Dear Viktor,
Your problem is not with the random coefficients, or other model assumptions, but with your interpretation of elasticities. They do not necessarily range between zero and one. The proper interpretation is the percentage increase in y for one percent increase in x. So it can positive or negative (in which case y decreases when x increases), and it can be smaller and larger than 1 (economists usually call elasticities between -1 and 1 inelastic and elasticities less than -1 or more than 1 elastic).
HTH,
Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z214
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
Viktor Slavtchev wrote:
> The model I'm using is linear and defined as followed:
> ln(ouput)=a+b*ln(input). Thus, the estimated coefficients could be
> interpreted as elasticities and therefore range in [0;1]. What I'm
> obtaining are group specific coefficients larger than 1 and/or even
> negative values. Coefficients in [0;1] are really an exception. The
> overall estimated elasticity is about .4. Are there any particular
> distributional assumptions required? Are there any other methods, which
> allow random coefficient estimation? Do you have some tip for me to
> solve that problem?
*
* 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/
