1. The report reviews the World Bank's experience in supporting the development of national agricultural research systems (NARSs) in developing countries in the 1980s and early 1990s, and makes recommendations concerning future Bank support for this agricultural subsector. It covers 32 countries that had one or more free-standing research projects, as well as eight additional countries that had important research components in other agricultural development projects. All of the Bank's operational regions are well represented.
2. The review concludes that Bank interventions in the subsector have had a significant positive impact, but serious deficiencies persist in most of the NARSs supported by the Bank. In many cases, the Bank could have done more to address these deficiencies, and certainly can do more in the future. In particular, too little emphasis has been given to system sustainability issues, and institutionalization of indigenous capacity in research planning, priority setting and evaluation has made slow progress.
Bank Intervention and Project Performance
3. The broad objectives of rural sector development are growth, poverty alleviation and improved management of the natural-resource base. Agricultural research is a key element in achieving these broad sectoral objectives. When successful it increases productivity, which is central to economic growth in the agricultural sector and typically induces strong effects in other sectors. The majority of economic impact studies on research programs show highly positive returns. Although most economic assessments have been done on programs in NARSs in developed countries, the results obtained in developing countries have also generally been positive. These considerations-importance in the development process, and favorable prospects for positive returns on invested funds-have encouraged the Bank's participation in NARS development.
4. Bank support for agricultural research increased in the review period and reached an annual commitment of about US$200 million. A commitment to provide long-term support to NARSs is evident. All regions have received substantial support. Formal cofinancing of projects with multilateral and bilateral agencies has not been significant overall, but in most cases support by other financiers/donors (often substantial) has been available. The Bank has been a major financial contributor to international research through the Consultative Group on International Agricultural Research (CGIAR).
5 . Although the Bank's contribution to research funding has been useful and substantial, there is no room for complacency. The proportion of Bank lending allocated to the agricultural sector is declining. The agricultural sector, and the importance of technological advancement to enable improved agricultural productivity, need to be integral elements in the Bank's dialogue with its borrowers.
6. In general, support for projects was within the framework of the 1980 Sector Policy Paper on Agricultural Research, which set the scene for involvement of the Bank in the 1980s. The majority of the guidelines were valid. The Policy Paper, however, failed to adequately balance expanding capacity with concern for efficiency and sustainability. Despite its emphasis on the importance of research relevance, the Paper also neglected to give sufficient attention to issues such as formal farmer involvement in the research process and to gender considerations. Unfortunately, one of the Paper's recommendations that was not properly developed by the Bank and supported by NARSs was the establishment of effective research monitoring and evaluation capacities. This has imposed limitations on the evaluation of NARSs' effectiveness and of the quality of the Bank's agricultural research project portfolio.
7. The literature reveals that empirical studies on research impact have mostly been on program-wide or commodity aggregates, and have generally indicated very favorable economic benefits. There is no reason to suspect that the programs supported by the Bank would not have acceptable economic returns, even though research in developing-country NARSs is likely to be less profitable on average and more variable in returns than in OECD countries. The expectancy that satisfactory economic benefits would be obtained, and the subjectiveness of ex-ante estimation of benefits that might be linked to the investments (which were usually of an institution-building nature), led to the Bank excepting research projects from its normal requirement for an economic rate of return estimate at project appraisal. Nevertheless, the need to make best use of scarce research resources (which should be a primary objective of both the Bank and the borrowing client) dictates that an economic evaluation capacity be developed for both ex-ante and ex-post analysis in NARSs. The evaluation process is not simple, but techniques appropriate to typical data, skill and resource constraints that may exist in a NARS are available. The report elaborates on this issue, but the fact is that impact studies were extremely rare among the supported NARSs in the review period.
8. The virtual absence of quantitative data has meant that ex-post evaluations of Bank supported agricultural research projects (and the results entered into the OED database) have been based on non-quantitative performance measures. The same data constraints have handicapped this review. Ratings are derived from aggregated assessments of achievements in aspects of the research system that are commonly accepted as being necessary for an effective, efficient and sustainable research organization-human resource development; adequacy of facilities; effective organization, management, planning and review procedures; linkages among the agencies of the NARS and with external research entities; linkages with clients and the production of relevant technology; adequate funding; and an appropriate incentive environment for researchers. This approach relies on the assumption that, if institutional arrangements and research processes are efficient in meeting the agricultural sector's technological needs, the investment will give favorable economic returns. The components of research projects and most of the guidelines of the Policy Paper were framed in these terms, even though their ultimate objective was to enhance technology development and improve agricultural productivity.
9. Of the completed and evaluated free-standing projects in the research portfolio, 72 percent are rated as having a "satisfactory" outcome, which is better than the average performance of agricultural sector projects in the review period. Taking into account the key role that technology development must have in improving agricultural productivity and alleviating both rural and urban poverty, the Bank should consider this level of portfolio performance as unacceptable. This concern is exacerbated I)y ex-post evaluations indicating that the sustainability of research investments was uncertain in almost one-half of those projects having a satisfactory outcome.
10. Even within the set of projects with satisfactory outcomes, there is considerable variation in the extent to which different aspects of institutional development have been attained. The net result has been an improved human resource base (albeit with mismatches between available and needed skills); a substantially expanded research infrastructure in facilities and equipment, combined with doubts about the appropriateness of some investments; improved links with external research entities; advances in coordination of agencies within NARSs, but inadequate attention to involvement of academic institutions; mixed results in improving research extension-farmer linkages; weak development of the incentive structure for researchers; and, despite considerable emphasis in the second half of the review period, slow progress in improving the efficiency of resource allocation in NARS agencies.
11. The performance record suggests that the Bank has been better at expanding the research capacity of particular agencies within the NARSs than at improving management and overall system efficiency. This is of concern, given the inadequacy of national funding support for public research institutions and serious doubts about the sustainability of many of them evinced by this review. Although expansion and efficient resource use should not be incompatible, expansion has proven to be the more likely outcome, and this may have even contributed to the weak results in improving efficiency.
12. Ultimately, the NARSs themselves and the governments to which they report must take responsibility for enhanced technology development and the transfer of knowledge to farming communities. However, the Bank, as the largest single source of external finance to agricultural research in developing countries, has a key role to play in helping to:
- enhance the prospects for sustainability of NARS investments;
- give priority to improvement in management of NARSs and their agencies so that most cost-effective use is made of resources and the quality of research output is improved; and
- ensure that effective and sustainable arrangements are in place to introduce appropriate technologies to the targeted farming communities.
The major findings of this review, assembled in chapter 5, can be summarized under these three broad headings.
Sustainable Investment and Research Funding
13. Inadequate national funding to operate the public agricultural research establishment is a critical constraint in most developing-country NARSs. The importance of the agricultural sector as reflected in its contribution to GDP, employment, and export income certainly varies across regions and countries. However, in all countries where the Bank is helping NARS, the sectoral contribution is important enough to warrant strong budgetary support of an efficient research system that produces mostly public goods. In the review period government support has been evident in a willingness to expand the research establishment, albeit in most cases with substantial loan or grant assistance. Unfortunately, while the staffing of public-sector components of NARSs has increased, the financing of research operations has not kept pace with staff expansion. Consequently, the funding per researcher has declined in most countries, with salaries consuming an unhealthy share of recurrent funding, so that efficiency and effectiveness have suffered and institutional sustainability has become doubtful. The Bank and donor community have strongly supported expansion and even taken exceptional measures to support operational funding, but the reticence by governments to provide funds to properly utilize research assets brings into question the level of their commitment to high-quality research; the frequently occurring, untimely release of approved budgets also demonstrates a lack of appreciation of either the damaging consequences to research productivity of delays in program execution, or the potential benefits of efficient agricultural research.
14. On average, research expenditure as a proportion of agricultural GDP (agricultural research intensity), typically at a level of about 0.5 percent did not increase in any region except AFR, where the trend is directly linked to the exceptional level of support by the Bank and donor community rather than to allocation of governments' internal resources, and sustainability is endangered by a critical recurrent-cost funding problem.
15. Governments seemingly appreciate a need for publicly funded agricultural research but, given resource scarcities, committment is not such as to give it priority over competing uses for recurrent outlays. The Bank, through its economic dialogue with borrowers, has the opportunity to highlight the benefits of investment in efficient research and recommend appropriate funding levels. The demand for public funds can be minimized by (a) transferring the cost of research to beneficiaries (through private research or cost recovery), (b) rationalizing public research investment and ensuring that the potential of all internal and external research resources is realized, and (c) ensuring that the management of public agencies is such as to maximize the cost-effective production of technology. The supply of funds, in the long term, must depend on governments being convinced that the operation of the public research system is efficient and that its social benefits are high enough to justify a high priority for funds.
16. Although individuals benefit by adopting new technology, rarely can technological knowledge be appropriated by an individual. In addition, use of the knowledge does not diminish the extent to which it is available to others. These factors support the public-good rationale for public investment in agricultural research. When this argument is combined with the fact that high proportions of farmers in many developing countries operate at subsistence or near subsistence levels (and, therefore, are less likely to be able or willing to pay for technology development), continuation of public sector investment in research is justified. Nevertheless, the possibility of transferring part of research costs to direct beneficiaries did not receive adequate attention in the Bank projects reviewed. Such transfer has occurred to a significant degree, however, in some NARSs, particularly in LAC, and frequently in the case of export commodity crops in all regions, the latter usually as a carry-over from a colonial period. Furthermore, the Bank has rarely used its public expenditure reviews to enhance allocations to research or to promote strategies that enable cost recovery.
17. The Bank has tended to assist expansion of assisted agencies rather than their rationalization, although some recent projects have attempted to rationalize institutions and reduce staffing levels. The Bank, in its interactions with governments and with the donor community, could also have given more attention to the totality of NARSs in defining appropriate intervention areas. The research potential in agricultural faculties at academic institutions is poorly utilized in most NARSs. Closer collaboration between the Bank's agricultural sector and human resource divisions would have led to more comprehensive treatment of the issue.
18. Scarce funding suggests advantages in collaboration amongst countries in producing technology for shared agroecological regions. The Bank and donors have made a commendable effort in AFR in establishing a Special Program for African Agricultural Research (SPAAR), but collaborative opportunities elsewhere have not yet been fully exploited. Regional cooperation aims to rationalize research and to initiate cooperation among donors and NARSs in which the comparative advantages and strengths of individual NARSs are utilized. Four regional frameworks for action have been initiated under SPAAR. This concept is consistent with the proposition that the small size of agricultural sectors and the scarcity of funding and experienced staff in many AFR countries preclude the independent development of a complete coverage of all technology needs in each. Nevertheless, there is still some reticence among donors to adjust their support (in accordance with defined priorities) away from research areas they have traditionally sponsored. Also, the willingness of national governments to fund research outside their own NARSs is not as strong as it might be.
19. Direct Bank/donor support for operational costs (para. 13), which has even reached 100 percent of some programs in recent AFR projects, reflects the acceptance of a critical need for technology development. However, there is a real danger of developing excessive dependency on external support. Fiscal sustainability must be a priority consideration; otherwise, substantial investments will not be efficiently maintained or utilized over the long run. High levels of contribution by the Bank and donors should not be undertaken without careful analysis of the government's fiscal support for the sector and for research within the sectoral allocation. Commitment should be demonstrated. The agreed research strategy should clearly show a willingness to adopt prioritization processes that limit programs to available resources and maximize efficiency, and a willingness to enter into and support legitimate regional research initiatives. Ideally, the policy framework should be conducive to agricultural investment and profitable use of technology.
20. Collaboration of NARSs with the CGIAR Centers has had mutual benefits for the continued development of relevant technology for farming communities. However, it is only in recent years that Bank project documents have been forthright in specifying participation of CG Centers in actions to improve NARSs; this needs to be expanded.
21. In the investment packages financed by the Bank, apart from some instances of inadequate attention to rationalizing agency investment in SAS and AFR, the type and scale of infrastructure investments proved to be largely justified. Scientific and field equipment also contributed to research capacity, but there were frequent problems of poor definition of equipment needs so that items were procured that were less than appropriate for the programs or for the institutional maintenance capacity. This was associated with insufficient attention at preparation/appraisal or to a sequencing problem in which equipment was ordered before the research programs were clearly defined.
22. Bank projects have strongly supported library facilities in the NARSs, but the continued usefulness of these facilities is in jeopardy owing to constrained funding for journal procurement in many systems. Recently, there has been some (but insufficient) attention to electronic communication systems and networking facilities. In addition to more effective use of scientific literature by researchers, the process of producing and reviewing scientific papers based on research findings in the NARSs. is another area which needs improvement.
23. Research as a component in agricultural projects can be a viable and sustainable investment if it is of sufficient size to warrant attention, if it is linked with a well organized existing public or semi-public research institution, and if it is complementary to the program of such an institution. However, many interventions in Bank projects did not have enough of these characteristics to generate useful results. Sectoral adjustment loans in ECA/MNA region had some positive impact on the reform of research institutions, but these were part of ongoing dialogues on research in the agricultural sector rather than isolated interventions. Elsewhere, research components in this type of adjustment lending were of little consequence.
Research Management and Efficiency
24. The most common research framework in the public research institutions supported by the Bank has been based on commodities with supporting scientific disciplinary units. Bank projects have commonly not attempted to modify the basic structure of these institutions, but have often introduced an agroecological dimension and promoted the creation or strengthening of an apex coordinating institution for the NARSs. Where the latter has not been feasible, improved coordination has usually been supported within the dominant group of institutes. Where they have been implemented, improved coordination mechanisms have facilitated the Bank's efforts to provide linkages to allow efficient use of resources, develop national research strategies, and institute program planning and review procedures, even though the outcome of these measures has not always been as favorable as anticipated.
25. With some notable exceptions, agricultural research by universities was not a major factor, as science and agricultural faculties were usually constrained by funding and facilities. Some success has been achieved, however, in involving academic (and private) research institutions in components of a research strategy through the creation of special funds under Bank projects, which can be tapped by these institutions.
26. Improved research management and planning efforts in Bank projects have almost always involved technical assistance provided under the loan or by donors. The International Service for National Agricultural Research (ISNAR) has been particularly active in this field. However, the overall outcome of this support can only be described as a "significant start" to a more rational planning process in most NARSs, or in major agencies within NARSs. At the broad strategy level, there has been considerable progress, although performance has been much weaker in analyzing resource availability and in establishing priorities amongst programs to implement strategies.
27. Weaknesses in the prioritization process have often resulted in documents that indicate what should be done considering the projected demand for technological improvement, but that do not face up to the task of setting priorities in the face of resource constraints. Frequent deficiencies of research master plans have also included a lack of real integration of planning between federal and state-level entities and among institutes, leading to unnecessary wasteful overlap, and a failure to properly incorporate the findings of farming system diagnostic surveys in the planning process. Even when there is a ranking or prioritization of programs based on some form of scoring model, use of the model, and often the model itself, frequently leave much to be desired. In addition, those responsible for policy and investment decisions are sometimes reluctant to make the hard choices raised by priority-setting exercises, so that little change occurs in the existing research agenda where failure to incorporate derived program modifications has occurred; this has likely been partly associated with the common practice of heavy reliance on external technical assistance in preparing master plans, and the "one-off' nature of many exercises where they are perceived as necessities to obtain loan or grant support. In essence, experience indicates that research master planning is unlikely to be very beneficial unless it is "owned" by funding and implementation agencies, is comprehensive in its coverage of potential contributors, adopts robust prioritization procedures in the face of resource constraints, and is able to link macroeconomic considerations with the real needs of farming communities.
28. In some NARSs, especially in AFR, there has also been a tendency to accept governments' stated sectoral policies without attempting to influence the priority-setting process when policies seem inappropriate. This, and the commonly occurring "softness" in the overall prioritization process, can be partly linked to a weak capacity in policy and economic analysis in most NARSs.
29. In developing programs to make best use of available resources, it is essential that primary consideration is given to the needs of research users, implying their involvement, or at least recognition of their requirements, in the planning process (as discussed below). In meeting these needs, however, there appears to be no substitute for the use of economic analysis, preferably as an economic surplus model (i.e., a representation of changed consumer and producer surpluses induced by adoption of research findings) in its comprehensive or simplified forms, to assist in priority setting among programs and subprograms. Subjective judgments will inevitably be involved, but organizing information in a manner consistent with such a model increases the probability of providing cogent assessments. A by-product of the approach can be a set of estimates of internal rate of return, benefit/cost ratio and net present value, which can be used to argue for budgetary allocations.
30. Simplified scoring methods can be used to compare programs where resources are not available for more complete analysis, provided appropriate weights are given to assessment criteria. However, at least rough economic efficiency indices must be calculated for these methods to be worthwhile. The tradeoff is in their qualitative nature in which no concrete opportunity costs are interpreted for alternatives. Informal methods that make structured judgments based on an analysis of the principal determinants of the net present value of research are likely to be more reliable than poor execution of simplified scoring methods, and other nonscoring methods such as congruence tests.
31. For between-project priority assessment, which has often to be done within institute programs, the full-blown quantitative economic measurement of likely impact is normally not warranted. Technical review in committees of peers, combined with an awareness and use of the principal determinants of the net present value of research, can go a long way to attaining efficiency in the use of resources available to individual research managers. The effectiveness of the peer review process should not, however, be taken for granted. Full expression of this critical process has been especially difficult where public-service and seniority-driven traditions tend to discourage the free exchange of cogent scientific criticism, particularly if addressed to the most established scientists.
32. The concept of independent program review using experts from other local institutions or international institutions has been successfully developed in a number of projects. In addition to enhancing local programming ability, this intensive review complements the postgraduate training and technical assistance programs that the Bank has funded to improve the quality of scientific research. However, more attention needs to be given to the institutionalization of needs-based training processes. The development of long-term training and research relationships with reputable education/research institutions in developed countries can also be profitably expanded to enhance scientific research capacity. The CG Centers also have a major role to play in both program review and hands-on training.
33. Many projects have stipulated that annual research plans be submitted to the Bank for review. This has not been a satisfactory method of ensuring appropriate research programming. The Bank has not had adequate resources for this to be done effectively. Recommendations are made to establish within the Bank a core of specialized research staff who, in addition to operational duties, will support less specialized research project task managers. However, considerable reliance will still have to be placed on on-site specialists to provide the required expert guidance in key "research quality" areas such as priority setting, participatory planning and evaluation, research rigor, and economic assessment of program impact. The effectiveness of such technical assistance, however, should not be measured in the production of a specific product such as a plan or impact analysis; rather, effectiveness is represented in the extent to which the required skills are developed in national staff, and appropriate procedures are instituted in research agencies.
34. As already mentioned (para. 8), the near absence of ex-post evaluation capacity needs to be remedied. Economic expertise, however, is likely to be a scarce commodity in most NARSs in the foreseeable future. Therefore, it will be essential that this scarce resource is used to best advantage in producing relevant research output. This precludes ex-post evaluation of all programs on a routine basis. Nevertheless, analysis of the economic benefits of selected programs or parts of programs should be undertaken to assist in future research planning, and to justify expenditures to public or private funding entities. Econometric methods are rarely feasible as routine measures in most developing countries because of their data requirements. However, economic surplus or cost reduction/profitability approaches can be used as for ex-ante estimations (para. 29), relying with appropriate cautions on experimental data and the opinions of scientists and extension workers to estimate unit cost changes (or yield improvements) and adoption rates for key technologies that have been developed over the relevant time period. The robustness of any analysis is enhanced by the extent to which the actual adoption of technology by farmers is measured, but this aspect of research management (monitoring of impact) has been sorely neglected to date.
35. Progress in evaluation will not be possible unless both Bank staff and research agencies cooperate in the design and implementation of practical methods of measuring performance and impact. As already stated, the Bank must put more resources into the ex-ante and ex-post economic evaluation of research. But this does not imply that the focus should be on estimating the economic return on a project supported by the Bank. Such a narrow interest would not add much to improving the efficiency of the technology development process. Rather the Bank should focus on the institutionalization of a capacity within the NARSs to use economic analysis in priority setting among program options. Even in the project preparation/appraisal phase, this principle should be applied. Favorable economic returns for major programs in the agencies being supported will provide economic justification for the Bank's support and a basis for subsequent evaluation. The availability of economic expertise in NARSs will determine whether the economic analysis at preparation/appraisal is no more than an introduction to its use, or an in-depth participatory exercise by social and biological scientists, which might cover virtually all programs and even form the basis of a master plan for specific institutes. In all cases, however, the Bank should pay close attention to this aspect of research management throughout supervision and project impact assessment.
36. A multi-donor financed unit (ESDAR) with the objective of promoting synergies between the different parts of the global agricultural research system was established in the World Bank in 1994. This unit has the potential to promote cooperative support by the international community for individual NARSs, as well as to bring pressure on NARS management teams to respond to the need for improved management and resource allocation. it could also have an important function in organizing the donor community to provide grant funding for technical assistance in key areas of research management (with emphasis on skill development), which many borrowers are loathe to contract with their own or borrowed funds.
37. The introduction of management information systems (MISs), which have the potential to facilitate research planning, monitoring, evaluation, and individual and unit accountability, has been promoted by the Bank in many of its projects in the latter part of the review period and by donors (especially USAID), often with the involvement of ISNAR. In general, where there has been some adoption of an MIS, it has been limited and has not had the full support of institute management and staff. A recent review by ISNAR of MIS activities highlighted problems and pointed to promotional/training procedures that should enhance effective adoption of MISS. However, even straightforward accounting systems need major improvement in many NARSs.
38. Improved facilities, postgraduate training opportunities and improved planning and review procedures provide not only a capacity for more relevant research but also an incentive for staff to stay in the public research organizations. However, they are unlikely to yield the intended dividends unless there is a suitable incentive framework for researchers to perform well. The scarcity of operating funds is an overriding disincentive to performance. There has been some limited successes in obtaining more attractive salary levels for researchers under autonomous institute arrangements, although the latter has generally not been sufficient to guarantee such consideration. Nor has the adoption of a semi-autonomous status normally been adequate to introduce reward and promotion systems based on performance, which in reality may only be achievable under completely autonomous arrangements-a rare occurrence.
Use of Relevant Technology
39. The best way to ensure that developed or adapted technology is quickly adopted by the targeted farming community is to ensure that it has been derived to address a clearly identified problem. The need for more relevant research that addresses the needs of smallholder farmers, especially in the more difficult production environments, has been highlighted in the majority of Bank projects. Interventions have included adjustments in resource allocations to correct imbalances, and a range of techniques to improve research-extension-farmer linkages. The latter have been designed to inform researchers of the real constraints facing farmers, and to interpret new technology as practical recommendations for use by extension services and adoption by farmers. Organization of regular meetings between research and extension personnel to achieve these objectives has had mixed results. On-farm research capacity has been substantially expanded, although it has suffered from a number of problems, the most important of which is that it is often the first program to suffer in times of funding constraints.
40. The adoption of a farming system perspective has been widely promoted, along with the use of socioeconomic considerations in planning and review processes. This perspective is necessary not only to make research more responsive to smallholders' needs, but also to ensure that the traditional commodity research emphasis adjusts to the technology needs to more diversified land use demanded by changing markets. Progress has been made, but there is still some reticence amongst established researchers to accept the need for this demand-driven emphasis, and the concept will need continued nurturing to become an established part of the research process. Assisting other researchers to attain an understanding of farming systems and their requirements arguably has the highest priority for use of scarce socioeconomic expertise in a NARS. The process is not complete, however, unless the adoption of technology resulting from research is measured in the targeted communities (para. 34). Greater attention to the farming system perspective in university training would be beneficial. Unfortunately, this has not been a focus in curriculum development in the Bank's education portfolio.
41. Women head a substantial proportion of farm households in many regions, and are key elements in most farming systems. The emphasis given to institutionalization of a farming system perspective implies consideration of gender in analyzing and responding to household economies, but this has not resulted in the required recognition of the gender issue. Gender considerations are important for effective farming system diagnosis, research planning and evaluation, and technology adoption studies in smallholder farming communities. The CG Centers have taken a lead in this area; this should be capitalized on in the Bank's dealings with borrowers.
42. The fullest expression of demand-driven research is to have the intended beneficiaries (farmers and other industry stakeholders) involved in the design and evaluation of experiments. The expansion of on-farm adaptive research has encouraged beneficiary involvement, although this has not always occurred in projects and often only in a limited way. A well-organized approach to beneficiary involvement has been adopted in the research project in Mali, and in a 1995 project in Colombia there is an attempt to get representatives of beneficiaries involved in research program review at the municipality level.
43. The Bank's concern for the transfer of developed technology is also evident in that onehalf of the free-standing research projects included a major extension component. Where this has not occurred, separate support for extension was usually provided in similarly dedicated projects or in area development projects. However, the end result of these technology transfer investments in adoption by farmers and in increased production and welfare, as for research investments, has not been measured. Nevertheless, a 1994 OED review of the Bank's extension portfolio and this review concur in finding that there is no single system that has sufficient merit to warrant its universal use in linking researchers, technology transfer agents and users of technology to ensure research relevance and quick adoption of improved technology. Local circumstances determine the most appropriate arrangements. It is clear that researchers must be made aware of the circumstances of farmers, whether this be through direct interaction with farming communities or their representatives, heavy dependence on intermediaries in public or private extension systems, or a combination of these approaches (which is most common). Regardless of methods used, this interaction has to be an integral part of the research process.
44. Recommendations of this review of the research portfolio can be summarized as follows. The Bank should:
Role of research
- elevate consideration of the agricultural sector, and particularly the contribution that agricultural research can make, in the policy dialogue, in economic and sector work, and in preparation of Country Assistance Strategies (CASS) with borrowers (with the objective of ensuring appropriate levels of budgetary support, timely release of budgetary allocations, and adoption of policies conducive to research output);
- within the agricultural sector, consider agricultural technology generation, acquisition and adoption as components in a single system, so that linkages between educators, scientists, extension agents (public and private), farmers and other industry stakeholders are taken into account when considering support for any component in the system;
- give more attention to the relationship between research and education sectors in client countries and in the Bank's servicing units, so that the education sector not only services training requirements, but also realizes its research potential;
- continue to support international agricultural research (CGIAR), establish alliances with CG Centers and other providers of assistance, and facilitate links between NARSs and cooperating research entities;
- be willing to substantially support lending for comprehensive development of research systems, but insist on demonstration of borrower commitment and be convinced that any supported investment has a high probability of sustained funding;
- as appropriate, rationalize existing resources instead of, or in addition to, expanding facilities;
- only finance high proportions of operating costs and specific research programs (on a temporary basis) when borrowers have demonstrated commitment to supporting an agreed, prioritized research agenda;
- ensure that policy and regulatory environments are favorable for private-sector research and technology acquisition;
- encourage client NARSs to introduce research cost-recovery systems, which will also enhance responsiveness of research to clients' needs;
- insist on regional cooperation in research by supported NARSs where this is advantageous and legitimate, and use Bank-supported projects to foster implementation of the frameworks for (research) action in Sub-Saharan Africa;
- be willing to take a leading role in donor coordination, where necessary, to encourage more efficient use of external resources;
- give greater attention to the phasing of support to NARSs for timely procurement of suitable equipment and facilities, and to engaging expert opinion in specifying equipment needs;
- where conditions are not appropriate for comprehensive support of a NARS, there may be cases where targeted support for specific research programs is justified;
Planning and evaluation
- re-double efforts to promote coordination mechanisms in NARSs as a vehicle for improved use of national institutional resources;
- continue to promote the use of research master plans and research prioritization procedures, but with greater attention to ownership of the procedures and outputs by national staff, to the quality and validity of the processes, and to their use in resource allocation; unless all of these elements are in place and policy makers are brought on board, the planning process may be of little value;
- give much greater emphasis to the development of monitoring, evaluation and socioeconomic analytic capability in NARS agencies-weakness in this area has limited the ability of agencies to undertake robust priority setting among programs, and has resulted in sparse ex-post evaluation of research work; in coincidence with more economic orientation by client agencies, put more effort into the development of ex-ante economic evaluation of programs and practical research performance indicators with clients prior to project initiation, define which research programs will be selected for measurement of economic impact, and ensure arrangements that take account of available skills are in place for such analysis; closely supervise implementation of monitoring and evaluation aspects of projects, and utilize program impact assessments in ex-post project evaluation;
- continue emphasis on improving the quality of scientific research, especially through needs-based training, technical assistance, more effective linkage with the CG Centers, both internal and independent, external review of research programs and projects, and local and international scientific networking;
- continue its insistence that research must be client-driven and responsive to defined client needs-usually well-facilitated by comprehensive stakeholder involvement in research planning and by adoption of a farming systems perspective, together with effective on-farin research capacities;
- emphasize the importance of client responsiveness by requiring, in the Bank project preparation phase, the formulation of critical research programs that meet diagnosed needs and opportunities of major farming systems; • provide guidance on the neglected areas of user involvement in research design and evaluation, and on implications for the planning and conduct of research of the prominent role women play in fanning systems;
- insist, in dialogue with a borrower's representatives and research agency management, that efficiency will not be attained unless adequate performance incentives are provided for staff, the most critical being sufficient operational support funding and appropriate criteria and opportunities for advancement in a research institution;
- accept that, to enhance the results of the research portfolio, the Bank will have to establish (through identification of staff and/or a change in staff skill mix) a core of staff with specialized research skills, who can maintain linkages with the international agricultural research community and assist, inform and train less specialized staff with task-manager responsibility for research projects, including demonstration of best practices in selected NARSs with different resource circumstances; even then, considerable reliance will have to be placed on long-term technical assistance within projects to guide NARSs in improving research processes.
45. Amongst these recommendations, priority items to be addressed by Bank management are:
- Country assistance strategies and programs should explicitly address the role of the agricultural sector and, within it, agricultural research.
- Strong support for international and regional research and NARS development should be continued. Comprehensive assistance to a NARS, however, should only be provided when there is a clear commitment from the borrower to adequately fund the system and to adopt sound research management principles. In the absence of favorable sectoral or research policies, limited research program assistance that is highly selective and addresses clearly defined research problems or gaps can still be appropriate. It should normally be combined with non-lending services designed to strengthen borrower commitment to research policy reform.
- External linkages by NARSs and strategic alliances between the Bank and appropriate development partners should be fostered (through lending and nonlending services) to enhance the effectiveness of national and international research programs.
- Economic analysis (normally using an economic surplus model or a simplified derivative) should be fostered in borrowers' processes for prioritizing research. By the same logic, economic analysis should be used by the Bank in ex-ante and ex-post evaluation of supported research programs.
- A core of Bank staff with specialized agricultural research system skills should be used to enhance the effectiveness of Bank assistance in NARS development, and research programs should be identified/developed in each region to serve as functional models of good practice.
- The efficient dissemination of research findings should be treated as an integral part of the technology development process.
- Research projects should foster procedures in NARS that enhance the relevance of research.
- M&E to facilitate planning and to underpin ex-post research evaluation should become a mandatory element of Bank-supported research projects.
- Measures conducive to enhancing scientific rigor should be required in all Bank supported projects-needs-based training, well directed technical assistance, external reviews of programs, competitive grant funding, scientific networking, and linkages with research entities of excellence.
As some of the recommendations represent a departure from current practices and others imply changes in emphases, the production of a strategic issues paper to guide managers and staff in the agricultural research subsector is recommended.
46. Undoubtedly, the Bank has made a significant contribution to NARS development. With closer attention to the issues raised in this review, future impact could be much greater. The importance of technology development dictates that the Bank must stay the distance in fostering agricultural research for the less-developed world, as well as continue to monitor counterpart developments in the more-developed world. International research can assist in the development of relevant technology, but the main task must rest with the NARSs themselves. If borrowers are willing to make the required commitment, there is a great opportunity to substantially improve the capacity of NARSs to enhance their research effectiveness, efficiency and relevance. There is scope for significantly improved performance in the agricultural research portfolio.