Puja,
Could you not add the value of the constant to the estimated dummy variables
(where the dropped dummy variable takes on the value of the constant)?
For example, the following creates a new variable of the fixed-effect for
each company:
use http://www.stata-press.com/data/r8/grunfeld.dta
xi i.com
qui qreg invest mvalue kstock _I*,
gen fe = .
forv i = 1/10 {
qui replace fe = _b[_Icompany_`i'] + _b[_cons] if company == `i'
*local j = `j' +1
}
replace fe =_b[_cons] if company == 1
tabdisp company, c(fe)
Scott
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Puja Vasudeva
Sent: Wednesday, June 23, 2004 7:43 AM
To: [email protected]
Subject: st: Estimating sqreg without a constant
Dear Statalisters,
I am trying to estimate a quantile regression of wages on a host of
explanatory variables but without a constant term. The explanatory
variables include the full set of industry dummies that serve as fixed
effects in the absence of a constant. The primary motivation is to obtain
an inter-quantile range for industry dummies.
I have looked at the qreg and sqreg and tried to modify the ado files
(saved separately from the original ado-files) by including the noconstant
option after the regress commands but I'm not sure this is correct and can
find no way to check this. Would be very grateful for suggestions!
Puja
*
* 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/
