Re: st: IVtobit estimation

["Austin Nichols" <austinnichols@gmail.com>]

Leonor--
Richard makes an excellent point.  I also wonder if instead of -ivtobit- you
can use -qvf- (-findit qvf- to install).  I suspect many -tobit-
models that impose
censoring at zero in fact have a theoretically nonegative dep var, and no
censoring at all.

On 4/15/07, Richard Williams <Richard.A.Williams.5@nd.edu> wrote:
> At 09:38 PM 4/14/2007, Leonor Saravia wrote:
> >Hello Statalist,
> >
> >I'm using the ivtobit6 command to do my regressions in Stata 8, but
> >when I run the complete model including the ll(0) (That's the
> >censoring point the dependent variable has), Stata tells me that I
> >can't use that extension, in fact, it tells: "invalid ll".  The model
> >which i'm trying to run is:
> >
> >ivtobit6 inversion, endog(impprom) iv(imppromVI) exog(PIB98
> >industria1-industria8 tamano1 tamano2 region2-region13),ll(0)
> >
> >When I run the same model without the ll(0), the command functions well.
>
> It looks like you have an extra comma before ll(0).
>
> However, why even use this in the first place?  The package
> description says "Stata 8.2 or 9.x users should use official ivprobit
> and ivtobit." Updating from Stata 8 to 8.2 is free.
*
*   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/