Dear statalist,
Email attempts to contact the authors of -movestay- (SJ 4(3)) have not
been successful, so I wonder if anyone else on the list might have
some thoughts.
My attempts to use this command for endogenous switching regressions
produce an error message which I can't decipher:
Fitting initial values .....initial vector: copy option requires
either a matrix or a list of numbers r(198);
Running the command on the authors's example dataset worked fine. (I
am using Stata 9 on Windows 2000, with 1 gig ram).
Since I couldn't get movestay- to work, I have been using heckman, one
equation at a time. When I use this approach on the author's example
dataset I get almost identical results (coefficients to about 2
decimal points), but the sign on rho_2 is negative, not positive, as
in the example shown in the Stata Journal.
Does anyone know why this might be so? My understanding is that the
heckman ml results will not be as efficient as the FIML approach, but
the sign on the correlation should not change, should it?
Here is my code using the authors's example dataset:
use movestay_example, clear
* first the authors' command:
movestay lmo_wage age age2 edu13 edu4 edu5 reg2 reg3 reg4, ///
select(private=m_s1 job_hold)
* now with switching and heckman
capture drop wage
* first private by making all public missing
gen wage=lmo_wage
replace wage=. if private==0
heckman wage age age2 edu13 edu4 edu5 reg2 reg3 reg4, ///
select(age age2 edu13 edu4 edu5 reg2 reg3 reg4 m_s1 job_hold)
* now public by making all private missing
replace wage=lmo_wage
replace wage=. if private==1
heckman wage age age2 edu13 edu4 edu5 reg2 reg3 reg4, ///
select(age age2 edu13 edu4 edu5 reg2 reg3 reg4 m_s1 job_hold)
--
Kind regards,
Ian
-------------------------------
Ian Watson
Senior Researcher
acirrt, University of Sydney
NSW, 2006, Australia
phone: 02 9351 5622
email:[email protected]
www.acirrt.com
-------------------------------
*
* 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/
