st: re: inference on coefficients

[Kit Baum <baum@bc.edu>]

Ben said

I use xtabond two-step, but I got 'Warning: Arellano and Bond
recommend using one-step results for inference on coefficients'. Could
anyone help to explain why it comes like this and how I can solve the
problems please?


It is well known that the regular GMM standard errors in two-step  
estimation are biased, so that making inferences from those SEs may  
be unreliable. Thus Windmeijer developed the "Windmeijer correction"  
which reduces their bias. This is available with official Stata's - 
xtabond- if you specify the twostep and vce(robust) options. It is  
also available in the user-written -xtabond2-, where it is the  
default if twostep estimation is performed. -findit xtabond2- for  
details.

Kit

Kit Baum, Boston College Economics and DIW Berlin
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/