Re: [Pen-l] Re: Question on overdetermination

["Jim Devine" <jdevine03@xxxxxxxxx>]

Sabri Oncu wrote:
> By the way Jim, I was expecting a response such as use SUR (Seemingly
> Unrelated Regressions) or 2SLS (Two Stage Least Squares) or maybe 3SLS
> if you want to make the model even nicer looking, find some IVs
> (Instrumental Variables), use many other so-called "controls
> variables"  and the like. Your response shows that you have not been
> dealing with econometrics for the last 30 years, as you said.

FWIW, I am familiar with the above. My attitude, however, is that
econometrics can't "prove" anything. Instead, if your model can't be
"verified" econometrically, you should rethink it. All 'metrics can do
is to provide extra evidence of one's case or to give
seat-of-the-pants estimates (such as "what is the trend of variable
X?") Thus, I saw no need to refer to instrumental variables, etc.

BTW, it's better to respond to what people say rather than to what you
"expect" them to say.

> But, suppose that you did everything there is to do and reached to the
> conclusion that yeah, both parties were responding to each other and
> more.
>
> Then what happens?

I don't understand this. what "parties" are you referring to?
-- 
Jim Devine / "Segui il tuo corso, e lascia dir le genti." (Go your own
way and let people talk.) -- Karl, paraphrasing Dante.
_______________________________________________
pen-l mailing list
pen-l@xxxxxxxxxxxxxxxxxx
https://lists.csuchico.edu/mailman/listinfo/pen-l