Using evaluation to improve management performance : some preliminary thoughts

This paper examines three possible reasons for this. (1) Evaluation methodologies have not adequately recognized the multiple and conflicting objectives of irrigation projects, and as a result, evaluations do not always correctly specify project problems and pitfalls. Evaluators should begin by evaluating the incentives associated with particular irrigation behaviors; they should also take the "bottom-up" view seriously and use criteria suggested by farmers in developing measures of project performance. (2) There is, on the one hand, a paucity of relevant data and, on the other, quantities of statistical data that shed little light on significant problems. As data collection is costly, evaluators should not overreact to data shortfalls by overcollecting. To maximize the utility of data, their collection for evaluation, monitoring, and research purposes should be dovetailed. The paper suggests some quick data collection techniques and some indicators of overall system performance that incorporate risk and change. (3) It is often difficult to translate the complex realities of an irrigation project into lessons learned. In this regard, the paper focuses on the evaluation team and presents suggestions for team composition and deployment. Also discussed are problems associated with the dissemination of and generalization from the findings of evaluations. A 9-page bibliography is included. (Author abstract, modified)