Review of Industrial Organization
12: 609–623, 1997.
1997 Kluwer Academic Publishers. Printed in the Netherlands.
RICHARD S. HIGGINS
Capital Economics, 1299 Pennsylvania Avenue, NW, Washington, DC 20004, U.S.A.
Abstract. Diagonal merger combines the assets of an input supplier and a downstream rival of the
input demander that does not use the input. Diagonal mergers are likely to be overlooked by federal
antitrust authorities as they are neither vertical nor horizontal mergers. Diagonal mergers are shown
to be nearly as anticompetitive as comparable horizontal mergers and, like horizontal mergers, the
welfare effects of diagonal mergers are predicted in the ﬁrst instance by a modiﬁed HHI calculation.
Key words: Merger, antitrust, Herﬁndahl Index.
Recently, several articles have claimed to provide an analytical basis for anticom-
petitive vertical integration, including Salinger (1989, 1991); Ordover et al. (1990)
[O-S-S]; and Hart and Tirole (1990). None of these attempts succeeds sufﬁciently
to justify invigorating federal antitrust enforcement of vertical merger as urged by
Riordan and Salop (1994). Both Salinger and O-S-S assume ﬁrm conduct that is
irrational or, at least, which has not been demonstrated to be individually rational.
Speciﬁcally, Salinger assumes vertically integrated input suppliers do not sell in
the merchant market, and O-S-S similarly assume commitment to a downstream
price by the vertically integrated input suppliers.
There are only two potential-
ly anticompetitive bases for vertical integration: (1) price discrimination; and (2)
variable proportions. In the ﬁrst case, as is true for price discrimination in gener-
al, the effect on economic welfare–and even output–is ambiguous depending on
higher order demand curvature.
Similarly, with variable proportions the effect of
vertical integration is ambiguous. Finally, in neither case, is pre-merger market
I wish to thank Raymond A. Jacobsen, Jr., David P. Kaplan, Paul H. Rubin, Marc G. Schildkraut,
and especially Gregory J. Werden for helpful comments on an earlier draft of this paper. I also wish to
thank the Center for the Study of Public Choice for inviting me to present this paper and the members
of its Wednesday Seminar for their critical comments.
Reiffen and Vita (1995) provide an excellent analysis of the Post-Chicago vertical integration
literature, except for the Hart and Tirole paper, which they do not address. Also, see Reiffen (1992),
which speciﬁcally criticizes Ordover, Saloner and Salop, and see Ordover, Saloner and Salop’s reply
The justiﬁcation for rejecting the Hart and Tirole analysis as a basis for antitrust enforcement of
vertical merger is that Hart and Tirole unrealistically assume contingent and multi-part price contracts
which effectively eliminate all but the anticompetitive gains from vertical integration.
See Joan Robinson (1933) and Ekelund et al. (1981).