Instrument Variable

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

Apple Tan
Posts: 29
Joined: Sun Apr 12, 2015 5:18 pm

Instrument Variable

Postby Apple Tan » Mon Jun 22, 2015 12:18 am

Hi

Using Eviews7. I want to do a TSLS regression. I read from the Eviews 7 user guide II that the instrument list show be as seen in image.

However, is log(perinc) one of the instruments? It is the variable we are instrumenting for, isn't it?


Thank you
Attachments
Untitled2.jpg
Untitled2.jpg (18.19 KiB) Viewed 3575 times

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Instrument Variable

Postby startz » Mon Jun 22, 2015 6:02 am

The version 9 manual says "The endogenous variable is the log of real after tax price per pack LOG(RAVGPRC)."

Apple Tan
Posts: 29
Joined: Sun Apr 12, 2015 5:18 pm

Re: Instrument Variable

Postby Apple Tan » Mon Jun 22, 2015 5:11 pm

Thanks for the reply. i missed out that line.

However, if my specification has an interaction with the endogenous variable, than how should the instrument list be?

model specification

log (A) = c + log (B) + log (C) + log (B) * log(D)
where log (B) is the endogenous variable.


Thank you

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Instrument Variable

Postby startz » Mon Jun 22, 2015 5:43 pm

You need the same number of instruments as you have variables, including the interaction. So you need four or more. None of the instruments can include B.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests