Title :

From partial derivatives of DEA frontiers to marginal products, marginal rates of substitution, and returns to scale

Author(s) :

Ouellette, Pierre [Auteur]
Vigeant, Stephane [Auteur]
Lille économie management - UMR 9221 [LEM]

Journal title :

European Journal of Operational Research

Pages :

880--887

Publisher :

Elsevier

Publication date :

2016-09

ISSN :

0377-2217

English keyword(s) :

Data envelopment analysis
Marginal products
Transformation function
First derivatives

HAL domain(s) :

Sciences de l'Homme et Société/Economies et finances

English abstract : [en]

The characterization of a technology, from an economic point of view, often uses the first derivatives of either the transformation or the production function. In a parametric setting, these quantities are readily available ...
Show more >The characterization of a technology, from an economic point of view, often uses the first derivatives of either the transformation or the production function. In a parametric setting, these quantities are readily available as they can be easily deduced from the first derivatives of the specified function. In the standard framework of data envelopment analysis (DEA) models these quantities are not so easily obtained. The difficulty resides in the fact that marginal changes of inputs and outputs might affect the position of the frontier itself while the calculation of first derivatives for economic purposes assumes that the frontier is held constant. We develop here a procedure to recover first derivatives of transformation functions in DEA models and we show how we can evacuate the problem of the (marginal) shift of the frontier. We show how the knowledge of the first derivatives of the frontier estimated by DEA can be used to deduce and compute marginal products, marginal rates of substitution, and returns to scale for each decision making unit (DMU) in the sample.Show less >

Language :

Anglais

Popular science :

Non

Collections :

• Lille Économie Management (LEM) - UMR 9221

Source :

Harvested from HAL