Governance can also be viewed at multiple levels; at the level of a research station, a national research system, a re­gional research network as well as at the global level. When we analyze the issues of governance of a research station, we take the external environment including the objectives given to the station as exogenous, and try to see how the governance of the station can be improved to meet these given objectives within the resource constraints. One can also analyze the larger question of governance at which one critically looks at whether the objectives defined by, or re­sources given to the station are appropriate and meet the criteria of good governance. Based on the conceptual frame­work as given here, one can develop a set of questions that are relevant for analyzing the governance of, and institu­tions involved in AKST investments (Table 8-19).

8.3.3 Analyzing the experience of governing AKST investments

8.3.3.1 Public funding/public sector research
The model of public sector research organization came to exist in many parts of the world during the second half of the nineteenth century. The founding of the public research organization was based on an assumption that nongovern­mental agencies (including private firms and farmers them­selves) are unable to mobilize adequate resources and skills required to generate agricultural research (see also 2.2.2). It was then assumed that farmers generally needed to be educated on the benefits of new technologies and did not have any major role in the generation of technology directly. Thus government, either national or regional, provided the resources for the establishment of these research establish­ments from the taxes, international aid or other assets such as state-owned land. This perception of the farmer, how­ever, has changed in recent years.
        Investment for AKST by governments has been success­ful on certain counts. It enhanced the capacity of a number

of countries to carry out good quality research. In many poor countries, there would not have been any significant level of agricultural R&D without these institutions due to the limited capacity as well as inadequate interest of the pri­vate or not-for-profit sectors to provide agricultural R&D, which mostly falls in the public good domain. Government-funding for AKST has also played an important role in en­hancing the awareness of farmers, in creating a wide pool of trained personnel and informing policy making at the national level in a number of countries.
         Despite such achievements, this model has had several problems. For example, it did not perform well in assess­ing the needs of farmers in many parts of the world and it has been fairly slow in responding to social and economic changes. There have been innumerable cases where research investments were directed in a way that they failed to meet set objectives, even if the uncertainty inherent in R&D ac­tivities is accounted for. Public organizations were not very successful in taking into account local agroclimatic and socioeconomic features in their research programs (Santha-kumar and Rajagopalan, 1995). Efficiency of public R&D organizations is also open to question, and one feature noted in many developing countries is the spending of a greater amount of financial resources to provide the salary of per­manently employed staff, with little left for actual research activities, which in turn affect the research output and hence the research efficiency (Eicher, 2001). This may not be di­rectly evident in ROR calculations of agricultural research. It is possible to have high ROR even with these levels of inefficiency. There have been inappropriate resource allo­cations between capital and operating expenditures in the public sector, resulting in a pool of inadequately trained and equipped personnel, research laboratories without sufficient operational and maintenance funds, or other inefficiencies.
          The fiscal problems of the governments of many devel­oping countries have led to a reduction of resources made available to public research systems, which often reduced