Estimating the hyperbolic distance function: A directional distance function approach

被引:25
|
作者
Faere, Rolf [2 ,3 ]
Margaritis, Dimitris [1 ]
Rouse, Paul [1 ]
Roshdi, Israfil [1 ,4 ]
机构
[1] Univ Auckland, Sch Business, Dept Accounting & Finance, Auckland 1, New Zealand
[2] Oregon State Univ, Dept Econ, Corvallis, OR 97331 USA
[3] Oregon State Univ, Dept Agr & Resource Econ, Corvallis, OR 97331 USA
[4] Islamic Azad Univ, Dept Math, Semnan Branch Islam, Semnan, Iran
关键词
Efficiency measurement; Data envelopment analysis; Hyperbolic distance function; Directional distance function; DATA ENVELOPMENT ANALYSIS; ENVIRONMENTAL EFFICIENCY; NONPARAMETRIC APPROACH; QUANTILE ESTIMATION; ROBUST VERSIONS; DEA MODELS; SCALE; OECD; EMISSIONS; RETURNS;
D O I
10.1016/j.ejor.2016.03.045
中图分类号
C93 [管理学];
学科分类号
12 ; 1201 ; 1202 ; 120202 ;
摘要
Fare, Grosskopf, and Lovell (1985) merged Farrell's input and output oriented technical efficiency measures into a new graph-type approach known as hyperbolic distance function (HDF). In spite of its appealing special structure in allowing for the simultaneous and equiproportionate reduction in inputs and increase in outputs, HDF is a non-linear optimization and it is hard to solve particularly when dealing with technologies operating under variable returns to scale. By connecting HDF to the directional distance function, we propose a linear programming based procedure for estimating the exact value of HDF within the non-parametric framework of data envelopment analysis. We illustrate the computational effectiveness of the algorithm on several real-world and simulated data sets, generating the optimal value of HDF through generally solving at most two linear programs. Moreover, our approach has several desirable properties such as: (1) introducing a computational dual formulation for the HDF and providing an economic interpretation in terms of shadow prices; (2) being readily adaptable to measure hyperbolic oriented super-efficiency; and (3) being flexible to deal with HDF-based efficiency measures on environmental technologies. (C) 2016 Elsevier B.V. All rights reserved.
引用
收藏
页码:312 / 319
页数:8
相关论文
共 50 条