I believe theta = sqrt(1/(sigma_u)^2)
Scott
----- Original Message -----
From: [email protected]
Date: Tuesday, May 4, 2004 6:18 am
Subject: st: SFA - variance parameter
> Dear Stata-users,
>
> Currently I am doing a Stochastic Frontier Analysis in Stata, as
> opposed to
> Limdep which I used in the past. The problem which arises with this
> transition to Stata is that I am not able to derive in what way the
> parameter 'theta' (one of the variance parameters) is calculated
> by the
> Limdep package.
>
> As this parameter is used in the formula by which the percentage
> of the
> error term assigned to inefficiency can be calculated, I would
> like to know
> how I can derive this term from the results of SFA by means of Stata.
>
>
> Limdep results:
> Variances: Sigma-squared(v)= .01407
> Sigma-squared(u)= .01068
> Stochastic Cost Frontier, e=v+u.
> +---------+--------------+----------------+--------+---------+-----
> -----+
> |Variable | Coefficient | Standard Error |b/St.Er.|P[|Z|>z] |
> Mean of X|
> +---------+--------------+----------------+--------+---------+-----
> -----+
> Primary Index Equation for Model
> Constant -2.398824811 .42999158 -5.579 .0000
> LNLOCSW .4676510927 .10859080 4.307 .0000
> 16.842014 LNNOLINE .4787993242 .11729499 4.082
> .0000 14.355184
> LNDUCT -.6791673073E-01 .36352854E-01 -1.868 .0617 -
> 5.5786894 Variance parameters for compound error
> Theta 9.676302022 5.3867624 1.796 .0724
> Sigmav .1186164475 .32301046E-01 3.672 .0002
>
> Stata results:
> -------------------------------------------------------------------
> -----------
> lnb1hca | Coef. Std. Err. z P>|z| [95% Conf.
> Interval]
> -------------+-----------------------------------------------------
> -----------
> lnlocsw | .4676511 .0813805 5.75 0.000 .3081482
> .627154
> lnnolines | .4787993 .0855751 5.60 0.000 .3110753
> .6465234
> lnduct | -.0679167 .0279463 -2.43 0.015 -.1226904
> -.0131429
> _cons | -2.398824 .3871222 -6.20 0.000 -3.15757
> -1.640079
> -------------+-----------------------------------------------------
> -----------
> /lnsig2v | -4.263715 .5266097 -8.10 0.000 -5.295852
> -3.231579
> /lnsig2u | -4.539367 .944926 -4.80 0.000 -6.391388
> -2.687346
> -------------+-----------------------------------------------------
> -----------
> sigma_v | .1186167 .0312324 .0707979
> .1987337
> sigma_u | .1033449 .0488266 .0409381
> .2608857
> sigma2 | .0247501 .0061499 .0126966
> .0368036
> lambda | .8712505 .0760584 .7221787
> 1.020322
> -------------------------------------------------------------------
> -----------
> Likelihood-ratio test of sigma_u=0: chibar2(01) = 1.06
> Prob>=chibar2 =
> 0.152
>
>
>
> Thanks and regards,
>
> Erik Brouwer
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
