Research World

Data  Envelopment  Analysis  (DEA)  essentially  provides a mathematical programming method for estimating optimal production frontiers and evaluating the relative efficiency of  different  DMUs  having  smellier  type’s  inputs  and  outputs.  The  basis  in  the  DEA method is the measuring of the efficiency of a decision unit in comparison with another one  operating  in  the  same  market.  The  constraint  in  this  analysis  is  a  necessity  for  all decision  making  units  to  be  above  or  below  the  efficiency  limit.  As  a  result,  while efficient units take the value 1, the value of the  inefficient ones is smaller than 1. The difference between 1 and the efficiency valve indicates that the same amount of output will  be  obtained  in  proportion  to  the  difference  with  less  input  (Ulucan,  2000).   DEA assigns  weights  to  the  inputs  and  outputs  of  a  DMU  that  give  it  the  best  possible efficiency. It thus arrives at a weighting of the relative importance of the input and output variables that  reflect the  emphasis that appears  to have been  placed on  them for  that particular DMU. For the estimation of efficiency score for each DMU in DEA, two type of models used which are on the basis of the production processes is CCR-model for CRS 
production processes and BCC-model for VRS production processes with two different types of orientations namely input orientation and output orientation.

DEA  technique is  computationally simple,  easy and  has the  advantage that  it can  be implemented without having any knowledge about the algebraic form of the relationship between outputs and inputs. In other words, we can estimate the efficient frontier in DEA without knowing whether an output is a linear, quadratic, and exponential or some other type of function of  inputs.  That  is  why  DEA  is  called  non-parametric  technique  in  a frontier  analysis.  There  is  an  alternative  method  for  frontier  estimation  that  assumes  a given function of output with inputs, called Stochastic Frontier Analysis (SFA). In SFA, we  need  econometric  techniques  that  can  use  to  estimate  the  unknown  parametric involved  in  the  function,  in  order  to  obtain  the  efficient  frontier.  Due  to  these requirements of SFA becomes more computationally demanding than DEA in the field of economics. This approach was used by Ainner and Chu (1968) and who consider a CobbDouglas production frontier.