Review of Industrial Organization (2007) 30:1–27 © Springer 2006
Measuring the Extent of Structural Remedy
in Section 7 Settlements: Was the US DOJ
Successful in the 1990s?
MIKHAIL S. KOULIAVTSEV
School of Business Administration, Philadelphia University, Henry Ave and Schoolhouse
Ln, Philadelphia, PA 19144. E-mail: firstname.lastname@example.org
Abstract. This paper suggests an innovative measure of structural relief obtained in a
typical Section 7 settlement. The fraction of competitive overlap subject to divestiture as
a condition of the settlement is modeled as a function of merger-speciﬁc efﬁciencies, the
proportion of the deal held “hostage” to antitrust review, the merger’s anticompetitive
potential, and other factors. The model is applied to data on 86 recent Justice Depart-
ment cases covering the period 1990–2003 and to the subsample of 1990s cases. All data
are collected from publicly available documents only. The government is found to secure
larger divestitures when the cost to the acquirer of delaying the settlement is high. The
resulting estimates are used to predict several out-of-sample observations.
Key words: merger policy, U.S. Department of Justice, structural remedies.
JEL Classiﬁcations: L44, C24.
Potentially anticompetitive mergers are challenged by the U.S. antitrust
authorities under the power granted to them by Section 7 of the
Clayton Act of 1914, as amended by the Celler-Kefauver Act of 1950.
Structural relief or asset divestiture is recognized as the preferred method
of resolving potential anticompetitive problems in Section 7 cases. Stud-
ies by Elzinga (1969), Pfunder et al. (1972), and Rogowsky (1982, 1986)
revealed that nonstructural remedies—court orders prohibiting certain con-
duct, for example—are often ineffective as ways of restoring competition in
affected (product and geographic) markets. This paper develops an inno-
vative technique to assess the extent of structural relief obtained in a
I would like to thank the General Editor and two anonymous referees for many
helpful suggestions on earlier drafts of this paper. The remaining errors are mine.