Aggregation bias and its bounds in measuring technical efficiency

Abstract

The discussion about the bias due to input (or output) aggregation in efficiency measurement based on data envelopment analysis, recently revisited by Tauer (2001) and Färe and Zelenyuk (2002) is continued. Attention is focused on the direction and the bounds of the aggregation bias.

Notes

Also, they reformulated the traditional Farrell-type technical efficiency measure into a subvector estimator.

 Input orientation was chosen for the purpose of illustration and similar analysis can be done for the output orientation as well. Also note that the framework of variable returns to scale technology is chosen, modelled by including the restriction

to the formulation. Similar analysis can be done for the other types of returns as well.

 To be precise, it is not appropriate to call the measure in Equation 3 ‘technical’ efficiency, since it is not independent of prices. A better name might be subcost efficiency.