Notice: On March 31, it was **announced** that Statalist is moving from an email list to a **forum**. The old list will shut down on April 23, and its replacement, **statalist.org** is already up and running.

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

From |
Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: Is it OK to interact an instrumented variable? |

Date |
Wed, 31 Aug 2011 03:31:24 -0700 |

You have three options: 1) Instrument X2 and X1*X2 by Z and Z*X1. This is what is commonly done, but problem with this approach is that it requires conditional independence rather than just uncorrelatedness of the instruments (included and excluded) with the structural error. 2) 2SLS with the second stage OLS using the regressors X2hat and X1*X2hat, where X2hat is the projection of the endogenous regressor on a full set of instruments. Here you will need to use a bootstrap or a similar technique to get valid standard errors. 3) Use X2hat and X1*X2hat as _instruments_. This requires only uncorrelatedness and gives valid standard errors. For an excellent discussion of these issues, see pg. 143 onwards in the 2nd edition of Prof. Woodridge's book, Econometric Analysis of Cross Section and Panel Data. T On Wed, Aug 31, 2011 at 12:13 AM, Joseph B. Creek <inkind1982@yahoo.com> wrote: > Hello, > > Hello, > > I am unsure about how to deal with this model. I want to estimate > > Y = B0 + B1*X1 + B2*X2 + B3*X1*X2 + controls + e > > X1 is exogenous and X2 is endogenous. The interaction of X1 and X2 has a clear economic interpretation that I want to investigate empirically. I intend to estimate using IV and have an instrument Z for X2. I have been told I can use Z and Z*X1 as instruments for the endogenous variables X2 and X1*X2. Is this true? Also, does it matter in this case if X1 and X2 are likely related? > > Is there any specific command I should use for this? > > Your help is much appreciated. > * > * For searches and help try: > * http://www.stata.com/help.cgi?search > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ > -- Tirthankar Chakravarty tchakravarty@ucsd.edu tirthankar.chakravarty@gmail.com * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: Is it OK to interact an instrumented variable?***From:*"Joseph B. Creek" <inkind1982@yahoo.com>

- Prev by Date:
**st: Testing whether the average partial effects are statistically different across groups?** - Next by Date:
**Re: st: Testing whether the average partial effects are statistically different across groups?** - Previous by thread:
**st: Is it OK to interact an instrumented variable?** - Next by thread:
**-stripplot- revised on SSC [was: Re: st: Transposing dotplot (changing axes)]** - Index(es):