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

st: using y's lags as instruments

Subject   st: using y's lags as instruments
Date   Tue, 02 Mar 2004 10:21:55 -0600


I am very new to statalist, and hope someone has any insight on two
issues related to xtabond:
(a) using select lags as instruments, and
(b) using lags vs differenced lags as instruments.

To a): I am using xtabond to run a dynamic panel data model. I want
to use lags of Y as instruments. But I want to choose myself which
lags to use as instruments. The code   maxldep(#) in the xtabond
statement would allow me to choose the maximum number of lags to use
as instruments. Suppose, though, that I want to use
y(i t-3) as instrument but not y(i t-2) or y(i t-1). Specifying
maxldep(3) would, however, force me to use as instruments all three
lags, no?

To b): Besides, when I use   maxldep(3), is it using as instruments
the lags or the differenced lags?? Because, as I understand it,
xtabond automatically first-differences all the variables in the
model (except for those included in diffvars( )).

Any thoughts??

Tewodaj Mogues
Dept. of Agricultural and Applied Economics
University of Wisconsin - Madison
427 Lorch St. #317, Taylor Hall
Madison, WI 53706

*   For searches and help try:

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