University Efficiency: A Comparison of Results from Stochastic and Non-Stochastic Methods

Efficiency scores are determined for Canadian universities using both Data Envelopment Analysis (DEA) and stochastic frontier (SF) methods for selected specifications. The scores are compared. Although there is some consistency, there is also considerable divergence in the efficiency scores and their rankings. Besides choice of the DEA or SF method, scores are sensitive to the definition of output, the inclusion of environmental (i.e., primarily firm-specific) factors, and the procedures for their inclusion (one-stage and two-stage methods). Despite the divergence among methods and specifications noted, the relative positions of individual universities across sets of several efficiency rankings (e.g., all the DEA and SF outcomes) demonstrate consistency. An analysis of rankings provides a range of potential rankings for each university. High and low efficiency groups are evidenced but the rank for most universities is not significantly different from that of many others. The results indicate that, while efficiency analysis can be helpful, decision makers need to be very cautious when employing efficiency scores for management and policy purposes and they recommend looking for confirmation of the usual efficiency analysis.