Comparative Study of Data Envelopment Analysis and Other Approaches to Efficiency Evaluation and Estimation.

摘要

Data Envelopment Analysis (DEA), a new methodology based on mathematical programming models, provides an approach to evaluation of the relative efficiency of organizations, especially not-for-profit organizations which have multiple outputs and inputs. This paper uses an artificial data base to evaluate DEA relative to other alternatives such as ratio and regression analyses. The results of this study generally favor DEA not only for identifying inefficiencies but also for locating their sources and estmating their amounts in particular DMUs (Decision Making Units). Statistical regressions performed very poorly, per se, as well as by comparison. Reasons for the poor performance of these customary statistical regression approaches are indicated along with possible ways of improving this performance.