# st: Re: reg3 and failure of the order condition

 From baum To statalist@hsphsun2.harvard.edu Subject st: Re: reg3 and failure of the order condition Date Wed, 31 Jul 2002 08:41:46 -0400

--On Wednesday, July 31, 2002 2:33 -0400 Joachim wrote:

Dear Stata users,

I was trying to use the reg3 command in Stata7 to determine a basketball
player's wage and team tenure via reg3:
The estimation is:
reg3 (salary tenure tenure2 experience experience2 draft player size all
star games direct performance indirect performance) (tenure draft all star
games direct performance indirect performance experience experience2
player size)

However an error message popped up:

Equation is not identified--does not meet order conditions
Equation salary : salary tenure tenure2 experience experience2 draft all
star game player size direct performance indirect performance

Exogenous variables: tenure2 experience experience2 draft all star game
player size direct performance indirect performance
r(481)
Any help out there?
Thank you very much
Joachim
In this syntax for the reg3 command, Stata is figuring out what are the endogeneous and exogeneous variables in your model. Since you have two equations, the endogeneous variables are taken as salary and tenure. For a simultaneous equation to be identified, it is necessary but not sufficient to satisfy the order condition, which may be stated in different ways, but most simply means that the exogeneous variables EXcluded from each equation must be at least as numerous as the endogeneous variables INcluded on each equation's RHS.

For the second equation, only tenure is endogeneous, so that this equation could be consistently estimated with OLS (e.g. the model is recursive).

For the first equation, both endogeneous variables appear; thus the order condition fails, 0 < 1, as there are no available instruments for tenure. You must have at least one exogeneous variable in the system excluded from this eqation, and it does not appear that this is so (well, Stata's calculation is better than my eyesight).

Kit
*
* 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/