ditions for the use of AKST at different levels. The trend towards privatization of AKST is particularly felt in the sphere of plant breeding and biotechnology.

Stimulating private investments in research. IPRs are meant to stimulate private investment in research, but even though evidence of such effects in various industries may be available, it is very weak in the agricultural sector, notably plant breeding (Pray, 1991; Alston and Venner, 2000). IPRs protected in LDCs with a limited research capacity are more likely to improve access to proprietary technologies from abroad (e.g., Bt cotton). There are claims of positive effects of protection of breeder's rights for a selective number of cases; however, without taking into account alternative explanations for the observed effects and without providing data for other crops in the same case study countries (UPOV, 2005). Other studies show inconclusive results of the value of IP protection for the plant breeding sector in LDCs. Based on evidence in five developing countries, IPRs may support the development of a private seed industry, but only when this sector has reached a certain level of maturity; IP protection is not a major stimulus for initial investments in the sector (World Bank, 2006).

Public-private partnerships in research. In a market system, IPRs provide a way to share benefits among the different chain partners through the transfer of technology fees (royalties). They are the basis for negotiating partnerships in research between private and public partners, notably private IPR-holders and public research institutions in accessing technologies in a certain country. However, the reliance on negotiated license agreements also introduces prospects for unequal sharing of benefits based on differences in negotiating capabilities and power of the partners.

Financial support to the public research systems through IPRs. Even though IPRs fit in a commercial approach to innovation, it is, in many countries, the public sector research institutions that promote the introduction of IPRs in agriculture. This promotion is based primarily on a perception that these institutes may obtain significant revenue when their inventions (e.g., plant varieties) may be protected. This revenue is welcomed when there is underinvestment in public research (common in many countries since the 1990s), but may be viewed differently if such benefits can only be obtained in commercial markets (e.g., seed markets). Reliance on IPR based revenues is likely to lead to a change in public research priorities, in some cases to commercial crops like maize and oil crops to the detriment of research on small grains and pulses, and to benign ecologies and market oriented farmers, to the detriment of a small-scale farmer focus (Fischer and Byerlee, 2002). Such research shifts may fit in market orientation priorities of national development strategies, but may at the same time challenge to some extent the public tasks of contributing to poverty alleviation and household nutrition security (Louwaars et al., 2006).

      The most common alternative strategy for a public research institute may be to publish its innovations, i.e., place in the public domain. This strategy reduces opportunities to obtain financial revenue and may limit public-private partnerships.

Challenges to technology transfer-thickets of rights. Even though license agreements may promote technology transfer by clarifying roles and responsibilities, IPRs may also pose serious limitations to research and the use of technologies in development. Particularly in advanced research, so-called thickets of rights lead to the tragedy of the anti-commons leading to underinvestment and underutilization of technologies (Heller and Eisenberg, 1998). Property rights on research tools, processes and products create very complex situations for researchers and their institutions, potentially leading to underutilization of technologies. Research institutes have to learn how to establish and negotiate their freedom to operate on these technologies. The quality and enforceability of the claims of a patent may significantly differ between jurisdictions; negotiating access to a technology can be very difficult when unequal partners are involved; so-called humanitarian use licenses (license on a technology for R&D for development with "soft" conditions) may be granted when the use of a technology is unlikely to challenge the commercial interests of the rights holder, but the "small print" license details can create significant obligations for the recipient.

     These are new policy challenges for most developing countries, the actual impact of which cannot be readily assessed yet (World Bank, 2006). The rights on enabling technologies create challenges for producing public goods, which has been the main focus of public research, and more specifically for the centers of the Consultative Group on International Agricultural Research. When more and more technologies are protected by IPRs in their target countries, producing international public goods may become more and more difficult (Fischer and Byerlee, 2002). Currently, these centers are venturing in license strategies on their protected technologies that provide a public good status for the purpose of poverty alleviation and food security in developing countries, while maintaining ownership in commercial markets both in developing and industrialized countries.

Costs of compliance. Compliance with the rights of IPRholders requires public and private research institutions alike to invest in capacities that they had not required in the past, notably legal and commercial specialists. There are already commercial seed companies that spend far more on legal services than on research. This preponderance of legal over research expense in fighting through the patent thicket may be a "warning" to public research institutions that emulating commercial plant breeding practices to produce public goods may be a less an optimal production pathway. Legal advice is not only needed to channel the use of research results in development oriented and commercial markets through contracts that need to be negotiated and concluded and court and settling disputes. Legal considerations are also more and more influencing the research itself. Scientists may be required to use old (free or cheap) technologies instead of effective ones which may be costly or not available.

      Scientists frequently feel stifled by the legal advisors who have to make sure that third party IP rights of contributive technologies are respected and that the IP produced by the scientists can be protected, by putting restrictions on scientific communication before a patent application is filed.