Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: clustered standard errors in quantile regressions


From   Giovanni Pica <pica@socsci.soton.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   st: clustered standard errors in quantile regressions
Date   Thu, 31 Mar 2005 10:43:44 +0100

Dear all,

a question on bootstrapping using the cluster option. I need to estimate the
following quantile regression:

w = a + bZ + e

where w is the wage of individual i in state s at time t; Z is the
unemployment rate in state s at time t.

In a standard OLS regression the standard error of the coefficient b should
obviously be clustered at the s,t level i.e. cluster(state_year).

The problem is that I estimate a quantile regression:

bs "qreg w Z, q(.90)" "_b[Z]", cluster(state_year)

Does this provide clustered standard errors in the quantile regression
framework?

I also notice (i) that when I run this command the number of replications is
always lower than the one specified and (ii) the lower bound of the
percentile and biased-corrected confidence intervals are always equal to
the observed coefficient. Can you tell me why?


Thanks a lot.

Best regards,
giovanni pica


*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index