Non-Metric Data : A Note on a Neglected Problem in DEA Studies

Data envelopment analysis (DEA) is widely used to compare the empirical performance of public institutions such as law enforcement agencies, judicial authorities or national health care systems. Many DEA analysts, however, ignore the fact that DEA efficiency values are non-metric. They consequently do not hesitate to compute (arithmetic) means. They do not hesitate either to treat DEA values as metric data in econometric analyses. Instead of providing useful insights into the performance of public bodies, the confusion of non-metric data with metric data constitutes a lack of internal validity that may cause serious fallacies. Against this background, we believe that a clear warning against an uncritical processing and interpretation of DEA values is pertinent and should be routinely considered by efficiency analysts as well as referees of efficiency papers