Date:         Tue, 20 Jul 2004 17:21:45 +0200
Reply-To:     Adelina Gschwandtner <adelina.gschwandtner@UNIVIE.AC.AT>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Adelina Gschwandtner <adelina.gschwandtner@UNIVIE.AC.AT>
Subject:      Panel Estimation
Content-Type: text/plain; charset=ISO-8859-15; format=flowed

Hello everybody!

I have used PROC NLMIXED for a panel of 88 companies, each having 23
observations. So in sum I have 2024 observations. I have used the
following code:

proc nlmixed data=in.stat50725u;
eta = a1 + b1*LP1;
model profmed ~ normal(eta, v);
estimate 'lrpp' a1/(1-b1);
run;

Now a colleague tells me that this might be wrong because in panels
there might be problems with OLS and I should use the ARELLANO-BOND
method?!!

Is this correct? Is it wrong to use PROC NLMIXED in panels???

If this should be true does anyone know a method for panel estimation or
a syntax for the Arrelano-Bond method?

Many thanks for any hint,
Adelina.

--
?º°`°º?ø,??,ø?º°`°º?ø,??,ø?º°`°º?ø,??,ø?º°`°º?ø?º°`°º?ø,??,ø

ADELINA GSCHWANDTNER DR.

University of Vienna
Department of Economics
BWZ, Bruennerstr.72                 Tel:(00431) 4277 37480
A-1210 Vienna                       Fax:(00431) 4277 37498

Adelina.Gschwandtner@univie.ac.at

http://www.univie.ac.at/Wirtschaftswissenschaften/gschwand/

?º°`°º?ø,??,ø?º°`°º?ø,??,ø?º°`°º?ø,??,ø?º°`°º?ø?º°`°º?ø,??,ø