st: stochastic frontier analysis efficiency scores

[Arne.Bueschemann@wiwi.uni-paderborn.de]

We have a panel dataset covering all Valuations for 30 companies for each 
year between 2001 and 2009 as well as several independent variables. 
Therefore, we applied a xtfrontier model to discover the technical 
efficiency for each company and each year. Now we have the problem that 
the efficiency scores vary in each year but the yearly ranking of all 
companies remains the same which we consider highly unlikely. 

We used following command: xtfrontier logdep indep1 indep2  l indep3  l 
indep4  l indep5  l indep6  l indep7, tvd i(id) t( idtime)

and for the post estimation we used: predict efficiency, te

What we try to find out: Does the efficiency of the companies increase or 
decrease over time and does the ranking between the companies change in 
regard to technical efficiency. Therefore, we are not quite sure how to 
use xtfrontier command or if we should apply frontier analysis for each 
year separately. 

We would appreciate your help and to let us know which model we should 
apply

Best regards, 
Arne
*
*   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/