2.1. Guiding Principles for TAC's Analysis
2.2. Medium- and Long-Term Vision of the CGIAR
2.3. CGIAR Priorities
2.4. Institutional Capacity, Governance and Management Considerations
2.5. Improved Systemwide Efficiencies
In pursuing its analysis, TAC recalled the mission statement of the CGIAR, which reads:
"Through international research and related activities, and in partnership with national research systems, to contribute to sustainable improvements in the productivity of agriculture, forestry and fisheries in developing countries in ways that enhance nutrition and well-being, especially of low-income people." (TAC/CGIAR, 1992)
TAC's analysis was underpinned by a set of guiding principles. These principles are related both to the CGIAR mission, goals, priorities and strategies, and to institutional attributes and opportunities for greater cost-effectiveness on a Systemwide basis:
CGIAR Mission, Goals, Priorities and Strategies
· restructuring should be planned against the broad framework of the medium-and long-term vision of the CGIAR;· structural changes should be consistent with CGIAR priorities and strategies;
· to strengthen research links between the CGIAR System and national programmes, and other partners in the global agricultural, forestry and fisheries research system.
Institutional Attributes and Cost-Effectiveness
· whenever possible, change should build on what exists that is of quality and efficiently operated, and has proven competence and delivery capacity;· CGIAR Centres should be institutions of excellence, focusing on research of high priority and international importance;
· overlap in responsibilities should be avoided;
· the basis for programme integration would be to achieve synergy between the activities involved;
· CGIAR programmes and institutes should have critical mass allowing them to address their mandates effectively;
· cost of governance, management and administration should be a reasonable share of overall expenditures in order to maximize resources available for research.
In 1990, TAC outlined a medium- and long-term vision for the evolution of the CGIAR System. The long term was defined in terms of the period when most national research systems in developing countries will be strong enough to meet their own national research needs. Realization of a long-term vision will therefore depend heavily on the development of capacity of national research systems and of effective regional and transnational mechanisms of cooperation, and for the private sector to become an important alternative supplier of research. Consequently, the CGIAR System will be expected to be smaller than, and different from, what it is at present. However, it is TAC's judgement, based on considerations of international public goods, economies of scale, and spillovers, that there will be a continuing need for international efforts in the long term in (TAC/CGIAR, 1990):
· germplasm collection, conservation, characterization, evaluation and enhancement, and basic genetic manipulation of plants and animals of transnational and/or global significance;· strategic research on global issues of natural resources conservation and management;
· strategic research on public policy and public management issues of global significance; and
· global information services related to research in agriculture, forestry and fisheries.
Currently, almost all CGIAR Centres are also involved in applied, and sometimes even adaptive, research on natural resources conservation and management, germplasm improvement and breeding, and development and management of production systems. However, this type of research is ultimately better carried out by national systems.
The rationale for developing options for change in the CGIAR is to expedite the process of implementing the long-term vision by improving the efficiency and effectiveness of the System. This would require the CGIAR System to be more selective in its goals and activities than it is at present. Clearly, such a change cannot be achieved instantaneously. A transition period is essential to allow for a gradual change.
In the 1990 expansion report, TAC attempted to define the possible evolutionary path, from the current situation to the long term, in the context of possible institutional arrangements to be implemented in the medium term. In the medium term, TAC envisaged the CGIAR System as having two major types of activity: global and ecoregional. Global activities would comprise strategic research on selected commodities and subject-matter areas, while ecoregional activities would focus on applied and strategic research on the conservation and management of natural resources, on development and management of production systems, and on commodity improvement. Both these types of activities need to be undertaken in partnership with non-CGIAR institutions and be closely interlinked.
The CGIAR has already taken an important step in 1993 by endorsing TAC's proposals to provide core support to a number of Systemwide programmes of high priority. Such 'programme' funding provides an attractive and innovative mechanism to promote effective partnerships. It should be noted that the original objectives of the CGIAR, adopted in 1971, did not refer to 'centres' as such. The objectives mentioned "examining the needs of developing countries for special efforts in agricultural research at the international and regional levels in critical subject sectors unlikely otherwise to be adequately covered by existing research facilities, and to consider how these needs could be met", and "reviewing the financial and other requirements of those international and regional research activities which the CGIAR considers of high priority." (First Review of the CGIAR, January 1977.)
2.3.1. Activities
2.3.2. Regions
2.3.3. Agroecological Zones/Ecoregions
2.3.4. Production Sectors
2.3.5. Commodities
In the 1992 Review of CGIAR Priorities and Strategies, TAC made recommendations on CGIAR priorities by activity, region, agroecological zone, ecoregion, production sector and commodity. At TAC 63, the Committee revisited these and its views are presented in the following sections.
The CGIAR activities are broadly categorized as:
Category 1: Conservation and management of natural resources including germplasm conservation (biodiversity). This area consists of two major types of activities - ecosystem conservation and management, and germplasm collection, conservation, characterization and evaluation;Category 2: Germplasm enhancement and breeding;
Category 3: Production systems development and management;
Category 4: Socioeconomic, public policy and public management research;
Category 5: Institution building. This area includes training and conferences, documentation, publication and dissemination of information, organization and management counselling, and capacity building networks.
The TAC-recommended balance of effort among these categories in the medium term was 18% for category 1, 22% for category 2, 29% for category 3, 11% for category 4, and 20% for category 5.
Among the major groupings of activities, TAC considered the following as possible candidates for rationalization of efforts as the CGIAR System matures:
a) Category 1, activity 1.2 - Germplasm collection, conservation, characterization and evaluation. This activity accounts for 7% of the System's core resource allocation. There could be economies of scale in the consolidation of these activities among the 13 centres that hold germplasm collections. In commenting on the "Stripe Study of Genetic Resources in the CGIAR", TAC has proposed modalities for the establishment of a Systemwide programme on genetic resources with a separate governance and funding mechanism.b) Category 3 - Activities in the area of production systems development and management. In the 1992 review of CGIAR priorities and strategies, TAC considered that, in the medium and long term, there would be a reduced need for these CGIAR activities because of increasing national capacity in developing countries to deal with them. These activities are mostly of an adaptive and applied nature, although the results obtained feed into the planning of more upstream strategic research. As national systems assume greater responsibility for research on production systems, the modes of operation could evolve towards greater use of networks and consortia.
c) Category 4 - Activities in the area of socioeconomic, public policy and public management research. The issues addressed by policy and management research have common features and common research approaches, and TAC is exploring mechanisms for a closer integration of these programmes and activities. A stripe study of public policy and management research is planned.
d) Category 5 - Activities in the area of institution building. In this area, there is a considerable amount of overlap between centres. The CGIAR is not equipped to provide comprehensive institution building services. TAC considers that some of these services could also be provided by other institutes and that considerable resources outside the CGIAR are available to support them. The CGIAR contributes to strengthening national research systems through its collaborative research activities as well as through the technology it helps to develop.
Currently all centres have training facilities and organize training programmes which are often targeted towards the same national scientists. While some of the courses are specific in relation to the mandate of particular centres, others are of a more general nature (e.g. farming systems, GIS, gender, etc.) and therefore could be organized on an inter-centre or Systemwide basis. The centres are already actively pursuing opportunities for cooperative activities in training which could result in System savings. All centres are involved in the area of information collection, dissemination and management. There may be room for consolidation and improvement in efficiency in this area if a System perspective were taken rather than a centre-by-centre perspective. Centres will be meeting at ISNAR in June 1994 to discuss opportunities in this regard. There may also be scope for rationalization of organization and management counselling activities as national research systems gain capacity, and other agencies are now equipped to provide this type of service.
In its 1992 priorities report, TAC suggested that, in the medium term i.e. by 1998, the distribution of CGIAR core efforts by region, should be 39% to sub-Saharan Africa (SSA), 33% to Asia and the Pacific, 17% to Latin America and the Caribbean, and 11% to West Asia North Africa (WANA).
Given the widespread poverty in the region and the limited capacity of national research systems, TAC considers that the sub-Saharan Africa region remains of highest priority to the CGIAR. The other three regions could be candidates for reduction in CGIAR support in the longer term. Economic growth and development should allow several countries in these regions to invest increasingly in agricultural, forestry and fisheries research, and to provide an increasing share of funding for CGIAR activities to their benefit. It is to be noted, however, that while scientific capacity has been enhanced strongly in many countries, lack of operational funds remains a major constraint to the effectiveness of most national research systems.
In WANA, national research systems have rapidly developed their capacity to undertake applied research. For example, Egypt, Morocco, Tunisia, Turkey and Pakistan have significant research capacity and should be able to assume increased regional roles. A similar situation is arising in Latin America with strong national research systems capable of assuming regional responsibilities, such as in Brazil, Colombia, Chile and Argentina. Other national systems in the region are also gaining research capacity.
With respect to Asia, countries such as India, China, Taiwan, Malaysia, Indonesia, Thailand, Philippines and South Korea are emerging as countries with strong national research systems. India and China particularly have strong national research systems which are not dependent on the CGIAR for their adaptive and applied research needs. Each of these countries could assume important regional research responsibilities.
In the 1992 review of CGIAR priorities, TAC recommended that the tropical agroecological zones (AEZs) and the cool subtropics with winter rainfall should receive increased attention. However, given the lack of data on previous allocation of CGIAR core resources by AEZ, the proposed allocations had no base line for comparison. Based on its knowledge of relative CGIAR efforts, TAC felt that the shifts in emphasis implied by the analysis were already well under way in the System.
TAC also allocated priorities by ecoregion and, in the framework of the 1994-98 medium-term resource allocation process, identified some priority programmes as Systemwide, core-supported initiatives (Table 1).
The highest priority ecoregions are the warm humid and sub-humid, and arid and semi-arid tropics and sub-tropics of Asia and sub-Saharan Africa. These four ecoregions account for more than two thirds of the priority index used by TAC in the 1992 priorities analysis. The warm humid and sub-humid tropics and sub-tropics with summer rainfall in Latin America and the Caribbean, and the cool sub-tropics with winter rainfall in WANA are also priority ecoregions. The ecoregions of lower priority are the cool tropics and sub-tropics of sub-Saharan Africa, Latin America and Asia when considered individually. TAC could consider core support for programmes for cool tropics and highland areas in different regions provided they are combined in a single Systemwide initiative. TAC also proposed core support for the Slash-and-Burn Programme which is considered a cross-ecoregional initiative.
Table 1: Priority Ranking of Ecoregional Programmes Proposed as Systemwide Initiatives
|
ECOREGION |
PRIORITY INDEX | |
|
1. |
Warm humid and sub-humid tropics and sub-tropics with summer rainfall in Asia and the Pacific (AEZ 2, 3, 6 and 7) |
212.6 |
|
2. |
Warm humid and sub-humid tropics in SSA (AEZ 2 and 3) |
166.8 |
|
3. |
Warm arid and semi-arid tropics in SSA (AEZ 1) |
136.8 |
|
4. |
Warm arid and semi-arid tropics and sub-tropics with summer rainfall in Asia and the Pacific (AEZ 1 and 5) |
121.1 |
|
5. |
Warm humid and sub-humid tropics and sub-tropics with summer rainfall in Latin America and the Caribbean (AEZ 2, 3, 6 and 7) |
113.6 |
|
6. |
Cool sub-tropics with winter rainfall in WANA (AEZ 9) |
81.1 |
|
7. |
Cool sub-tropics with summer rainfall in Asia and the Pacific (AEZ 8) |
63.6 |
|
8. |
Cool tropics and sub-tropics with summer rainfall in Latin America and the Caribbean (AEZ 4 and 8) |
42.0 |
|
9. |
Cool tropics in SSA (AEZ 4) |
33.1 |
|
| ||
|
Cross-ecoregional - Slash and Burn Programme |
N/A | |
SOURCE: Review of CGIAR Priorities and Strategies. Part II. Revised Chapter 13. 1993
The CGIAR has traditionally conducted research on crops and livestock of major importance. Since 1992, the CGIAR has incorporated in its portfolio research on forestry and fisheries. Of the four production sectors, fisheries is the smallest one, accounting for only 5% of global value of production, compared to 19% for forestry, 19% for livestock and 57% for crops. Currently, the CGIAR is allocating approximately 2% of its core resources to fisheries research, and work on inland aquatic resources and coastal resources in particular is considered to be of high priority.
With respect to forestry research, TAC has previously argued that there is a strong case to be made for the integration of forestry and agroforestry research, although the Group has agreed to the development of separate institutes. For livestock research, the CGIAR is already proceeding with the recommended structural change, i.e., integrating the two existing institutes into a new one with a global mandate.
TAC also recalls the recommendation in the 1992 priorities report that CGIAR programmes in agroforestry, forestry and fisheries should not be funded at the expense of critical research needs in crops and livestock.
The outcome of TAC's 1992 analysis on priorities by commodity and the implications for resource allocation by 1998, resulting from the 1994-98 MTP process, are presented in Table 2. It is important to note that the percentage shares, shown in this table, only refer to the sum of resources allocated to activity categories 2 and 3 which account for approximately half of the CGIAR core resources. The data should, therefore, not be treated as absolute figures but rather as a broad indication of the relative priority of the commodity. Further, it should be noted that in those cases where TAC has recommended phasing out or reduced relative priority for a specific commodity in the longer term, this would not include activity category 1.2 on germplasm collection, conservation, characterization and evaluation.
The livestock data represent an amalgamation of the multiple products of livestock and research on forage crops. In its 1992 analysis, TAC recommended that the CGIAR investment in livestock research should be limited to research on ruminants, and focus on smallholder mixed farming systems. The future strategy and scope for livestock research in the CGIAR is currently under discussion by the Group. TAC considers that the relative allocation of CGIAR core resources to livestock research should not be reduced below the current relative level.
Table 2: CGIAR Resource Allocation by Commodity (% Share of Core Resources to be Allocated by 1998)
|
COMMODITY |
% |
|
Livestock |
18 |
|
Rice |
17 |
|
Maize |
9 |
|
Wheat/Barley |
9 |
|
Cassava |
8 |
|
Potato |
5 |
|
Phaseolus Beans |
4 |
|
Sorghum |
3 |
|
Banana/Plantain/Yam |
3 |
|
Millet |
3 |
|
Groundnut |
3 |
|
Sweet Potato |
3 |
|
Chickpea |
2 |
|
Soybean |
2 |
|
Fababean/Lentil |
1 |
|
Pigeonpea |
1 |
|
Cowpea |
1 |
|
Coconut |
- |
|
Vegetables |
- |
|
Agroforestry/Forestry |
6 |
|
Fish |
2 |
|
Total |
100 |
SOURCE: CGIAR Medium-Term Resource Allocation 1994-98
Rice is the single most important food commodity in developing countries. TAC recommended that it should continue to be a commodity of high priority to the CGIAR, although some rice production systems are of higher priority than others. Since the marginal cost of research geared towards obtaining yield increases is high in upland rice and flood-prone rice (deepwater and tidal wetlands) farming systems, TAC considers that, in the long term lower priority could be given to them, particularly in Asia. TAC recognizes that upland rice is grown in particularly complex farming systems, and considers that research on these systems should be organized in the framework of a consortium, to ensure the necessary partnership in the research with other research agencies, who also need to provide expertise in the other dimensions of the upland farming systems.
The CGIAR will also continue to give high priority to maize and wheat research. While the private sector is increasingly important for research on hybrid maize, few international efforts, outside the CGIAR, focus on open-pollinated varieties which, in the long term, will remain of highest priority in the work of the CGIAR because of their importance for the poor. Even for hybrids, the private sector relies on tropical germplasm collected by CIMMYT. TAC has also re-affirmed that high priority should be given to wheat research. However, within wheat a distinction can be made between durum and bread wheats, and within bread wheats, between winter, facultative and spring wheats. Given its limited regional importance, the priority of research on durum wheat could decline in the long term. It accounts for only 5% of developing country wheat production, of which 70% is grown in WANA, but of the latter area, only 40% is grown in dry rainfed farming systems. Moreover, considerable research efforts are made in several Mediterranean countries, in particular, Italy and Spain. Similarly, the priority to be assigned to research on triticale should be carefully assessed. Despite considerable investments by the CGIAR, there has been only minor adoption of the crop by developing country farmers. While triticale is no doubt a superior substitute for rye, it is a poor substitute for wheat or barley. To date triticale is grown on 2.5 million ha of land but on only 165,000 ha of land in developing countries, of which 90,000 ha in Brazil and 36,000 ha in North Africa.
Cassava is an important crop, particularly for low income producers and consumers. A large share of its production, particularly in Asia and Latin America, is used as livestock feed. Cassava has negative income elasticity of demand in most regions, and postharvest technology (which is mostly outside the scope of the CGIAR) is an important element of research efforts. On balance, TAC recommends that in the longer term, CGIAR efforts in cassava research could be reduced in Asia and Latin America, and maintained or expanded in sub-Saharan Africa.
The importance of potato in developing countries is growing, and substantial payoffs have been obtained from CGIAR investments in this commodity. However, a significant amount of potato research, which is often highly relevant to developing countries, takes place in developed countries. Research results can be transferred. The CGIAR has also over-invested in potato research, relative to the TAC priority index. Sweet potato is primarily grown in China, which produces over 80% of total developing country output, and has a strong national research system. Because of its low income elasticity of demand, the importance of sweet potato, as a food, is decreasing steadily as incomes grow. In recent years, there have also been shifts in utilization of sweet potato as a staple food to livestock feed. In its 1992 report on CGIAR priorities, TAC considered that in the medium term, CGIAR efforts in potato and sweet potato could be maintained. In the longer term, however, on balance TAC considers that the priority of potato and sweet potato could be reduced.
While recognizing the potential importance of soybean for resource-poor farmers, particularly in sub-Saharan Africa, TAC considers that there are many other suppliers of research on this commodity. The level of CGIAR core support to this commodity could, therefore, be reconsidered in the long term. The priority of millet research was re-affirmed, particularly for sub-Saharan Africa. In Asia, core support for millet research could be reduced because of the strength of the national research system in India. Sorghum is an important crop in semi-arid farming systems where poverty is dominant. However, progress in obtaining research results is slow because of the difficulty of the crop leading to a high marginal cost of research. There are also other suppliers of research on this crop. TAC therefore recommends reduced core support for research on this crop in the long term. The relative priority of banana and plantain has been re-affirmed. Most of the CGIAR investment in this area is through a network approach and collaboration with other institutes.
TAC also considered the priority of food legumes. The phaseolus bean is an important commodity on equity and sustainability grounds, but the CGIAR has over-invested in this commodity relative to the TAC priority index. It is also a crop which predominantly is of importance in certain areas of sub-Saharan Africa and Latin America. In the long term, its priority to the CGIAR could be reduced. Chickpea is an important dietary item in south-east Asia, India and the WANA region. Lentils are important in WANA, Bangladesh and India which together account for nearly 90% of the production in developing countries. In 1992, TAC recommended the phasing out of CGIAR support to lentil in the medium to long term on the basis of its geographically-limited importance. Turkey also has substantial research capacity for this crop and could assume regional responsibilities for WANA. Production of cowpea is concentrated in West Africa and Nigeria accounts for over 50% of total production and for over 80% of the area grown to cowpea. Nevertheless, cowpea is an important crop in mixed farming systems of semi-arid Africa, and TAC considers that there is no obvious alternative source of supply of research on this crop. TAC has already recommended that CGIAR support for fababean should be phased out. Pigeonpea is an important crop in one country only, India. ICRISAT has made rapid progress in obtaining research results on the commodity, particularly through the development of a hybrid variety. TAC recommended that in the medium term, CGIAR support for this crop could be maintained but considers that, in the long term, CGIAR support to pigeonpea research could be phased out. Overall, TAC considers that legume research should be conducted in an ecoregional research framework because of the nature of their integration in complex farming systems and localized importance of food legumes.
TAC has re-affirmed the priority to be assigned to research on coconut and on vegetables, although to date, no core resources have been assigned to commodity improvement programmes for these commodities. The Committee also re-affirmed the relative priority it has assigned to research on forestry and agroforestry, and to fisheries, particularly with respect to the resource management issues.
2.4.1. Institutional Capacity
2.4.2. Governance and Management
In the 1994-98 medium-term resource allocation process, TAC made an assessment of each centre's institutional capacity to deliver an effective research outcome. TAC evaluated each centre according to a set of largely supply-oriented criteria, which were equally-weighted. These were:
· the strategic character of the centre's research programme;
· the potential for breakthroughs;
· past performance and likelihood of continuance (or improvement);
· the external environment, institutional health and quality of management; and
· capacity to collaborate with NARS, other CGIAR Centres, and advanced institutions.
The outcome provided TAC with an indication of the broad perception of the centres' capacity to deliver an impact-oriented, quality research programme of a strategic international character.
TAC also concluded that several centre programmes were operating below critical mass. Furthermore, the scale of operations of several centres was below the optimum resulting in high overhead costs. The share of centre governance and management costs in several centres seems to be high relative to comparable institutes outside the CGIAR. If a centre becomes too large, bureaucracy is likely to creep in and the centre may no longer be able to function effectively. In considering a future structure of the CGIAR, TAC has therefore also taken into account the need for a critical scale of CGIAR programmes so that they can operate efficiently with minimum overhead and transaction costs.
In an average CGIAR Centre, the cost of governance, management, and administrative and finance units is estimated at well over US$ 2 million per year. The CGIAR's decentralized operating structure of independent centres accounts for these relatively high costs. For example, costs arise from the operation of independent Boards of Trustees or Governors, from external reviews, from planning and resource allocation activities, from consultations with the CGIAR's central mechanisms, from participation in CGIAR meetings, and from activities related to external contacts. These activities are considered essential to maintain the crucial balance between freedom for centres to maximize their creativity, and the accounting requirements imposed on the CGIAR as an international publicly-supported System. However, the examination of options for cost reductions across the System cannot ignore the exploration of savings opportunities in these areas.
In addition to the governance costs, independent operations of individual centres also require maintaining a service infrastructure at most centres. This infrastructure typically consists of training and publication facilities, administrative and financial services, in addition to executive leadership.
Options for maximizing the use of donor funds to support CGIAR research programmes should also take into account future needs for physical plants and facilities. For field research to be productive in often difficult conditions, CGIAR scientists must be well supported by appropriate centrally-provided tools and facilities. In some instances, the remote location of research stations requires the provision of essential services such as living accommodation and schooling facilities for scientists and their families.
The costs of these services historically may not have been an important parameter in CGIAR resource allocation decisions since the high returns from CGIAR research investments far outstripped the costs involved. Furthermore, centres wanted to make the work environments attractive to high quality staff. In the present financial climate, all costs should be carefully evaluated but it may be inappropriate to give them undue weight in the overall decision-making process.
Efficiencies have been sought in staffing as well as in the operation of research facilities in the past two or three years. However, the continuing high costs of centre operations due to political and economic instability in host countries, or high costs of physical plant renovation of aging facilities should be considered in decision making. The latter perhaps should be of greater concern as it is evident that the level of resources to renovate, in the short term, all of the CGIAR's aging facilities is unlikely to be available. Maintenance of aging physical plant facilities may claim a larger share of resources in the future. Already, six centres use more than 10% of their core resources for this infrastructural cost. CGIAR Centres have already actively pursued opportunities for efficiency gains in this area.
The fixed nature of these facilities allows limited room for scaling down at the margin, however. A facility has to be operated even if the number of scientists is reduced. As a consequence, in several instances, in addition to considerations of programmatic critical mass, there may come a point where it is no longer cost-effective to operate a downsized physical plant.
2.5.1. Research
2.5.2. Research-Related Activities
2.5.3. Administration, Governance and Management
Opportunities for greater cost-effectiveness should be explored by taking a Systemwide rather than the narrower centre perspective. Centres have historically been independent in the management of their resources and activities. As successive external reviews have indicated, centres have generally managed CGIAR resources efficiently and have taken every opportunity to make even more effective use of their resources. In the light of measures taken by individual centres in response to financial stringencies in the last few years, there are very few opportunities left to improve efficiency at the centre level. However, because of their independence, every centre has been required to provide a full spectrum of research support and research-related activities. Further efficiency gains could possibly be achieved by taking a System perspective in the organization of some of this work. In particular, it should be examined whether the System can be re-organized so as to produce the same output mix at a lower cost. This issue is addressed in subsequent sections.
In the 1994-98 CGIAR medium-term resource allocation process, the proposed centre budgets had the following characteristics: on average 62% of the System's effort was allocated to research and research support; 14% to research-related activities such as training and information; and 24% to administration, management and governance. The opportunities for Systemwide efficiency gains in each of these areas are addressed in the following sections.
While the overall share of core resources assigned to research and research support averages 62%, it ranges from a low of 52% to a high of 67% among centres. When the share of research support is subtracted, no less than 8 out of 18 centres currently allocate less than 50% of their core resources to research per se. Given that research should be the major thrust and focus of the centres' activities, attempts should be made to increase the share of resources allocated to research. Also, the operations of several centres overlap in terms of commodities, activities and subject matter areas. There could be substantial scope for rationalization of efforts, elimination of duplication, and for the achievement of economies of scale. Moreover, the System will need to give much greater attention to the activities of other actors in the global agricultural research system for developing countries. These include the possibility of some stronger NARS and regional associations, taking greater responsibility for specific areas of research currently carried out by the CGIAR. Greater account should also be taken of the capacity of developed country research institutions for research of relevance to developing country situations. The critical issue for the CGIAR is the provision of research results that will assist developing countries in overcoming their agricultural, forestry and fisheries problems. The CGIAR should not be concerned about how or by whom these results are provided as long as the supply is efficient, timely and of low cost.
For these reasons, TAC will accelerate the series of stripe reviews proposed in the medium-term resource allocation process. These reviews will examine current activities in particular areas of research, suggest where appropriate, changes in the CGIAR strategy, and explore alternative ways for organizing the work. The Stripe Study of Genetic Resources has already been completed. The other on CGIAR Regional Commitments in West Africa is in progress, and reviews on public policy and public management, roots and tubers and cereals research are being planned. Each of these reviews will also pay attention to the supply dimension of research and critically assess the role of the CGIAR with respect to those of other institutes.
The System currently allocates 14% of its core resources to training, conferences, documentation, publication and dissemination of information. There is substantial scope for rationalization of these activities. Many centre training courses address topics of wider interest than those of the particular centre, such as farming systems research, geographic information systems, gender analysis, data processing, statistics, and biotechnology. Further, centre training efforts often target the same recipients in national research systems. In the area of information, substantial efficiencies could be obtained by making greater use of electronic communication facilities.
Centre Directors have already acknowledged the scope for efficiencies in both training and information, and are considering ways and means to offer the same products and services at a lower cost. Furthermore, the CGIAR is not the sole provider of training and information services. With respect to publications, every centre has its own production infrastructure and facilities. Opportunities for efficiencies in this area should also be explored.
Approximately 2% of the System's core resources is presently allocated to institution building networks. Several of these networks overlap and could be organized more efficiently, thereby also streamlining relations between centres and national research systems.
The CGIAR currently assigns 24% of its core resources to administration, governance and management including expenditures on physical plant operations. This share ranges from 16% to 31% among individual centres.
Overall, the average cost of administration, governance and management in a CGIAR Centre amounts to US$ 2.25 million a year. Opportunities should be sought to reduce these costs, for example, through sharing of administrative facilities, use of computer networks, simpler management structures, etc.
The cost of physical plant operation of centres ranges from 18% of core resources to zero for those centres using rented facilities. Six centres use more than 10% of their core resources for this infrastructural cost. This issue will merit careful attention as the CGIAR plans its future operations.
CGIAR Boards range in size from 8 to 17 members, with an average of 14. The aggregate cost of operating these Boards is in excess of US$ 3 million per year. It should be possible to govern centres more efficiently with smaller Boards, or to use common Boards for more than one centre.
Rationalization of central services, i.e. those required by the System, should also be considered. For example, by increasing the efficiency of the review process, and extending the period between reviews, TAC believes that the total cost of external reviews could be reduced considerably.
