instrumental variable help

[econ]

hi there,

i have searched through other threads but can't manage to find any answers to the problem i'm having. i'm not good with econometrics, and i am unable to do something extremely simple using eviews.

i specify the equation, having selected 2SLS:

y c d q d*q

(where d is a dummy and d*q is the interactive term).

i then need to instrument for q and d*q, as both are endogenous and so use r to do so. I have been told that i use the instrument list of:

r d*r

when i do so, i get an error message coming up saying that there are insufficient instruments. even if i have other instruments (which i can add if need be), the message still comes up. i have used the constant, as well as including my dummy in the instrument (not sure if this is right) and it still won't work. the only time it is successful is if i move q d*q into the instrument list as well. am i doing something stupidly wrong and i haven't realised?

any help would be much appreciated, thanks.

[EViews Glenn]

You need d as well...

[startz]

You need four instruments, including the constant. Make sure they are all exogenous.

[EViews Glenn]

c is added automatically unless you tell EViews not to...

[startz]

Yeah, yeah, yeah. So add three instruments plus the constant that EViews include.

[econ]

thanks for your help. everything i tried didn't work, so have had to do things a bit differently. thanks for the quick replies.

[ramadhan23]

Screen Shot 2014-10-16 at 7.50.26 AM.png (29.32 KiB) Viewed 6732 times

Screen Shot 2014-10-28 at 7.34.17 AM.png (38.38 KiB) Viewed 6732 times

i want to ask how to use instrument variables into a model? i am working on a thesis about international capm and need to incorporate instrument variables to the estimation model as attached. as you can see the lambda m is the instrument variables, In my research i use 5 instrument variables (constant included) , can anybody explain how its done in eviews? thanks