AID experience in agricultural research : a review of project evaluations

agricultural research (AR) policy, this paper summarizes A.I.D.'s past and present AR activity and identifies major issues affecting AR projects. First, major trends in AR over the past 10-15 years are reviewed: the development of new technologies adapted to both adverse and optimal conditions; and greater emphasis on onfarm research, communication between researchers and farmers, support for national AR systems in developing countries, and longer time frame requirements. A review of A.I.D.'s AR experience reveals a changing pattern. In the 1950's, the belief that adequate technologies already existed made AR a low A.I.D. priority, while congressional restrictions and limited in-house expertise inhibited AR activity in the 1960's. The 1970's New Directions strategies emphasized other development initiatives. From 1978 to 1981, however, a new awareness of the lack of appropriate technology needed for food production increases and the success of the green revolution led A.I.D. to increase funding for AR by almost 70% to $143 million, with most of the funds going for mission projects. Next, 131 evaluations of 48 A.I.D. AR projects are reviewed. Standard evaluations were found to focus on inputs and outputs; to provide neither an adequate picture of project performance nor the data needed to compare projects; and to not fully treat problems entailed in doing AR in developing countries, the need for more off-station AR and dissemination of findings. Only the four Impact Evaluations included in the sample measured project effect on beneficiaries. The documents did, however, reveal several recurring issues in AR projects. For the 39 regional and mission projects reviewed, these focused on problems in conducting AR and involving farmers in it, AR quality and priorities, construction delays and the time allotted to AR, and inadequacies in inputs and A.I.D. supervision. Issues in the nine centrally funded projects reviewed concerned linkages with host country institutions, performance of long-term staff, project scope and funding, and AR quality. The study indicated the need for more than the standard evaluations of A.I.D. AR projects. Fourteen appendices include a 246-item bibliography (1963-81) of documents examined for this report.