5.1. The Evolution of the National Programs and the Role of the Centers
5.2. Duplication
5.3. The Ecoregional Approach and Alternative Organizations
5.4. Governance
5.5. Relations with Partners
5.6. Center-Specific Recommendations
Here we repeat the main questions and summarize the Panel's answers and recommendations with appropriate illustrations from individual centers. Where we lack adequate information on some issues to present justifiable recommendations, we note the issue as one for future external review in the recommendations about individual centers in the following section.
In view of the growing strength of the national programs, the Panel recommends shifting more Center effort to basic and strategic research.
Conservation and management of natural resources; and Germplasm enhancement and breeding (Categories 1 and 2). The Panel recommends that IITA and ICRISAT shift resources into these categories out of Category 3. The other Centers, including WARDA, do so little in West Africa that this recommendation cannot practically apply to them.
Production systems and management research (Category 3). The Panel recommends that activity category 3 be the subject of an explicit devolution policy in IITA and ICRISAT in order to augment work in Categories 1 and 2.
Institution building (Category 5). The Panel recommends that the activities of the IARCs, with the exception of ISNAR, in Category 5 be limited to training and information because the Centers lack comparative advantage in institution building as such. It recommends that the Centers restrict their activities in capacity building of national programs to a strict minimum, with the obvious general exception of training.
The Panel concluded that duplication on commodities, as might be said to have existed on maize, cassava, or rice in the past, is not a significant inefficiency. The appearance of duplication on themes - striga, cropping systems, soils, economics and public policy, to cite a few examples-occurs because the variability of common problems across the region makes it inevitable. The Panel found numerous collaborative mechanisms to avoid real duplication on such themes, concluded that this form of duplication was not a major cost to the system, and recommended no novel steps to eliminate this form of duplication. There probably was some duplication of training, but the Centers have recently begun to harmonize training activities.
The Panel contends that the "decentralized network" alternative would be inefficient for three reasons because (i) it would not reduce information costs so as to impose accountability and scientific performance; (ii) it would create too much uncertainty about resource availability and accountability; and (iii) the suggested benefits of the decentralized network approach are already gained in the Centers or in related efforts, notably that of SPAAR.
Should center boards become political bodies? The Panel found occasional support for this among national programs, but not general support.
Board size. Several interviews argued that consolidation of some Boards would be effective. The obvious regional example is IITA and WARDA, as those of ILCA and ILRAD have been merged. The Panel therefore recommends a common Board of Trustees for IITA and WARDA, with ex-officio representation of ICRISAT, ICRAF and IRRI (section 4.3.2) as a means of bettering the integration of research among those four Centers.
Staff composition. Some NARS staff proposed that the IARCs should hire more regional scientists. The Panel strongly discourages any such policy because it might lead to political appointments and debase the quality of Center scientists.
National programs (section 4.5.1). The contacts with national programs are on the whole efficient. The Centers have many mechanisms to inform themselves about national activities, to receive input into their research planning, and to collaborate substantively on common problems.
This section lays out the analysis, recommendations and suggestions for external reviews of the leading Centers, excepting WARDA, which is discussed in Annex 1.
Maize improvement. The next external review of IITA should make a definitive recommendation about devolving humid forest maize research to one or several national programs. IITA has already considered this internally, so there is information for a prompt decision.
Interactions with tree crop research centers. Professor Carl Eicher has written that "IITA should develop scientific partnership with tree crop research centers in the NARS in West Africa" (Eicher, 1992, p. 30). The Panel has the admittedly superficial impression that IITA has already done so, but it is a point that the next external review can easily verify.
The Niamey site. A grave problem in the institutional culture of ICRISAT was always the domination of Hyderabad over West Africa. That domination forced the selection of the Niamey site and the insistence on research directions that should have been known to be fruitless (eg, the operational scale on-station trials that were mechanically copied from India to Africa). Much of the lack of production impact of ICRISAT in West Africa is due to the concentration at Niamey, which prevents effective work on sorghum and groundnut and relegates millet work to a fringe of the mandate area. Being in western Niger isolates millet improvement from such management issues as intercropping, mechanization, complex cropping patterns, and rotations because it works in a farming system which is quite unrepresentative of major systems in the West Africa SAT. Niamey is unlikely ever to be a regional center of academic excellence, in the way that Zaria would obviously have been. We are well aware that being at Niamey does not restrict operations to western Niger, but location determines much of what scientists do and Niamey is not a fully representative location 18.
18 ILRI, in commenting on this paragraph in the draft report, contended that the Panel's remarks on the location of the ISC "will not improve the location of the CGIAR's immovable investment at Sadore" and "that criticism of past decisions is not helpful". The really important investments - staff - are of course movable and, in criticizing past decisions, we are trying to help the system avoid violating the second rule of medicine: if something isn't working, stop doing it.
ICRISAT has started to fix this error. It has built regional sorghum programs in Nigeria and Mali. It is moving groundnut staff to Bamako and Kano. It has sought to widen the prospective benefits of research at Niamey by building a training and information center and by broadening collaboration with ILCA on crop-livestock management, with ICRAF on agroforestry, and with IFDC and others on soils and land use. The Panel recommends that the ICRISAT research resource allocation model be continuously applied to the benefits and costs of the Center's investment at Niamey with the specific objective of justifying it not only within ICRISAT, but across IARC and NARS partners in the region. The Panel recommends that ICRISAT sorghum work should be strengthened in Nigeria. It now consists of only a breeder, an entomologist and an agronomist and lacks capacities to do more basic work. The planned transfer of ICRISAT groundnut staff from Niamey to Kano should be strengthened by addition of new staff.
Despite the decisive and correct steps taken by ICRISAT management to shift emphasis in West Africa, there remain two fundamental problems that do not appear to be adequately treated by ICRISAT; the first problem affects ILCA as well.
Crop and livestock management. There is still too much crop and livestock management and characterization research in the IARCs. That research (CGIAR category 3) has little expectation of additional benefit because it often duplicates what farmers already know, what they can easily learn without research, what is a matter of extension, or what the national programs can do. The Panel is highly confident in its critique of this first problem and the report is definite about what to do about it (section 4.4.2).
Crop improvement. We are less confident in our critique of the second problem. The breeding work of ICRISAT has not produced materials for general or even location-specific use in West Africa and so we begin from that fact, which ICRISAT, in a 20 page reply to the draft report, did not really dispute. A millet breeder in the national program in Niger said that ICRISAT had made no progress since the late 1970s. Interviews with ICRISAT staff and national staff, reading on the subject and previous knowledge suggest that a more basic approach is needed. While we do not like to recommend additional reviews, we do not have the technical competence to say what to do about ICRISAT crop improvement work in West Africa. The Panel recommends a very high-level review of ICRISAT's crop improvement programs in West Africa, including that of CIRAD for sorghum, one that brings independent biological scientists with no previous connection to ICRISAT staff, management, or Board.
ILRI is, of course, uncertain about its mid-term program in Africa. It is hazardous to speculate about what it will do. Nonetheless, we get the profoundly depressing feeling that the CGIAR needs yet another rethinking of what it is trying to do with livestock research in West Africa.
Effects of the new ILRI Mandate on CIRDES and ITC
The expansion of ILRI's mandate outside Africa, and the contraction of its resources from the sum of those of ILCA and ILRAD, means that ILRI will do less in West Africa than ILCA and ILRAD did. That fact and the completion of major ILCA and ILRAD work - the Malian Delta program, the trypanotolerance network, the subhumid zone studies in central Nigeria, various studies at ITC - implies greater responsibilities for national and regional efforts, including those of ITC and CIRDES. Growing roles for CIRDES and ITC without CGIAR funding are consistent with devolution and with a greater strategic element in the ILRI mandate, given that CIRDES and ITC work is applied. CIRDES and ITC do not have the financial base or scientific stature to join the system and the Panel does not recommend any CGIAR support to CIRDES or ITC, but they should be encouraged to compete for funds allocated by ILRI (ILRI, 1994, p. 25).
ILRI Program
The interactions of animals with crops, pastures, trees, and land and water resources impose a research organization in livestock distinct from that in crops. Scattering staff - 1 in Bobo-Dioulasso, 1 in Bamako, 2 in Kaduna, 2.5 in Ibadan, 3 in Niamey - has not produced and will not produce usable results. A small bilateral crops program (eg, ICRISAT in Mali in the 1980s) of a breeder and an agronomist can succeed because it can distribute materials for testing. It can execute a large diverse program with low supervision costs per experiment, which is impossible with livestock studies. A small animal research team in an analogous fashion cannot succeed unless system characterization is all that is sought and that now can be done more cheaply by the national programs.
Concentrating on animal production per se - genetics, especially - will not pay except in rare circumstances (eg, ranches or specialized dairy small holdings in the cool highlands) that are irrelevant for most African producers. The focus has to be on primary production for greater feed output. ILCA now has 2.5 staff (1 at Niamey, 1.5 at Ibadan, 0 at Bobo-Dioulasso, 0 at Kaduna, 0 at Bamako) working on primary production. This is not enough.
The ILRI presence at Bamako - one staff doing only liaison with national and regional programs - should be converted into a scientific post at Niamey or Ibadan because it is too expensive in isolation. Liaison offices of this type are only relevant for the germplasm Centers (the IRRI INGER coordinator at IITA is a good example) in which a lone scientist can have a full program managing the distribution, conduct, and analysis of trials. A strong indicator is that some national and regional programs visited by the Panel were unaware of the liaison office.
The ILRI group of two scientists in central Nigeria is too small to be effective. It requires more staff to achieve critical mass, better physical facilities for program continuity, and close interactions with national or international crop scientists. There are at least three alternatives for that group.
Merge ILRI/Kaduna into NAPRI (National Animal Production Research Institute). NAPRI staff told the Panel that the chief barrier to an effective ILCA program had been discontinuity and argued that continuity could be achieved by basing ILRI staff at NAPRI. The Panel rejected this suggestion on the grounds that, while continuity, critical mass and interactions with crop scientists are problems, an ILRI program in NAPRI would not have a specific international comparative advantage - it would not add anything different to what NAPRI does, as staff of the latter stated forcefully to the Panel.
IFPRI has neglected Nigeria in favor of Burkina Faso, Niger, Senegal, and the Gambia. The Panel understands that the funding cycle has not lately permitted IFPRI to post staff in the region, though this may change as new special projects develop, including one in Ghana. However, even when IFPRI had 12 staff in West Africa, it had none in Nigeria. Moreover, its Washington staff have written little on Nigeria. Whatever may be the reasons for that past neglect, the Panel recommends that IFPRI post 1 or 2 permanent scientists at Kano or Ibadan, with the staff having the same links to the Nigerian universities that IITA and ICRISAT have. One of those scientists would be the regional coordinator of the social science work.
IFPRI replied to the draft report that: i) it did not consider that policy research now had a positive expected impact in Nigeria; ii) it vigorously seeks possibilities for additional work on Nigeria; and iii) it has developed a joint research proposal with Nigerian institutions. IFPRI has further objected to the recommendation of putting staff with a regional coordination role in Nigeria, apart from doing research on that country from abroad, because of logistical difficulties and because qualified Nigerians can be brought to IFPRI, and to other external institutions.
