I cannot find the right procedure to do the job

[applemonster]

Hi all the genius,

I need to run a two-way fixed effects regression on a panel dataset and enroll instrumental variables (IVs) and 2SLS.

I thought I had two options but it seems neither of them work well.

Option 1: The syslin procedure.

This procedure seems to have a good package to deal with endogeneity, IV and 2SLS but be not compatible with panel data.

Option 2: The panel procedure

I use this procedure a lot. It works well with panel data and has a statement named "INSTRUMENTS". However, according to the documentation, I find this statement is deeply binded with the moment condition equations of the dynamic panel estimator, which is not my case.

So can someone suggest me a good procedure to handle my case?

Thanks.

[sbxkoenk · Accepted solution]

Hello,

I have moved your question to this "Econometrics" board.

Over here, you have a better chance for an appropriate answer.

Do you have SAS VIYA Econometrics?

If yes, you could try PROC CPANEL (instead of the PROC PANEL from SAS/ETS).

Cheers,

Koen

[Ksharp]

Searching it at

https://support.sas.com/kb/30/333.html