[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: ivreg2 test of exogeneity
> -----Original Message-----
> From: firstname.lastname@example.org
> [mailto:email@example.com] On Behalf Of
> Sent: 21 April 2006 10:51
> To: firstname.lastname@example.org
> Subject: st: ivreg2 test of exogeneity
> Dear Statalist members,
> i have a question concerning the command ivreg2.
> I am estimating a rudimental fixed effect model,
> where the
> dependent variable is wage, and the independent v. of
> interest unemployment rate.the observation units are groups
> of individuals differentiated by skills. i want to control
> for unobserved heterogeneity of the groups.some of my
> covariates are constant, and so they get treated as part of
> the error term.i want to test the strict exogeneity
> assumption on the explanatory variables that underlies the
> fe/re model by using ivreg2.in particualr i guess that
> education, which is constant over time and hence part of the
> individual component of the error term,correlates with the
> unemployment rate.
> i used the lag value of unemployment rate as an instrument,
> and pooled the data.
> is the following command correct?
> ivreg2 wage d_educ2-d_educ4 exp /*
> */ unemploym ( = unemploym-1) [aweight=weight],orthog(unemploym)
That's right, but the latest version of ivreg2 (2.1.15) has a new
option, "endog", which lets you do the same test but in a slightly more
intuitively understandable way:
ivreg2 wage d_educ2-d_educ4 exp /*
*/ (unemploym = unemploym-1) [aweight=weight], endog(unemploym)
You will get the same test statistic as your proposed command, but this
is a little clearer: you are asking whether unemploym needs to be
specified as endogenous.
> Thanks for your suggestions,
> Analog-/ISDN-Nutzer sparen mit GMX SmartSurfer bis zu 70%!
> Kostenlos downloaden: http://www.gmx.net/de/go/smartsurfer
> * 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/
* For searches and help try: