by seed multiplication on station. Similar achievements have been recorded in southwest China for maize (Vernooy and Song, 2004).

Local research and innovation: the contribution of occupational education. Local level innovation can be promoted if appropriate investments are made in educating farmers but this has been a relatively neglected area. One of the major breakthroughs has been the development and spread of farmer field schools (FFS) (Braun et al., 2006). Based on adult education principles, the schools take groups of farmers through field-based facilitated learning curricula organized in cycles of observation, experimentation, measurement, analysis, peer review and informed decision-making. FFS are making in aggregate a significant and influential contribution to sustainable and more equitable small farm modernization, particularly in the rain fed areas where twothirds of the world's poor farm households live. Kenya, Tanzania and Uganda have included the approach in national research and extension strategies, as has India. Systematic review of available impact data (Braun et al., 2006; van den Berg and Jiggins, 2007) and area-based impact studies (Braun et al., 2006; Pontius et al., 2002; Bunyatta et al., 2005; Mancini, 2006) demonstrate positive to strongly positive achievements. Contributing effectively to farmer empowerment also contributes to the strengthening of civil society and self-directed development (Mancini et al., 2007). Others have criticized their cost in relation to the scale of impact (Quizon et al., 2000; Feder et al., 2004ab), noted the weak diffusion of specific technologies, lack of diffusion of informed understanding (Rola et al., 2002) and failure in some instances to develop enduring farmer organizations (Bingen, 2003; Tripp et al., 2005). Further experimentation is warranted to test if combining farmer education such as FFS with complementary extension efforts will overcome the perceived shortcomings (Van Mele and Salahuddin, 2005).

     World Learning for International Development, the Alaska Rural Systemic Initiative project and the Global Fund for Children similarly have documented gains (World Bank, 2005a) in the effectiveness and efficiency of local research, school-based science education and the development of agroecological literacy at the grass roots brought about by investing in farmers' occupational education (Coutts et al., 2005).

Farmer-funded R&D and extension. Innumerable examples exist of effective technological advances pioneered by farmers themselves; e.g., grafting against pests, biological control agents such as the golden ant in citrus in Bhutan (Van Schoubroeck, 1999) and soil management and farming system development in the Adja Plateau, Benin (Brouwers, 1993).Yet the economic value of local and traditional innovations has not been much researched. One study in Nigeria in the early 1990s estimated the contribution of the informal agricultural sector where farmers are using mostly indigenous innovations at about US$12 billion per year, providing income for an estimated 81 million people (ECA, 1992). This estimate, however, does not include the cost of opportunities foregone or traditional practices that do not work. Recent literature begins to sketch out the strengths

and weaknesses that might be taken into account if a more comprehensive cost-benefit analysis were to be attempted (Almekinders and Louwaars, 1999).

2.2.2 Producers of AKST at national level

Countries have developed a complex array of public institutions, IAs and actors responsible for planning, funding, implementing, assessing, and disseminating public interest agricultural research. They include national, regional/municipal agricultural research institutions, universities and other higher education institutes and extension services. Most of these arrangements historically have been publicly financed because agricultural research investments involve externalities, and are subject to long gestation periods (cfr. Chapter 8, Table 8.1; Lele and Goldsmith, 1989; Beintema and Stads, 2006; Pardey et al., 2006ab).

     In the 1960s and 70s National Agricultural Research Systems (NARS) in developing countries (see 2.1), especially agricultural research institutes (ARIs), received strong financial support from governments and international donors to launch agricultural modernization through the dissemination of Green Revolution technologies (Chema et al., 2003). In the 1980s, as a result of budgetary crises and adjustment programs, public funds for agricultural research failed to keep up with expanding demand. Public expenditure declined as proportion of total research and development spending; expenditure per researcher declined much more because staffing continued to expand faster than budgets. From the 1980s onwards, the main drivers of institutional development of the NARS were structural reforms in national economies and adjustment policies, global political changes; ideological demands for reduced public sector involvement and intervention; a greater private sector role and significant biotechnological breakthroughs (Byerlee and Alex, 1998; Iowa State Univ., 2007). These events have given rise to a diverse institutional landscape responding to both domestic and global priorities and opportunities. Brazil's EMBRAPA, for instance, has become an exporter of capacity, in 2007 opening liaison offices in West and East Africa whereas NARS in sub-Saharan Africa continue to face many constraints (Jones, 2004).

Sub-Saharan Africa's National Research Systems. Overall budget constraints throughout the period have weakened public sector NARS in most African states. The general panorama today is of deep attrition of human resources, equipment facilities, capital funding and revenue, despite islands of promise such as the revitalization of capacity in Uganda under vigorous decentralization policies, in Ghana in relation to agroindustrial developments and in post-apartheid South Africa. Nongovernmental organizations, the CGIAR, private commercial actors and recently the establishment of sub-regional bodies (Central African Council for Agricultural Research and Development, CORAF), (Association for Strengthening of Agricultural Research in Eastern and Central Africa, ASARECA) and similar arrangements for southern Africa supported by the Forum for Agricultural Research in Africa (FARA), have filled the gaps only in part. An alliance largely funded by a US-based philanthropic trust recently has been established to transfer germplasm and advanced biotechnology skills to African NARS to catalyze