Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: instruments of differing frequency

From   Abhimanyu Arora <>
Subject   st: instruments of differing frequency
Date   Thu, 12 Apr 2012 20:12:54 +0200

Dear statalist

I am doing a monthly panel-data analysis. Now to take care of
endogeneity concerns, I am proceeding with IV estimation and seek some
references/opinion on a couple of conceptual points.

I have two kinds of exogenous variables I am thinking of taking as
instruments—those that are at a monthly level and those that are at a
yearly level, both of which have their unique issues

A. For IV available at a monthly frequency, I need an autoregressive
equation in the (first-stage) reduced form with the lags (which I do
not know) being panel specific. For ivreg, the functional form is
irrelevant. One might also see some similarity with -xtabond- but is
not that exactly.

B. For the second type of IV- Is there any issue if I use yearly IVs
to instrument for monthly endogenous variable, given that I clustered
by year?

Many thanks indeed

Best regards


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index