The RATS Software Forum

Hello,

I would like to apply the pooled mean group estimator as proposed by Pesaran, Shin and Smith(1999), "Pooled Mean Group Estimation of Dynamic Heterogeneous Panels", JASA, vol 94, no. 446, pp 621-634.
How have I to modify the pesaranshinsmithjasa.zip file to apply it to 5 instead of 3 variables such as in their study?
I already tried to change the names of the variables, the data file, the amount of observation...
But I receive following error message at the line "end do iters":
"## SR10. Missing Values And/Or SMPL Options Leave No Usable Data Points
The Error Occurred At Location 0261 of loop/block
Line 8 of loop/block".
Please find attached my (wrongly) modified version of the RATS procedure pesaranshinsmithjasa, the original one as well as my data set.
Thank you very much in advance for any reply.

Regards,
Frauke

First, you have your calendar mis-specified. It should be:

cal(panelobs=92,q) 1983:4

(the :4 means 4th period, but doesn't set the frequency per year). As a result, your data set doesn't get read properly.

Also, there's an initialization for theta which needs to have the proper dimensions, such as:

compute [vector] theta=||1.0,0.0,0.0,0.0||

You still had it as the 2-vector from the original program.

Thank you very much!
After correcting for that and some other mistakes the program runs.
However, the results are very unusal since the income elasticity should be around unity and the interest rate semi-elasticity negative.
In contrast, the mean group estimator shows exactly those results.
Could you please check whether there is still something wrong in my code?
That would be great!

Regards,
Frauke