st: New version of felsdvreg.ado to estimate large number of two-wayfixed effects

[Thomas Cornelissen <cornelissen@ewifo.uni-hannover.de>]

Dear all,
a new version of my program -felsdvreg- for a memory saving fixed 
effects estimation of the linear two/three-way
error components model is available.

-felsdvreg- computes a linear regression model with two fixed effects 
(e.g. person and firm effects).  One
effect (e.g. the firm effect) is included as dummy variables while the 
other effect (e.g. person effect)
is eliminated by substracting group means (within-transformation).
(A third effect, e.g. a time effect, or more effects can be specified by 
including the corresponding dummies.)

Of course this could be done "manually", but the program is "memory-saving"
in that it doesn't explicitly generate all the dummy variables for the 
second effect,
but still compute the exact least squares estimates. Therefore the 
program can be used in
large datasets.

At least Stata 9 is required to run the program. The program is 
available by typing
net from http://www.thomas-cornelissen.de/stata/
in Stata.
Best,
Thomas

--

Thomas Cornelissen
Institut für Empirische Wirtschaftsforschung
Leibniz Universität Hannover

Institute of Empirical Economics
University of Hannover, Germany

Tel.: +49 (0)511 762-5661

*
*   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/