Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: ivreg and indicator variables


From   "Millimet, Daniel" <millimet@post.cis.smu.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: ivreg and indicator variables
Date   Wed, 25 Sep 2002 09:41:07 -0500

you need two instruments.  if the indicator x2 is exogenous, then the second instrument should be your original instrument also interacted with x2.  now you can use ivreg.
 
dann

	-----Original Message----- 
	From: faparici@gmu.edu [mailto:faparici@gmu.edu] 
	Sent: Wed 9/25/2002 9:36 AM 
	To: statalist@hsphsun2.harvard.edu 
	Cc: 
	Subject: st: ivreg and indicator variables
	
	

	Dear stata users,
	
	I am estimating a model
	
	y = a + b1x1 + b2x1x2 + b3x3 + e
	
	x1 and x3 are continuous variables. x2 is an indicator variable taking
	the value 0 or 1. x1 is correlated with the error term e, thus I want
	to estimate this model via two stage least squares. I have an
	instrument z which is correlated with x1.  How do I estimate this model
	in Stata?
	
	this does not work since I only have one instrument:
	
	ivreg y (x1 x1x2 = z1) x3
	
	but is it correct to estimate the first stage, then get the predicted
	value x2hat and to multiply the predicted value with x2, and run a
	regression with
	
	y = a + b1x1hat + b2x1hatx2 + b3x3 + e
	
	If this is correct, how do I get the proper standard errors?
	
	Thanks in advance for your help,
	
	-Javier
	
	*
	*   For searches and help try:
	*   http://www.stata.com/support/faqs/res/findit.html
	*   http://www.stata.com/support/statalist/faq
	*   http://www.ats.ucla.edu/stat/stata/
	

<<winmail.dat>>




© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index