st: ml maximize---identical z-stats across all coefficients

[A Marco <marcoecon@gmail.com>]

Dear users,

I am using Stata/MP 11 on unix. I have an admittedly complex maximum
likelihood problem, for which I'm using the "lf" estimator. I'm frequently
getting results across different samples and specifications where *all* the
z-stats for the coefficient vector in one of the equations are identical or
nearly identical. This can't be coincidence (not this many times in a row),
and I'm wondering if it is symptomatic of a deeper problem (probably
identification? We're doing some simulations now).

Has anyone come across this?

Secondly---this may or may not be related---when the algorithm instead
gives missing SEs, the offending variable is almost always for the constant
term in the first equation. My only thought was to demean the X's so that
the constant term is meaningful, but that didn't seem to help in most
cases.

ML model command and output below.

Many thanks!
Alan Marco, US Patent and Trademark Office

---------------------

version 10 /* version 11 does not allow for two techniques in the options */
ml model lf tmoptions ///
   (x0: mark_renewed = class age age2 count addage) ///
   (sigma: ) ///
   (delta: ) ///
   [fw=weight], technique(bhhh 10 bfgs 10)
ml search
ml maximize

(snip)

Iteration 95:  log likelihood = -2022.0431
Iteration 96:  log likelihood = -2022.0406
Iteration 97:  log likelihood = -2022.0389
Iteration 98:  log likelihood = -2022.0387

                                                 Number of obs   =3229
                                                 Wald chi2(5)    =0.02
Log likelihood = -2022.0387                      Prob > chi2     =1.0000
------------------------------------------------------------------------------

mark_renewed |      Coef.   Std. Err.      z    P>|z|     [95% Conf.Interval]

-------------+----------------------------------------------------------------
x0          |
      class |   .6346415   4.894683     0.13   0.897    -8.958761 10.22804
        age |  -5.188072   39.47252    -0.13   0.895    -82.55278 72.17664
       age2 |   .0789744    .600916     0.13   0.895    -1.098799 1.256748
      count |   2.110165   16.27507     0.13   0.897    -29.78839 34.00872
     addage |   .7506584   5.731206     0.13   0.896     -10.4823 11.98362
      _cons |   90.61308   689.2631     0.13   0.895    -1260.318 1441.544
-------------+----------------------------------------------------------------
sigma       |
      _cons |   .6657538    .177967     3.74   0.000     .3169449 1.014563
-------------+----------------------------------------------------------------
delta       |
      _cons |   .0564934   .2982348     0.19   0.850     -.528036 .6410228
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/