[R] System/SUR equations and panel

[Arne Henningsen]

Hi Alok

On 3 October 2012 17:28, Alok K Bohara, PhD <bohara at unm.edu> wrote:
> I was wondering if there is any R package to estimate a two-or three
> equation system in a panel setting.
>
> Say, I have 100 firms and their expenditure on wind generated power and
> solar-generated power.
>
> S_i = X1_i*b1 + u1_i
> W_i = X2_i* b2 + u2_i
>
> (X1 and X2 need not be the same.. but for simplicity, they could be assumed
> to be similar)
>
> Now, I have this firm data set for 10 years...
>
> I looked at systemfit package and its "investment"  example and the way it
> uses plm package was not quite clear.  Any help will be much appreciated.

Unfortunately, systemfit cannot (yet) apply panel data techniques
(e.g. random effects or fixed effects) to estimate a system of
equations. However, you can manually estimate a fixed effects model if
you add dummy variables for individuals and/or time periods or
estimate the model with "within transformed" data.

Best wishes,
Arne

-- 
Arne Henningsen
http://www.arne-henningsen.name