Is it efficient to analyse efficiency rankings? |
| |
Authors: | Uwe Jensen |
| |
Institution: | Institut für Statistik und ?konometrie, Christian-Albrechts-Universit?t, Olshausenstr. 40, D-24118 Kiel, Germany (e-mail: Jensen@stat-econ.uni-kiel.de), DE
|
| |
Abstract: | When production functions are estimated as frontier functions, the deviations from the frontier can be interpreted as individual
inefficiency estimates. Unfortunately, it has recently been shown that efficiency differences across individuals are very
often statistically insignificant. In this paper, we will analyse the consequences of the consideration of confidence statements
for the reliability of efficiency rankings. The stochastic frontier and confidence intervals derived by Horrace and Schmidt
are compared to the COLS approach and bootstrap confidence intervals. The membership function is proposed as a simple Monte-Carlo
approximation for the probability for an individual to be the most efficient in the sample.
First version received: May 1998/final version accepted: July 1999 |
| |
Keywords: | : Bootstrap confidence intervals frontiers inefficiency membership function ranking |
本文献已被 SpringerLink 等数据库收录! |
|