RADII OF CLASSIFICATION PRESERVATION IN DATA ENVELOPMENT ANALYSIS - ACASE-STUDY OF PROGRAM FOLLOW-THROUGH

Sensitivity and robustness of efficiency classifications for the addit ive model and its geometric equivalents in Data Envelopment Analysis ( DEA) are addressed. The minimum distance (measured by a Tchebycheff no rm) separating an organization from reclassification is computed via l inear programming formulations and shown to constitute a generalized ' residual' for each organization. Without this sensitivity information, findings can be distorted when marginally efficient or inefficient un its are distinguished solely on the basis of their classification. Ana lysis of these residuals from an earlier (inconclusive) DEA study furt her reveals how substantive differences in a sample's underlying group s can be detected. Properties of group efficiency and group proximity to the efficient frontier are investigated using these new indicators.