Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Re: Frontier analysis

From   "Scott Merryman" <>
To   <>
Subject   st: Re: Frontier analysis
Date   Thu, 5 Feb 2004 09:08:48 -0600


Technical efficiency can be calculated 3 ways:

    TE = exp(-E(u|e)]

    TE = exp(-M(u|e)]

    TE = E(exp(-u|e)]

These will not equal to since exp[-E(u|e)] != E[exp(-u|e)]

However, the last technical efficiency measure (by Battese and Coelli (1988)
"Prediction of Firm-Level Technical Efficiencies with a Generalized Frontier
Production Function and Panel Data"  Journal of Econometrics, 38) is preferred,
particularly when u is not close to zero and it is consistent with the
definition of technical efficiency, see Kumbhakar and Lovell (2000) Stochastic
Frontier Analysis.


----- Original Message ----- 
From: <>
To: <>
Sent: Thursday, February 05, 2004 7:59 AM
Subject: st: Frontier analysis

> Dear all,
> First of all, thanks for all your advices, they were very useful to me.
> After having solved the problems, Stata is now generating results
> comparable with those from Limdep (same data, same coefficients, same
> standard errors (by approx.). Still a final question lasting, my assumption
> is that the inefficiency value is given by 'u' in Stata. The reason for
> this is that these values are the same as given by Limdep (via the command
> eff, this is E[u/e]). Stata's reference guide gives the following
> definition on u:
> u gives estimates of minus the natural log of the technical efficiency via
> E[u_i|e_i]
> te gives estimates of the technical efficiency via E[exp(-su_i)|e_i]
> Can you tell me which is the correct variable for the inefficiency
> component? u or te?
> thanks and regards

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index