Stochastic Nonparametric Envelopment of Data: Combining Virtues of Sfa and DEA in a Unified Framework

MTT Discussion Paper No. 3/2006

53 Pages Posted: 2 Jun 2006  

Timo Kuosmanen

Aalto University School of Business

Date Written: June 2006

Abstract

The literature of productive efficiency analysis is divided into two main branches: the parametric Stochastic Frontier Analysis (SFA) and nonparametric Data Envelopment Analysis (DEA). This paper attempts to combine the virtues of both approaches in a unified framework. We follow the SFA literature and introduce a stochastic component decomposed into idiosyncratic error and technical inefficiency components imposing the standard SFA assumptions. In contrast to the SFA, we do not make any prior assumptions about the functional form of the deterministic production function. In this respect, we follow the nonparametric route of DEA that only imposes free disposability, convexity, and some specification of returns to scale. From the postulated class of production functions, the proposed method identifies the production function with the best empirical fit to the data. The resulting function will always take a piece-wise linear form analogous to the DEA frontiers. We discuss the practical implementation of the method and illustrate its potential by means empirical examples.

Keywords: frontier estimation, productive efficiency analysis, stochastic frontier analysis (SFA), data envelopment analysis (DEA), nonparametric regression

JEL Classification: C14, C51,C61, D24

Suggested Citation

Kuosmanen, Timo, Stochastic Nonparametric Envelopment of Data: Combining Virtues of Sfa and DEA in a Unified Framework (June 2006). MTT Discussion Paper No. 3/2006. Available at SSRN: https://ssrn.com/abstract=905758 or http://dx.doi.org/10.2139/ssrn.905758

Timo Kuosmanen (Contact Author)

Aalto University School of Business ( email )

P.O. Box 1210
Runeberginkatu 22-24
Helsinki, Finland 00101
Finland

HOME PAGE: http://www.aalto.fi

Register to save articles to
your library

Register

Paper statistics

Downloads
887
rank
23,657
Abstract Views
2,507
PlumX