dear listers,
I'm pretty new to the subject of stochastic frontier models and from
the STATA reference texts I didn't quite understand what exactly the
panel specific difference is that an xtfrontier estimate generates
compared to a simple pooled estimate using frontier, d(t) when applied
to panel data.
For an experiment of comparison I applied frontier, d(t) and xtfrontier
to a set of panel data. the estimates tend to differ more than i
expected, and in the frontier, d(t) case the standard errors tend to be
considerably smaller.
To me it seemed from the text that the only difference is that
xtfrontier treats the inefficiency-term uniformely over a group, whereas
frontier treats it separately for each observation - might this be
responsible for the difference, or do I miss some (possibly completely
obvious ;-)) fact?
*
* 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/
