. In the USA, the 1980 Bayh-Dole Act mandated that the US Government cede ownership of intellectual property, emanating from government-sponsored research, to the recipient institution. As a result, some universities now hold significant IP portfolios. Negotiating use rights for publicly held intellectual property can be more problematic than for IP held by private firms: public agencies like universities may be hamstrung by regulations or bureaucracies, or royalty sharing arrangements with faculty (Nottenburg et al. , 2002).

The party transferring the proprietary input is likely to possess superior information about its cost and potential, and may exploit this information advantage in negotiations. Hence, it is valuable for the Centres to have information about cost, commercial value, and jurisdictional validity of proprietary inputs. It may be cost-effective for such information to be provided through System-wide services such as CAS or the informatics systems related to intellectual property being developed at CAMBIA.

Exchanging materials for data

Contractual provisions often mandate data-for-materials exchange. A standard MTA reads: ‘Recipients are requested to furnish [Centre] with data and information collected during evaluations of the material’ (SGRP, 2000, p. 12). The International Network for the Genetic Evaluation of Rice (INGER), established in 1977, is an example of a system of transfers of germplasm and information unencumbered by IP. Four Centres and scientists from NARS involved in rice breeding participate in INGER. The requirement – included in MTAs – that recipients supply INGER with relevant varietal performance data on the material distributed for evaluation via the network is crucial in this arrangement, which can be seen as a repeated game. Failure to collect and report data would save costs in the short run, but would eventually lead to exclusion from the network. On the other hand, interactions in a network like INGER may often be based on a vague sense of quid pro quo and on a culture of information sharing and cooperating for a common purpose, rather than on explicit costs– benefit calculations. An increase in ‘territorial’ behaviour in recent years has reduced the scope of INGER.

Funding issues

A fundamental issue in funding is the degree of influence exerted by the funding entity on the recipient’s research. For the past decade, the CG’s budget has failed to grow while the demands on the System in terms of its commodity coverage, research problem orientation, and accountability, have continued to expand. At the same time, an increasing proportion of the budget consists of restricted (programme and project) funding. In addition to concerns about undue influence, this has also raised the transaction costs of funding (CGIAR, 2000b, pp. 17–18). In the late 1990s, there was a growing awareness in the System that nontraditional sources of funding may have been neglected. This awareness led to Future Harvest, a joint initiative of the Centres, intended to deal with the problems of stagnant overall funding and declining unrestricted support. Future Harvest promotes the CG and its Centres to the public and others, targeting fundraising activities to foundations and companies not traditionally aligned with agriculture or the CGIAR (CGIAR, 2000c).

Another hitherto untapped source of funding is competitive grants:

[I]t is possible that new sources of finance derived from Ministries of Science could also be accessed by the CGIAR if the funds were internally allocated on a competitive basis. At the moment, the CGIAR is not able to receive these funds because it does not have an internal competitive allocation mechanism (CGIAR, 2000b, p. 18)

The Global Challenge Programs are designed to be such an internal competitive allocation. The blueprints for Global Challenge Programs address the inter-Centre competition issue by stipulating that they ‘require cooperative research, going beyond individual Center mandates’ mechanism (CGIAR, 2001, p. 10).

Some of the Centres’ clients reach beyond subsistence or low-income agriculture and may be willing to contribute to Centre R&D. A variation of this scenario is the catalysis of R&D consortia operating outside of the CG as an alternative to in-house R&D. At least two such consortia (FLAR and CLAYUCA, discussed below) demonstrate the potential for engaging agroindustries based in developing countries in research programmes similar to those at CG Centres. A sound IP arrangement may be critical to the success of such consortia.

Privately sponsored public R&D may combine elements of giving with elements of purchasing or collaboration. A prominent issue in sponsored research involves the disposition of rights to any IP arising from the research. For example, an agreement whereby Novartis (now Syngenta) funds plant biotech research at the University of California, Berkeley, grants Novartis the first right to negotiate for IPR to a fraction of the research results equivalent to the share of the budget provided by Novartis. Such provisions strengthen sponsor incentives but tend to spark controversy among the recipient’s stakeholders, as has happened in this case. Centres are confronted with similar, potentially controversial, funding opportunities. Such difficulties are not confined to arrangements with the private sector. Increasingly government agencies that sponsor CG research seek a say over resultant IPR. A wide variety of such contractual clauses exist, some stating that IPR should not be sought on results of R&D funded by the donor; others that some might be expected, perhaps with defensive objectives; and still others that IPR be ceded to the donor. Some Centres have paid insufficient attention to such clauses, although recent IP audits have heightened awareness of these matters.

Technology transfer to the developing world

The CG System’s technology transfer mechanisms function reasonably well – at least as far as the System’s own research products are concerned. Still, the System’s technology transfer role could be enhanced by an improved technology information system, including invention disclosures. The transfer of technology developed by others is an even greater challenge. ISAAA and ABSP are examples of other non-profit initiatives that may be more specialized than the CG System in the transfer of technology developed by others. Both benefit from links to US universities, to the for-profit sector (especially ISAAA), and to USAID (in the case of ABSP). CIIFAD (like ISAAA, hosted by Cornell) is an example of a university institute that is active in technology transfer to developing countries. A particularly interesting case concerns transgenic virusresistant papaya. 6

The more advanced NARS have the capacity for an office of technology transfer. The Brazilian Agricultural Research Corporation (EMBRAPA), which accounts for about half the country’s agricultural R&D spending, already possesses such an office, called the Intellectual Property Secretariat (Maredia et al. , 2000). A natural role for the Intermediary Biotechnology Service of ISNAR (IBS) and CAS is to aid an international network of technology transfer offices.

Multinationals’ presence in LDCs is mostly due to purchases of and partnerships with local firms rather than direct investment. Monsanto has reportedly purchased 16 local seed companies in Brazil, and most large Indian seed companies have formed alliances with global life sciences companies. Only a few LDC companies have a capacity in biotechnology research, and in nearly all cases, this research is carried out as part of an alliance with one of the global companies (Byerlee and Fischer, 2001).

In the traditional CG model of technology transfer, involving public entities in developing countries, there was no need for commercial development of Centre R&D outputs. This has begun to change. In certain kinds of technology, the for-profit sector often possesses superior development, production and distribution capacity. Examples include vaccines and biopesticides (see the LUBILOSA project below). In crop breeding, traditionally the CG’s most important R&D activity, technology transfer via developing-country firms, including a downstream R&D role for those firms, is becoming more common in the Centres. For example, a partnership between Centro Internacional de Agricultura Tropical (CIAT) and Papalotla, a Mexican seed firm, involves the development and commercialization of new hybrid grass varieties of the genus Brachiaria for cattlefarming. Multiplication of the grass seeds and other downstream activities require investments that are too large for most NARS. Thus, CIAT considered its best option to develop, multiply, and distribute the new grass varieties widely would be to partner with a firm such as Papalotla with established distribution channels. In this arrangement, Papalotla funds R&D through advance payments on future royalties; registers CIAT as the owner of the new grass varieties in relevant countries; and may sub-license the grasses to local firms. Without this IP arrangement – apparently novel in the CG System – this partnership might not have been possible. Moreover, because of Papalotla’s exclusive rights, the firm can count on long-term relationships with its farmer customers. This makes it more feasible and attractive for the firm to play a role that includes farmer education, extension, and feedback from farmers (E. Binenbaum, P.G. Pardey and B.D. Wright, 2002, unpublished). Technology transfer in cases like LUBILOSA (see below) and Papalotla is similar to the prevailing pattern (after the Bayh-Dole Act) of university–industry technology transfer in the USA: the non-profit IP owner licenses to a for-profit partner. Here, IP serves as an incentive tool for private-sector development and commercialization of the technology (Mowery et al ., 2001; Parker et al ., 2001). To maintain an effective system of technology transfer, the CG System must make market segmentation an integral part of IP arrangements with technology suppliers. Such IP arrangements may be triangular – e.g. a research licence for a Centre plus a licence for NARS distribution of research products (Byerlee and Fischer, 2001, p. 18).

R&D outputs of interest to for-profits

R&D for the poor, a core activity of the CG System, may yield – as by-products – outputs of interest to the for-profit sector. Such outputs include varieties and genetic traits with food and non-food (e.g. fibres, energy) applications; methods for genetic conservation or transformation; methods and genes related to propagation and reproduction (e.g. apomixis); vaccines and (bio-)pesticides; and mechanical equipment. R&D outputs are a subset of valuable Centre assets. Well-organized direct and indirect information about the costs and benefits of these to private-sector players would be highly useful in order to notice opportunities for funding and collaboration, and anticipate strategic interactions. Given the present situation of ad-hoc information provision, a marginal investment in an information system would probably yield a substantial pay-off.