2
$\begingroup$

I am planning on running a 3 equation simultaneous equation model where each of the dependent variables depend on each other (i.e. Y1 is based on Y2 and Y3; Y2 is based on Y1 and Y3; Y3 is based on Y1 and Y2).

I came across the method of 3SLS but I can't find much information on its applicability.

When is it appropriate to use 3SLS instead of 2SLS (what are the assumptions underlying 3SLS)? And what are the benefits and drawbacks of using 3SLS over 2SLS? If you could explain in very basic terms, that would be great!

$\endgroup$
2
$\begingroup$

Three stage least squares have some efficiency gains with respect to two stage least squares but it might not always be applicable. It's really a huge field that depends, among other things, on the quality of your instruments.

Two good applied resources are:
1. This paper on estimating oil demand and supply
2. The introductory paper for the systemfit package which explains well the kind of tests you want to run to discriminate between 3SLS and 2SLS

Both seem targeted to the economics graduate but that's just a consequence of the field's need to deal with simultaneity and shouldn't discourage you.

$\endgroup$
  • $\begingroup$ I see. So the benefit is smaller standard errors. $\endgroup$ – F4N May 3 '16 at 21:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.