[Limdep Nlogit List] Problem running Multinomial Logit Model

[broniakc at onid.orst.edu]

Hi,

I am attempting to run a MLOGIT model on multiple choice outcomes.  When I
include all my independent variables in the model, the program returns the
following error:

Error 135:  Models-Singular Hessian During Newton Iterations

Here is the output:

--> LOGIT;Lhs=LU92;Rhs=ONE,ARC8792,ARP8792,ARF8792,ARU8792,ARR8792,D2,D3
    ,CINTTERM,PINTTERM,FINTTERM,UINTTERM,RINTTERM;Wts=XFACT;Marginal Effects
    ;PrintVC$
Hessian is not definite at current values.
Switching to BFGS (gradient based) method.
(Not a failure. Just looking for a better algorithm.)
Line search does not improve fn. Exit iterations. Status=3
Error:   806: Line search does not improve fn. Exit iterations. Status=3
Function=  .21810229037D+04, at entry,  .14280957686D+06 at exit
Error:   135: Models - singular Hessian during Newton iterations.

I specify this using all but my RINTTERM variable,(including the weighting
variable) and the model runs properly.  I have checked this variable for
missing data or other anomalies that might explain this.  Does anyone know why
this is happening?

Sincerely,
Christine Broniak