1999), care is needed to ensure that institutional changes in public sector and changing sources of funding do not under­mine the research agenda of public institutions, especially the generation of knowledge, which may not seem to be profitable and viable by the private firms.

8.4 Investment options
The goal of this international assessment of AKST is to provide policy makers with investment options for meet­ing the development and sustainability goals. Since no single investment can meet all goals at once, a portfolio of AKST investments are needed. Countries are likely to have differ­ent weights on the importance of the different objectives and so alternative combinations of AKST investments will be presented based on whether countries place more weight on environmental goals, improving health and nutrition, reducing poverty  and hunger,  or maximizing economic growth.
       This subchapter focuses on the research investment options of governments, international organizations, and foundations that support AKST in order to achieve devel­opment and sustainability goals. The questions that these organizations would like to be answered include:
•   How much should governments invest in AKST versus other public goods?
•   How should AKST resources be allocated? Which com­modities? Where—for example, less favored land, small poor countries? What type of technology—for exam­ple, labor using, land saving, or water saving technolo­gies? Which disciplines? Which components of AKST? Which institutions?
•   What methods should be used to decide how much money to invest and how to set AKST priorities?
The answers to the first two questions need to incorporate multiple criteria, which should include at least public RORs to research as well as the impacts on poverty, human health, and environment  (see 8.2).  Societies and policy makers who place more emphasis on poverty reduction rather than economically  sustainable  development  or  environmental sustainability could place more weight on the AKST invest­ments that reduce poverty than societies that favor improv­ing the environment. Societies with more poor people may place more weight on research to improve the livelihood of the poor than on research to reduce greenhouse gases Coun­tries in which agribusiness plays a big role in the economy and a large role in governance of the public research insti­tutes, may invest more in developing productivity-increas­ing change.         Formal priority-setting methods, including those based on ROR studies, are in practice only occasionally used to set research priorities, and formal multi-criteria techniques for research resource allocation are used even less (Alston et al., 1995). This is because they are expensive, time consuming, and some factors are difficult, if not impossible, to quantify. The impacts of agricultural research on environment, health and poverty have been particularly difficult to measure (see 8.2.5, 8.2.6, 8.2.8). As a result, most of the studies that we were able to assess and base our policy options on are those of the ROR type. However, making mistakes when invest-

ing in AKST can also be a problem. Investing money in an AKST project that has little social or economic importance, large negative consequences, or very little chance of succeed­ing can be even more expensive than formal priority setting. Thus investing in formal priority setting can save money and have high payoffs. Changes in governance that incorporate users of this technology into the priority setting and evalua­tion processes can also be productive.         When looking forward—particularly 50 years forward— people who decide on AKST investments often simply have to look for major problems that appear to be coming and invest to fill gaps in knowledge.

8.4.1 Criteria and methods for guiding AKST investments
"Any research resource allocation system, regardless of how intuitive or how formal in its methodology, cannot avoid making judgments on two major questions. What are the possibilities of advancing knowledge or technology if re­sources are allocated to a particular commodity, problem or discipline? What will be the value to society of the new knowledge or the new technology if the research effort is successful?" (Ruttan, 1982).

ROR studies and broader comprehensive impact assess­ments can be undertaken before initiating any AKST invest­ment (ex-ante) or after completion of the R&D activities (ex-post) depending on the purpose. The purpose of under­taking ex-ante assessments is to study the likely economic impact of the proposed investment, to formulate research priorities by examining the relative benefits of the differ­ent AKST investments, to identify the optimal portfolio of investments and to provide a framework for gathering information to carry out an effective and efficient ex-post assessment. Thus the greatest benefit of ex-ante assessment is derived from its power to assist decision makers to make informed decisions on investments i.e., in setting priorities to allocate the scarce resources.        AKST investment priorities are set at both micro and macro levels. More formal quantitative methods are used at the macro level and participatory methods are increas­ingly being used at the micro level. Priority setting is carried out explicitly or implicitly in all AKST investments through allocation of research resources to different commodities, regions, disciplines problems and type of technology. Since priority setting occurs at various levels of decision making, the resource allocation questions and methods employed vary depending on the level at which priorities are set. Pri­ority setting also requires intensive consultation among and between politicians, administrators, planners, researchers as well as the beneficiaries. Formal procedures facilitate this process as they systematize the consideration of key vari­ables and multiple objectives in the analysis and allow an interactive process to develop.          Priority setting based on ex-ante assessment employs a range of methods that can be broadly classified into supply-and demand-oriented approaches; although some combina­tion of these approaches is often used in empirical studies. Supply-oriented approaches to priority setting and resource allocation often are conducted at the more aggregative re­gional and national level and use a variety of methods from