availability, accelerated global trade liberalization, concerns on food safety and demand for standardization of agricul­tural practices all make the production, marketing and trade of agricultural produce more complex (see chapter 3).

National  public  research   and   development   institutions within ESAP. Most national agricultural research systems (NARS) in ESAP were established in the 1960s. They are typically organized under a ministry, as an autonomous agency or as a coordinating council (Dar, 1995). Although they differ in operation, they are similar in policy and pro­gram formulation. Each has research agencies and stations dedicated to a specific commodity, and they are usually at­tached to the ministries of Agriculture, Natural Resources, Science and Technology or Higher Education. Most NARS are organized top-down and are government funded but have the autonomy to craft their own research programs.
     NARS are organized nationally, regionally and locally. The national research organizations conduct basic and ap­plied research on areas strategic to national interest and importance. Regional centers undertake applied research of regional significance and local research stations perform adaptability verification trials and fine-tuning of technology generated by the national or regional research centers. This system allows for work specialization and complementarity and provides for location-specific technology. Collaborative research is common among members, as well as with the private sector, civil society, Consultative Group on Interna­tional Agricultural Research (CGIAR) centers and interna­tional donors. Collaboration fosters task sharing to provide scientific solutions to common agricultural problems, ex­pands the sources for research and development investment and cultivates long-term partnerships and links.
     Private sector participation in research and develop­ment with AKST is quite limited and mostly complements, rather than substitutes, for continued public research. The bulk of private research and development has been in devel­oping new crop hybrids, animal breeds, chemical pest and disease controls, veterinary medicines, commercial livestock feeds, food storage, packaging and processing technology. The technology is often most suited to a small subset of the needs of small-scale farmers, is typically capital intensive and is covered by intellectual property rights.

National extension systems within ESAP. Every country in the region has a public department that provides agricul­tural extension services. Four models of extension systems prevail in most ESAP countries (Sulaiman and Hall, 2005) with approaches that are centralized, decentralized, NGO led or private sector led.

Centralized approach. Under this scheme, extension services are centrally planned, funded and implemented by units at­tached to the Ministry of Agriculture. Programs are mainly supply driven and use a top-down approach, with little par­ticipation from farmers or other stakeholders and with little or no accountability to clients. Technology dissemination is the primary objective. It is unclear if extension has been responsive to the drastically changing information and sup­port needs of farmers in recent decades (Sulaiman and Hall, 2002; van den Ban, 2005).

     This is true in India, China and a number of other Asian countries where extension policy is developed centrally in a fairly prescriptive fashion. Although approaches have evolved over the long term, it is not clear how lessons from their experiences are used in developing policy. In fact, de­velopment fads and encouragement from international do­nor agencies seems to be a major source of implementation. While these programs might be conceptually laudable, mak­ing them work on the ground is much harder. Furthermore, these major shifts often lock up extension in a particularly operational mode until yet another new idea comes along.
     China illustrates quite a different and interesting ap­proach to agricultural extension. The National Agricultural Extension Center under the Ministry of Agriculture formu­lates national extension policy. The center draws up exten­sion strategies that link agricultural programs with other agencies and provides training and supervision over pro­vincial agents. With the country's move toward a market-oriented economic system, rural extension services have ex­panded and diversified according to local resource and mar­ket development needs (Yonggong, 1998). Arrangements have been restructured to help farmers relate to new market opportunities more effectively. An incentives structure has been developed to allow profit sharing between extension workers and farmers. The policy, while insufficient to pro­vide specific courses of action, allows extension agents and farmers to pursue local, pragmatic program innovations. This has been important in responding to the rapidly chang­ing socioeconomic environment.

Decentralized approach. In response to demand for decen­tralized governance, this approach promised to improve farmer control and make extension services more demand driven. However, the lack of sufficient preparation by ex­tension management and the institutional inertia in most government bureaucracies has failed to deliver on these promises. Despite this, widespread clamor for decentraliza­tion suggests implementation problems might eventually be overcome.
     The cases of Indonesia and the Philippines highlight the complications of making broad policy prescriptions. The foreseen benefits of decentralization, primarily the devolu­tion of authority and decisions locally, have not yet been fully realized. The effectiveness of this approach depends on the skills and vision of local government officials. This suggests that policy instruments such as decentralization need to be accompanied with capacity development. Also, local stakeholders need to understand the importance and rationale for strengthening local knowledge networks. Since the performance of extension is dependent on these systems, stakeholders need to have the skills to analyze them, diag­nose system failure and design remedial measures. Capacity development is not only necessary to successfully implement decentralized approaches; it is indispensable if local stake­holders are to be more active in the policy process.
     For the Philippines, inadequate funding curtailed the ef­fectiveness of devolved extension. Experience suggests that with decentralization came a trade-off between the effective­ness of technology transfer, which seems to have suffered, and the accountability of the system to its clients, which seems to have improved. This has led to the emergence of