On the Practice of Lagging Variables to Avoid Simultaneity
| Date | 01 December 2015 |
| Author | William Robert Reed |
| DOI | http://doi.org/10.1111/obes.12088 |
| Published date | 01 December 2015 |
897
©2015 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 77, 6 (2015) 0305–9049
doi: 10.1111/obes.12088
On the Practice of Lagging Variables toAvoid
Simultaneity*
William Robert Reed†
†Department of Economics and Finance, University of Canterbury, Private Bag 4800,
Christchurch 8140, New Zealand (e-mail: bob.reed@canterbury.ac.nz)
Abstract
A common practice in applied economics research consists of replacing a suspected simul-
taneously determined explanatory variable with its lagged value. This note demonstrates
that this practice does not enable one to avoid simultaneity bias. The associated estimates
are still inconsistent, and hypothesis testing is invalid.An alternative is to use lagged values
of the endogenous variable in instrumental variable estimation. However, this is only an
effective estimation strategy if the lagged values do not themselves belong in the respec-
tive estimating equation, and if they are sufficiently correlated with the simultaneously
determined explanatory variable.
I. Introduction
Simultaneity is a concern in much of empirical economic analysis. One approach that
has been employed to avoid the problems associated with simultaneity is to replace the
suspect explanatory variable with its lagged value. The practice is widespread, as can be
confirmed by searching for variations of ‘avoid simultaneity lagged variables’ on Google
Scholar. Recent examplesare Aschhoff and Schmidt (2008); Bania, Gray and Stone (2007);
Bansak, Morin and Starr (2007); Brinks and Coppedge (2006); Buch, Koch and Koetter
(2013); Clemens et al. (2012); Cornett et al. (2007); Green, Malpezzi and Mayo (2005);
Gupta (2005); Hayo, Kutan and Neuenkirch (2010); Jensen and Paldam (2006), MacKay
and Phillips (2005); Spilimbergo (2009); Stiebale (2011); and Vergara (2010). The practice
is common across a wide variety of disciplines in economics and finance. Many appear
in top journals including the American Economic Review, the Journal of Finance, the
Economic Journal, and the Journal of Banking & Finance, and are highly cited.
The rationale for the practice is explicitly identified in statements such as the following:
‘We avoid poor-quality instrumental variables and instead address potential biases from
reverse and simultaneous causation by… lagging’ (Clemens et al., 2012); ‘The vector of
*I acknowledge helpful comments from Kuntal Das, Chris Hajzler, Steven Stillman, JeffreyWooldridge, James
Ziliak, Arthur Grimes and seminar participants at the University of Otago and the 2014 New ZealandAssociation of
Economists conference. I am also grateful for the comments from anonymous reviewerswhich substantially improved
the analysis. Remaining errors are my own.
JEL Classification numbers: C1, C5, C15
To continue reading
Request your trial