[Andrew Musau]

Write down the expression of the elasticity and then use lincom or margins with the expression option.

[Fadi Ansar]

Thankyou so much Andrew Musau .Can you please help me with the code I can use if I need to estiamte for each bank and year in stata

[Andrew Musau]

Look at Kumbhakar's book A Practitioner's Guide to Stochastic Frontier Analysis Using Stata for illustrative examples. You need to know how the estimated coefficients relate to your cost function parameters and then calculate the derivatives needed to be plugged into the elasticity formula. Once you have this expression, use lincom or margins to calculate the elasticities and the corresponding standard errors.

[Anushiya Thanapalan]

Hi,
I am using STATA 15 for my analysis. I am estimating profit efficiency using Cobb Douglas function.I am wondering if anyone can help me in estimating technical efficiency based on farm-specific characteristics? I mean the marginal effect of the technical inefficiency variables on E(u).
Thank you in advance.
Anushiya

[Lindikaya Myeki]

Hi Andrew, see attached my Stata output. i worried about the mean for Central and Western Africa, they seem to be too high. What could be wrong????

[Andrew Musau]

The technology gap ratio should be between 0 and 1. Check your estimation.

[Alba Estevez Bauluz]

You should have a look at Huang et al. (J Prod Anal 42:241–254, 2014) "A new approach to estimating the metafrontier production function based on a stochastic frontier framework" which criticizes Battese et al. (J Prod Anal 21:91–103, 2004) and O’Donnell et al. (Empir Econ 34:231–255, 2008) procedure of using linear (quadratic) programming techniques in the second step. Their suggested procedure is easily implemented in Stata. Given the output "y", 3 inputs "x1", "x2", and "x3" and environmental variables z1-z8 and assuming a translog functional form, first generate dummies for the groups reflecting different technology possibility sets, e.g., group1, group2, group3. The ensuing syntax is

frontier lny c.lnx1##c.lnx1 c.lnx2##c.lnx2 c.lnx3##c.lnx3 c.lnx1#c.lnx2 c.lnx1#c.lnx3 ///
c.lnx2#c.lnx3 if group1, distribution(tnormal) cm(z1-z4) ///
predict te1 if group1, te
predict xb1 if group1, xb

This is the Battese and Coelli's (1992) frontier model for panel data implemented by Stata's -frontier- command. Repeat the same for the rest of the groups. In the second step, use the predicted values as the outcome for the metafrontier equation

gen pred= .
replace pred= xb1 if group1
replace pred= xb2 if group2
replace pred= xb3 if group3


frontier pred c.lnx1##c.lnx1 c.lnx2##c.lnx2 c.lnx3##c.lnx3 c.lnx1#c.lnx2 c.lnx1#c.lnx3 c.lnx2#c.lnx3, distribution(tnormal) cm(z5-z8)
predict te_meta, te

Note that the environmental variables in the second stage differ from those in the first. Read Huang et al. for reasons why you need to include environmental variables.

https://link.springer.com/article/10...123-014-0402-2

Dear Andrew,
I'm trying to estimate a stochastic metafrontier model with two groups in Stata. I have followed the guidelines in the answer to John Ng`ombe and obtained the technical efficiency of the first stage for each group (predict te1 if gruop1, te ...), as well as the technical efficiency of the metafrontier in the second stage ( predict te_meta, te). I understand that the technical efficiency of the metafrontier must be unique for all groups.
However, after reviewing the articles by Huang (2014) and Ng`ombe (2017), I have observed that the results tables show a TE, TGR and MTE for each group. This leads me to wonder: should a metafrontier be calculated per group?
I thank you in advance for your attention and help.
Sincerely,
Alba Estevez Bauluz

[Andrew Musau]

Once you estimate the meta-frontier equation, you can calculate efficiency scores for each group separately.

[Alba Estevez Bauluz]

Once you estimate the meta-frontier equation, you can calculate efficiency scores for each group separately.

Thank you very much for answering so quickly! Could I ask you another question? How can I calculate the efficiency scores for each group separately?

[Andrew Musau]

Whatever is reported are just mean values. So you can take the mean over all observations or restrict these to subsamples. Below, consider the definition of a large firm as one with size> 2000. I can estimate a model and predict efficiency for all firms as well as large and small firms.

webuse frontier1, clear
frontier lnoutput lnlabor lncapital, vhet(size)
predict te_vhet, te
gen largefirms= size> 2000
lab def size 1 "Large firms" 0 "Small firms"
lab values largefirms size
*ALL FIRMS
sum te_vhet
*LARGE, SMALL AND ALL FIRMS
table largefirms, stat(mean te_vhet)

Res.:

. *ALL FIRMS

.
. sum te_vhet

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     te_vhet |        756    .2691245    .2071925   .0001571   .7643759

.
. *LARGE, SMALL AND ALL FIRMS

.
. table largefirms, stat(mean te_vhet)

-------------------------
              |      Mean
--------------+----------
largefirms    |          
  Small firms |  .2650408
  Large firms |   .272759
  Total       |  .2691245
-------------------------

[Alba Estevez Bauluz]

Whatever is reported are just mean values. So you can take the mean over all observations or restrict these to subsamples. Below, consider the definition of a large firm as one with size> 2000. I can estimate a model and predict efficiency for all firms as well as large and small firms.

webuse frontier1, clear
frontier lnoutput lnlabor lncapital, vhet(size)
predict te_vhet, te
gen largefirms= size> 2000
lab def size 1 "Large firms" 0 "Small firms"
lab values largefirms size
*ALL FIRMS
sum te_vhet
*LARGE, SMALL AND ALL FIRMS
table largefirms, stat(mean te_vhet)

Res.:

. *ALL FIRMS

.
. sum te_vhet

Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
te_vhet | 756  .2691245 .2071925 .0001571 .7643759

.
. *LARGE, SMALL AND ALL FIRMS

.
. table largefirms, stat(mean te_vhet)

-------------------------
| Mean
--------------+----------
largefirms |
Small firms | .2650408
Large firms | .272759
Total | .2691245
-------------------------

Thank you very much Andrew! I really appreciate your help!