Erkie Asmare 1 and Andualem Begashaw

Types and Natures of Nonparametric Methods of 
Efficiency Analysis
Under nonparametric methods of efficiency, there are two ap
-
proaches namely; Data Envelopment Analysis (DEA) and Free Dis
-
posal Hull (FDH) methods [11]. However, in the field of research the 
most commonly practiced approach is data envelopment analysis. 
As an original non-parametric approach of efficiency measurement, 
data envelopment analysis has been introduced by Charnes [15], 
Cooper and Rhodes (CCR) in 1978 [11]. It was originally developed 
to measure the performance of various non profit organizations, 
such as educational and medical institutions, which were highly 
resistant to traditional performance measurement techniques due 
to the complex and often unknown relations of multiple inputs and 
outputs and non-comparable factors that had to be taken into ac-
count [13]. 
In addition, DEA was first developed in public sector analysis 
of technical efficiency, where price information is not available or 
nor reliable [16]. Moreover Kuosmanen et al. [1] argued that pub-
lic providers have objectives and constraints different from those 
of private providers and so the only common ground on which to 
compare their performance is on the basis of their technical effi
-

---

3/7
How to cite this article:
Erkie A, Andualem B. Review on Parametric and Nonparametric Methods of Efficiency Analysis. Open Acc Biostat Bioinform
. 2(2). 
OABB.000534. 2018. DOI:
10.31031/OABB.2018.02.000534
Open Acc Biostat Bioinform
Copyright © 
Erkie Asmare Beyene
Volume 2 - Issue - 2
ciency. However; in recent years, DEA has been successfully applied 
in measuring the efficiency of both profit and non-profit organiza
-
tions, such as the effectiveness of regional development policies 
[13]. In contrast to the econometric approaches, non-parametric 
methods are based on the hypothesis that the efficiency frontier is 
generated from the empirical results’ of the most efficient decision 
making unit (DMU’s) or from the benchmarks [11]. Nonparametric 
inference has gained popularity because of several reasons: 
a. 
The computations are easy to estimate.
b. 
The data need not be measured quantitatively but could 
be in a qualitative format.
c. 
The data could also be in an ordinal ranking. 
d. 
Does not have as many restrictive assumptions as para-
metric statistical inference.
On the other hand, nonparametric inference is deficient in the 
sense that it does not utilize all the information in the sample and, 
thus, will be less efficient than parametric inference [17]. Since the 
DEA model is non-stochastic, noise is reported as inefficiency and 
makes the mean technical efficiency lower [17].