Methodologies for assessing the impact of agricultural and rural development projects : a dialogue

sponsored an international workshop to discuss key issues concerning impact assessments of agricultural and rural development projects. Major findings and conclusions were as follows. 1. Current strategies for computing the net impacts of agricultural and rural development projects - statistical control and quasi-experimental design techniques - are impractical and should be replaced by rapid, low-cost data collection methods (e.g., informant interviews, community meetings, surveys) to determine whether anticipated or unanticipated change occurred in core impact areas, and whether these changes can plausibly be related to the project. 2. Production impact should be assessed by examining not only a project"s impacts on targeted crops, but also the project"s effect on non-targeted crops and activities, the use of the targeted crop, the persons responsible for production and marketing, the cost opportunity of the new crop, and gender issues. 3. Impact on household income can be assessed by conducting small, sample surveys for extended periods of time. These surveys should focus on household expenditures, key standard-of-living indicators, and intrahousehold control of income. 4. Besides food production, factors such as seasonality, crop mix, income, the role of women, crop labor requirements, marketing, and government policy are vital in assessing the attainment of food consumption objectives. Nutritional impact can be measured by simple and easily learned anthropometric methods. Also pertinent are questions about the frequency of meals consumed during different agricultural seasons and the number of nights the respondent or his/her acquaintances went to bed hungry. 5. Agreement is lacking among scientists and donors on suitable models for assessing project impact on natural resources and the environment. Such models, when developed, should include not only biological and physical aspects, but social and economic aspects as well. Excessive use of technical language and evaluators" sense of their inability to affect environmental issues are also problems. 6. More needs to be done to devolve assessment efforts to host country agencies, building, where necessary, the required institutional capacity.