# [Limdep Nlogit List] Retrieving and interpreting individual effects in panel regression

Mattia Landoni mattia.landoni at gmail.com
Thu Apr 3 13:26:25 EST 2008

```Hi,

I am trying to run a stochastic frontier model to determine
revenue-generating efficiency (limdep 8). Suppose I fit a linear (ols)
random effects regression model [I know there is also the frontier model,
but bear with me for now. Also bear with me if what I am asking for is
written somewhere in the manual, it's my second day with LIMDEP and there's
a limit to how much I can absorb... :-)]:

Y = Xb + e + u

Where "u" is the individual disturbance and "e" the residual.

REGRESS ;Lhs=X_REVENU ;Rhs=XLIST ;Pds=NI ;Random Effects ;res=E ;keep=XB\$

1) To create a variable containing the individual effects, will the
following work?
CREATE ;U = X_REVENU - XB - E\$

1.1) Is there a way to create the variable U directly within the REGRESS
command (analogous to ";res=E")?
1.2) Will it work also in a Fixed Effects model?

2) Now let's go back to the frontier specific estimation command. The
frontier model is

Y = Xb + v - u (revenue)
Y = Xb + v + u (cost, adding the ;cost option)

FRONTIER ;Lhs=X_REVENU ;Rhs=XLIST ;Pds=NI ;eff=U ;keep=XB\$
Note the "eff=U" term. My variables are in logarithm form. Please choose
"True" or "False" for the following statements :-)

- Revenue model: large U corresponds to large inefficiency (T/F)
- Cost model: large U corresponds to large inefficiency (T/F)
- To obtain an inefficiency coefficient for the individual firm: CREATE
;INEFF = exp(U)\$ (T/F)

Thanks for the attention!

Mattia

--
Mattia Landoni
1201 S Eads St Apt 417
Arlington, VA 22202-2837
USA
Greenwich -5 hours

Office: +1 202 62 35922
Cell: +1 202 492 3404
Home: +1 360 968 1684

Govern a great country as you would fry a small fish: do not poke at it too
much.
-- Lao Tzu

```