st: Interpreting the dropped variables of clogit

["Rienzo, C" <C.Rienzo@rhul.ac.uk>]

Dear all,

I wonder whether you could help me to confirm/understand the  
interpretation  of variables dropped in a conditional logit model.

I have been reading some past FAQ:
http://www.stata.com/statalist/archive/2009-03/msg01161.html
http://www.stata.com/statalist/archive/2005-08/msg00036.html

but I am still have a doubt.

I am using a long  (unbalanced) panel data; say I am interested in 
estimating the marginal effects to be in part time  (PT)  job for a 
worker  given a set of workers and firms characteristics.
Therefore PT=1 if worker  is in part time job
          PT=0 if Not in part time job.

Now when running the clogit STATA will report that “all positive and 
all negative values dropped”.

My understanding in this specific case  is that STATA will drop workers 
that either ALWAYS had PT job or NEVER had PT job, that is when there 
is no within variation.
Is this correct? 
Or the dropped observations are due to any other variable that has not within variation not only for the part time one?

Many thanks,

cinzia

*
*   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/