Agricultural production support project

Mid-term evaluation covers 1982-87 and was based on interviews, document review, and site visits. The project has been plagued with organizational problems, both in terms of coordination among its many discrete components and in working with - rather than against - the Government of Niger (GON) agencies and services the project is designed to strengthen. Matters are complicated by the fact that the GON did not provide the human resources necessary to create the GON-project linkages needed to ensure project success, and that, in the case of some GON agencies, there simply is not much to help. For example, the extension component, which was designed to strengthen the national extension service, in fact may have done the opposite as it established competing (and better) extension services in the field. The input supply component did succeed in improving financial management at the GON"s agricultural input supply agency (Centrale d"Approvisionement, or CA), but efforts to privatize CA are plagued by its high operating costs and lack of autonomy. The seed multiplication component has created an infrastructure to distribute seed, but so far the seed it has developed has not been much better than the seed farmers can produce themselves, and its links with researchers has been minimal. The cooperative component, on the other hand, has been a success. TA and training (from basic literacy to inventory management) has been provided to 114 (vs. 200 targeted) co-ops, and credit has been extended to 37, with no defaults and relatively small arrearage overall. The major lesson of this project is to think carefuly before creating a project with discrete components, especially when each is managed by a different government agency or parastatal. Such projects need to be carefully designed to ensure effective coordination among all the agencies involved. Lines of authority need to be carefully and clearly established, and the risks of clustering must be outweighed by the benefits. This lesson is being applied to the design of a follow-on project, which will have fewer components, all related to the cooperative movement. Other lessons are: always work with an existing institution, no matter how weak, rather than create a new one; national-level participation must be ensured, even enforced, in institution-building projects; moving a parastatal toward self-financing will work only if the agency is allowed to operate as a business; cooperative training is most successful when defined by member input and given precedence over credit; contract growers can multiply seed but require close supervision. USAID/N notes that neither the GON nor project staff have accepted all the evaluation findings.