Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: xtivreg with FE and state-year interactions

From   Scott Cunningham <>
Subject   Re: st: xtivreg with FE and state-year interactions
Date   Wed, 3 May 2006 16:36:20 -0400

Sorry - that should say, "I suspect two variables are endogenous". And I should als0 note that "fip" is the state indicator variable.

On May 3, 2006, at 4:30 PM, Scott Cunningham wrote:

My data is individual level panel data. Approximately 800 individuals surveyed over 3 years. I have data on 50 American states. I have been estimating various models in which I control for state, year, state-year interaction and individual fixed effects. I suspect one independent variable is endogeneous and am using 2SLS to instrument. -xtivreg- works fine until I use the full model with year, state, year-state interactions and individual fixed effects. When I run this, with -xtivreg-, I get the following error:

. xi:xtivreg safesex srfm age age2 hgc hhd1 i.year i.fip i.year*i.fip (rp5 core=rp5iv coreiv), fe i(id) vce(boot)

i.year _Iyear_1998-2002 (naturally coded; _Iyear_1998 omitted)
i.fip _Ifip_1-55 (naturally coded; _Ifip_1 omitted)
i.year*i.fip _IyeaXfip_#_# (coded as above)

(running xtivreg on estimation sample)
_Iyear_2000 present in model more than once

Ordinarily, Stata drops dummies to deal with multi-collinearity created by the dummy trap problem, so I'm not sure what is happening here. Apologies if the answer's obvious.

* For searches and help try:
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index