Panel data regression with random effects

[hkgejd]

Hi everyone

I want to do a regression analysis with random effects on a panel dataset. I know it is possible with the help of the 'proc panel'-procedure, but unfortunately I don't have that package. Is it possible to do this without the 'proc panel'-procedure?

To those who are familiar with R, I want to do something similar with:

"data("Produc", package = "plm")
zz <- plm(log(gsp) ~ log(pcap) + log(pc) + log(emp) + unemp,
data = Produc, index = c("state","year"))
summary(zz)"

Thanks in advance 😀

[sbxkoenk]

Hello,

If you do NOT have the SAS/ETS PROC PANEL ,
You won't have the SAS Econometrics PROC CPANEL either , I believe.

Then you can work with PROC MIXED or PROC GLIMMIX or PROC NLMIXED (all SAS/STAT).
The results will be the "same" , ... with slight variations because of different estimation algorithms!

Let us know if you cannot figure it out.

Cheers,
Koen

[hkgejd]

Hi sbxkoenk 

Thanks for your help! 😀

You are right, I don't have the SAS/ETS-package.

How would you "translate" the following PROC PANEL to one of the mentioned procedures?

proc panel data=data;

id id t;

model lwage = exp exp2 wks ed /ranone;

run;

[sbxkoenk]

Hello,

Here are two SAS usage notes about using PROC MIXED for panel data

(well, for the 1st one PROC MIXED is not the primary focus, but it is covered) :

Usage Note 22114: Analyzing unbalanced panel data using PROC PANEL or PROC TSCSREG
https://support.sas.com/kb/22/114.html

PROC MIXED can be used as an alternative to the Parks method as discussed in the following note:
Estimating a model using the Parks method with unbalanced panel data
https://support.sas.com/kb/22/115.html

Remark : The estimates from PROC MIXED will not be identical to those from PROC PANEL since they use different methods — PROC PANEL uses Seely's method while PROC MIXED uses REML.

Koen