Review of Industrial Organization 15: 205–217, 1999.
© 1999 Kluwer Academic Publishers. Printed in the Netherlands.
Effects of Assumed Demand Form on Simulated
, LUKE FROEB
and STEVEN TSCHANTZ
Vanderbilt University, Nashville, TN 37203, U.S.A.
GREGORY J. WERDEN
U.S. Department of Justice, Washington, DC 20530, U.S.A.
Abstract. This paper investigates the properties of four demand systems used to predict the effects
of differentiated products mergers: the Almost Ideal Demand System (AIDS), logit, linear, and log-
linear (constant elasticity). In Monte Carlo experiments, these demand systems yield signiﬁcantly
different predictions, although all are calibrated to the same the same, randomly generated, premerger
relative quantities and demand elasticities. The predicted price increase is greatest with log-linear
demand, followed by the AIDS. The linear and logit demand forms result in signiﬁcantly lower
postmerger prices. The results highlight the importance of the inherent higher-order properties of
demand systems, i.e., their “curvature.”
Key words: Mergers, antitrust, AIDS, logit, Computed Nash Equilibria.
JEL Classiﬁcation: L41-horizontal anticompetitive practices.
In 1992 the U.S. Department of Justice and Federal Trade Commission issued new
Merger Guidelines, which divide potential anticompetitive effects into those that
involve some form of coordination and those that do not. The latter are termed
unilateral effects, and in differentiated products industries, the federal enforcement
agencies have been focusing primarily on unilateral effects (see Shapiro, 1996).
Merger simulation has been advanced in policy debates and extensively used in ac-
tual merger investigations to assess the unilateral effects of differentiated products
mergers (see Hausman and Leonard, 1997; Werden, 1997a,b,d; Werden and Froeb,
Department of Mathematics, firstname.lastname@example.org.
Owen Graduate School of Management, email@example.com; support for this project
was provided by the Dean’s Fund for Faculty Research.
Department of Mathematics, firstname.lastname@example.org.
The views expressed in this paper are not purported to represent those of the U.S. Department
of Justice, email@example.com.