Sensitivity and stability analysis in DEA: Some recent developments

This paper surveys recently developed analytical methods for studying the s ensitivity of DEA results to variations in the data, The focus is on the st ability of classification of DMUs (Decision Making Units) into efficient an d inefficient performers, Early work on this topic concentrated on developi ng solution methods and algorithms for conducting such analyses after it wa s noted that standard approaches for conducting sensitivity analyses in lin ear programming could not be used in DEA, However, some of the recent work we cover has bypassed the need for such algorithms. Evolving from early wor k that was confined to studying data variations in only one input or output for only one DMU at a time, the newer methods described in this paper make it possible to determine ranges within which all data may be varied for an p DMU before a reclassification from efficient to inefficient status (or vi ce versa) occurs. Other coverage involves recent extensions which include m ethods for determining ranges of data variation that can be allowed when al l data are varied simultaneously for all DMUs, An initial section delimits the topics to be covered. A final section suggests topics for further resea rch.