Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: Interactions using -xtivreg- Instrumental Variables estimation


From   David Torres <writeon4truth2@msn.com>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Interactions using -xtivreg- Instrumental Variables estimation
Date   Wed, 31 Jul 2013 23:41:27 -0400

Thanks so much, Vidhya.  It's nice to know I was on the right track.
D


----------------------------------------
> Date: Wed, 31 Jul 2013 19:32:51 -0400
> Subject: Re: st: Interactions using -xtivreg- Instrumental Variables estimation
> From: vidhyahere@gmail.com
> To: statalist@hsphsun2.harvard.edu
>
> Hi David,
>
> That is correct. You typically include your interact your instruments
> with the same variables as the endogenous variables are interacted
> with. Unless of course you have other non-interacted instruments as
> well, which can be added to the list.
>
> I think thought you need to manually generate these interacted
> variables (both endogenous and instruments) and include them. I do not
> think STATA accepts interaction operators in its list of endogenous
> regressors or instruments (may be newer version do, I may be wrong).
>
> What is forbidden in this case though is "the forbidden regression" -
> see Wooldridge (2000), Econometric Analysis of Cross Section and Panel
> Data, section 9.5, esp. pp. 236-7 and a similar earlier question on
> statalist - http://www.stata.com/statalist/archive/2003-11/msg00795.html.
> There are many more similar questions like this on statalist and
> useful discussions surrounding this, as well that can help you
>
> Best,
> Vidhya
>
>
> On Wed, Jul 31, 2013 at 7:11 PM, David Torres <writeon4truth2@msn.com> wrote:
>>
>> I'm still hoping someone can answer the question below, which I sent yesterday.
>>
>>>
>>> Hello, all,
>>>
>>> So I'm using an instrumental variables approach with two-level data using Stata's -xtivreg- command. My single IV relates to a level-one time-varying variable. I'd like to include interactions between the level one variable and dummies at level two denoting sex and race/ethnicity.
>>>
>>> How is this done? Do I produce interactions between the instrument and my dummies, as well as between the variable being instrumented for and my dummies, and then simply insert everything in my model like this:
>>>
>>> xtivreg mathsc year male black hispan othrace (var1 var1*male var1*black var1*hispan var1*othrace = instrument instrument*male instrument*black instrument*hispan instrument*othrace), re
>>>
>>> or should I be setting the model up differently?
>>>
>>> Thanks in advance for your help.
>>>
>>>
>>>
>>> *
>>> * For searches and help try:
>>> * http://www.stata.com/help.cgi?search
>>> * http://www.stata.com/support/faqs/resources/statalist-faq/
>>> * http://www.ats.ucla.edu/stat/stata/
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/faqs/resources/statalist-faq/
>> * http://www.ats.ucla.edu/stat/stata/
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/faqs/resources/statalist-faq/
> * http://www.ats.ucla.edu/stat/stata/ 		 	   		  
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index