Impact assessment of the SAFGRAD commodity networks

Impact evaluation, conducted 6 months after the project's 7/92 conclusion, is based on in-depth reviews of 8 of 26 participating countries. SAFGRAD research on maize and cowpea in West Africa has had substantial impact. In Ghana, for example, the area in improved maize cultivars increased from 20% in 1982 to 55% in 1991. From 1985-92, the annual social benefits (lower food prices for consumers, lower production costs for farmers) from maize research ranged from $5.5 million to $84 million, with an estimated rate of return of 73%. The estimated social benefits for maintenance research on cowpea in Mali and Burkina Faso were from $800,000 to $12.3 million during the 1984-91 period. While farm-level diffusion of new sorghum varieties was substantially less (possibly due to an overemphasis on breeding) than for maize and cowpea, S-35 has been successfully introduced into northern Cameroon -- where, over 7 years, social benefits are estimated to range from $288,000 to $831,000 -- and, more recently, into Chad. Since sorghum and millet productivity is critical to improving nutrition and raising agricultural income in much of sub-Saharan Africa, future research should focus on these two food grains, which are much more important than maize in drier areas such as the Sahel. SAFGRAD regional research networks have been a major mover of new technologies. In the countries examined, about 50% of the maize and sorghum varieties and almost 75% of the released cowpea varieties had been in SAFGRAD trials. These results are striking, given that varieties distributed through SAFGRAD cowpea trials represented only 14-54% of the total number of varieties tested in the five study countries. The same trend exists for maize. The most significant spillover effects among the member countries of SAFGRAD networks occurred in the maize network, where nine varieties were adapted and released in one country and, in turn, tested and released in other countries. The cowpea network had spillover of four varieties in the countries examined. The West and Central Africa Sorghum Network had spillover of one variety, the Eastern African Sorghum Network of three. In West and Central Africa, a total of 103 new maize, sorghum, and cowpea cultivars were released, while in East Africa, 26 new sorghum and millet technologies were released in two study countries (Ethiopia and Kenya). The proportion of germplasm contributed by the NARS to SAFGRAD trials has also increased, accounting at present for about 50% of cowpea germplasm and 60% of sorghum germplasm (maize is not indigenous to Africa). These results confirm the networks' success in sharing technology between countries and the increasing role of the NARS in the networking process. Significant building of research capacity has occurred. More than 30 scientists received M.S. or Ph.D. training, and several are now research leaders in their respective countries. In addition, short-term training was provided to nearly 400 NARS researchers and technicians in more than 22 countries, and 165 individuals participated in scientific monitoring tours to NARS or IARC programs. SAFGRAD also: organized biennial conferences and network meetings attended by more than 2,500 researchers and produced about 500 publications -- 52% on technology generation and 48% on technology transfer. Despite this increase in research capacity, the operating funds made available to researchers have declined significantly in the last 10 years; national policymakers need to be made aware of the high social benefits of agricultural research such as are exhibited by SAFGRAD, especially since they will have to bear an increasing share of research costs. There is also a need to keep NARS (especially the smaller ones) linked to the international scientific network and to keep the work of the NARS free from domestic political pressures.