[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Reconciling xtabond2 and DPD98

From   John Rolle <>
To   Stata List <>
Subject   st: Reconciling xtabond2 and DPD98
Date   Wed, 24 Oct 2007 08:56:36 -0700 (PDT)


I have been doing a systems GMM replication of a paper using the GAUSS program DPD98.  I specified the same model in  Stata (with identical lag lengths for the instruments), but get entirely different results for the coefficient estimates.  Is anyone familiar with why these routines would give different estimates?  I have already experimented with the "robust" option and alternative specifications of the "h()" matrix but these do not bring me any closer to the DPD98 results.  If you know the answer directly of some paper that discusses this issue, please let me know.


John Rolle
PHD Student
American University

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index