programs to teams of Filipino scientists (Pray, 1987). The domestic supply of skilled personnel is heavily dependent on the level and composition of public and private expenditures on education.
         Several aspects of the business environment affect the level and productivity of research costs. Industrial policy can influence the degree of market concentration, the intensity of competition, and the prices of research inputs and out­puts. Various government incentive programs, such as gov­ernment contracts for new products and processes, grants and concessional loans, technical information services, and tax incentives, reduce research costs. Indirectly, the devel­opment of capital markets makes it easier for firms to raise funds for research (for example, venture capital). Bilateral and multilateral agreements also improve trade opportuni­ties by facilitating access to intermediate technologies.
        Regulation such as product quality standards, qual­ity testing regulations  and seed certification procedures can greatly increase the costs of commercializing research output and they can delay the adoption to new technology which reduces the incentive to innovate and reduces the ben­efits to farmers. Regulations that have been put in place in many countries to ensure that products developed using bio­technology are environmentally benign and safe for human consumption are necessary to gain consumer acceptance, but they have greatly increased the cost of developing and releasing transgenic plant varieties. For example, one seed company spent US$1.6 to 1.8 million to obtain regulatory approval for Bt cotton in India. This is more than the an­nual research budgets of most Indian seed companies. As a result, only the largest companies can afford to attempt to commercialize genetically modified crops (Pray et al., 2005). Bangladeshi regulations that required irrigation pumps and diesel engines meet efficiency standards of wealthy coun­tries delayed the commercialization of inexpensive Chinese irrigation equipment and slowed the spread of high-yielding rice varieties by 5 to 10 years (Gisselquist et al., 2002)

8.1.3 Investments in other AKST components
Investment data for other AKTS components, such as edu­cation and mainstreaming traditional knowledge, are dif­ficult to obtain.
        Due to the public good attributes of extension services, it not surprising that the great majority of official extension workers worldwide are publicly-funded and most extension is delivered by civil servants. Universities, autonomous pub­lic organizations, and NGOs deliver perhaps 10% of exten­sion services, and the private sector may deliver another 5% (Anderson and Feder, 2003).
        The structure and function of national extension sys­tems continue to change, particularly as the level and source of funding, especially public funding, changes across differ­ent countries. In many countries, there is a continuing ef­fort to shift the cost of extension to farmers, although these different approaches to privatizing extension or to increase cost recovery by public extension systems have met with different levels of success (Anderson, 2007), private sector involvement remains small.
        Given the numbers of extension personnel and the likely costs incurred in the different country contexts, agricultural extension investment is of the same order of magnitude

(although likely lower) as the agricultural research world presented in expenditure terms (Table 8-1); so it is surpris­ing that it has been subject to relatively little critical data collection and analysis. In contrasting differences between developing and more industrialized countries, one feature is the even more extreme differentiation between public and private entities; however, the situation is not fully clear (World Bank, 2006; Anderson, 2007).

8.1.4 Funding agricultural R&D in developing countries
Although various new funding sources and mechanisms for agricultural research have emerged in recent decades (see 8.3), the government remains the principal source of fund­ing for many developing countries. For example, the prin­cipal agricultural research agencies in the largest countries (in terms of agricultural R&D investments) such as Brazil, China, India, Mexico, Nigeria, and South Africa are still mostly funded by the government. In contrast, the principal agencies in a number of countries have been able to diver­sify their sources of support through contract research (for example, Chile and Cote d'Ivoire) or a commodity tax on agricultural production or export (for example, Uruguay, Malaysia, Colombia) (ASTI, 2007).
          Bilateral  and  multilateral  funding has  been  an im­portant source for agricultural R&D for many countries. Since 1970, both multilateral and bilateral assistance grew in real terms, but began to decline after the early 1990s to only US$51.2 billion by 2001. In recent years, ODA has increased again (Table 8-7). After several decades of strong support, international funding for agriculture and agricul­tural research began to decline around the mid-1980s. This decrease is mostly related to the significant increase in the share of ODA spent on social infrastructure and services (FAO, 2005a). Data on the sectoral orientation of aid are available for bilateral funds only. The agricultural compo­nent of bilateral assistance grew steadily and accounted for 16% in 1985, declining thereafter to 4% in2003. Regionally the largest proportional reductions in assistance occurred in Asia. ODA to agriculture halved in SSA and decreased by 83% in South and Central Asia during the period 1980-2002 (FAO, 2005a).
           Data on aggregate trends of donor funding for agricul­ture and agricultural research are unavailable, but informa­tion on agricultural R&D grants and loans from the World Bank and the United States Agency for International Devel­opment (USAID) is accessible. The amount of funding that USAID directed toward agricultural research conducted by national agencies in less-industrialized countries declined by 75% in inflation-adjusted terms from the mid-1980s to 2004. Again, Asian countries experienced the largest losses, but funding to Africa and LAC was also cut severely (Pardey et al., 2006b). Over the past two decades, World Bank lend­ing to the rural sector has been erratic, but after adjusting for inflation, the general trend has been downward as well. The exception is the large amount of lending in 1998, which resulted mostly from loans with large research components approved for India, China, and Ethiopia (Pardey et al., 2006b).
            There appears to be no single cause for the decline to the donor support for agriculture between1980-2003, although