st: re: odd "repeated time values" error in glm

[Kit Baum <baum@bc.edu>]

Clive reports

I've recreated the problem using one of Stata 9's toy datasets. Below,
-unemp- is the unemployment rate and -region- is a continuous  
variable with
no upper limit:

. webuse productivity, clear
(Public Capital Productivity)

. tsset state year
       panel variable:  state (strongly balanced)
        time variable:  year, 1970 to 1986

. glm unemp state public, link(logit) vce(hac nwest 1) t(year) robust  
eform
repeated time values in sample
r(451);



I believe all that you are seeing is that although -glm- knows how to  
compute a Newey-West covariance matrix, it does not know how to do so  
in the context of panel data. For that matter, neither does -newey-.  
David Roodman's -newey2- is better behaved---it works with panel  
data---but as Clive pointed out everything -newey2- can do, -ivreg2-  
can do (and then some).

I am not familiar with the help on -glm-, but I imagine the manual  
entry nowhere claims that the 'nwest' suboption works in a panel  
context. Perhaps it should clearly state that this is so if it is the  
case.


Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html


*
*   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/