Review of Industrial Organization
12: 291–294, 1997.
1997 Kluwer Academic Publishers. Printed in the Netherlands.
Identifying Participants in a Price-ﬁxing Conspiracy:
Output & Market Share Tests Reexamined – Reply
ROGER D. BLAIR and RICHARD E. ROMANO
University of Florida, School of Business Administration, P.O. Box 117140, Gainesville, FL
In Blair and Romano (1990), we argue that a simple output test can be employed to
distinguish participants from nonparticipants in a price-ﬁxing conspiracy. Partici-
pantswill decrease their outputs to supporta market price increase (and higher prof-
its) as in the received theory of cartel behavior but modiﬁed to account for actions
of nonparticipants. Any nonparticipant, facing increased price (or demand), will
expand output. Hence, though both participants and nonparticipants charge higher
prices and enjoy increased proﬁts, the participants induce the price increase with
their output reduction, while nonparticipants mute the price increase with their uni-
lateral choices to produce more output. We illustrate the result by taking a standard
competitive (or Bertrand-pricing) equilibrium as the benchmark, and let a domi-
nant subset of ﬁrms conspire to ﬁx price, who then act collectively as a textbook
price-leading dominant ﬁrm.
We also provide an appendix that veriﬁes the output
test for the analogous characterization of competition (i.e., having prices as strategy
choices) but for differentiated products; and we have made available analysis con-
ﬁrming the output test should producers have quantities as their strategic variable
(i.e., be Cournot competitors).
Karaaslan challenges the robustness of the output test, claiming that it might
lead to both type I and type II errors. We do not take his criticism seriously because
it is developed in the context of an inconsistent model, one which artiﬁcially
changes the strategy space and timing of choices at the time the partial conspiracy
arises. It is the contrived changes that cause the output test to fail, not the partial
conspiracy as we will argue.
Since strategy spaces and timing of choices are
widely and appropriately regarded as unalterable structural properties of oligopoly
In the Bertrand interpretation of our benchmark equilibrium, we assume rationing of market
demand is such that a higher-priced ﬁrm would obtain zero market share. For example, all consumers
attempt to purchase from the lowest-priced ﬁrms. If they stock out, then consumers postpone purchas-
ing. Such a speciﬁcation eliminates the existence-of-equilibrium problem that arises in the Bertrand
model with decreasing returns to scale. See Tirole (1988, pp. 212-216) for a discussion of rationing
and existence in the Bertrand model.
Blair and Romano (1990, p. 40, n. 12).
We conﬁne our remarks to his criticism of our work.