Dominance Thresholds: A Cautionary Note

• Author: Douglas A. Herman,Shawn W. Ulrick,Seth B. Sacher
• Date: 01 December 2014
• DOI: 10.1177/0003603X1405900407
• Published date: 01 December 2014
• Subject Matter: Article

THE ANTITRUST BULLETIN: Vol. 59, No. 4/Winter 2014 : 855
Dominance thresholds:
A cautionary note
BY DOUGLAS A. HERMAN,* SHAWN W. ULRICK**
AND SETH B. SACHER***
As a threshold matter, high market shares are considered informative of
potential underlying competitive dynamics, especial y issues of market
power and dominance. However, as is wel known, high shares may not
tel the entire story. Beyond issues related to market definition, entry, and
efficiencies, this article notes an additional reason why caution may be
warranted before drawing conclusions about dominance or market
power from share evidence. Specifical y, such conclusions can be statisti-
cal y unsupportable in the presence of smal sample issues. In markets
with relatively few transactions (that is, “thinly” traded markets), the
implications of observed high market shares are much less clear than in
more thickly traded markets. It is entirely possible that situations that
appear to implicate a dominant firm actual y reflect pure random chance
in a competitive process involving equal y matched or nondominant
firms. This article discusses theories and methods for distinguishing
between outcomes that exhibit strong statistical evidence of dominance
as opposed to those that merely reflect the random distribution of win-
nings among nondominant firms.
KEY WORDS: antitrust, market structure, monopolization, dominance, smal
samples, statistical significance
* Economist, Encino, CA.
** Economist, U.S. Federal Trade Commission, Washington, DC.
*** Economist, U.S. Federal Trade Commission, Washington, DC.
AUTHORS’ NOTE: This article reflects our views only and does not necessarily
represent the views of the Federal Trade Commission or any individual commissioner.
Mark Glick, Dave Schmidt, Mike Vita, Nicholas Franczyk, Malcolm Coate, and John
Yun provided helpful comments and suggestions. Remaining errors are ours.
© 2014 by Federal Legal Publications, Inc.

---

856 : THE ANTITRUST BULLETIN: Vol. 59, No. 4/Winter 2014
I.
INTRODUCTION
It is general y maintained in competition matters that a high market
share, such as seventy percent, provides prima facie evidence of
monopoly power or dominance.1 Such a conclusion is consistent with
the way U.S. courts as wel as U.S. and international antitrust agencies
evaluate general structural evidence.2 Of course, such structural evi-
1
Under U.S. competition law, evidence of market power or monopoly
power is a requisite condition for a monopolization or attempted
monopolization offense. See, e.g., United States v. Grinnell Corp., 384 U.S. 563,
570–71 (1966). The courts have defined market power and monopoly power
in related but different manners. The Supreme Court has stated that “market
power exists whenever prices can be raised above the levels that would be
charged in a competitive market.” Jefferson Parish Hosp. Dist. No. 2 v. Hyde,
466 U.S. 2, 27 n.46 (1984). Monopoly power has been defined as “the power to
control prices or exclude competition.” United States v. E. I. du Pont de
Nemours & Co. (Cel ophane), 351 U.S. 377, 391 (1956). EU competition law
generally defines a firm as dominant if it behaves “to an appreciable extent
independently of its competitors, customers and ultimately of its consumer.”
Case 27/76, United Brands Cont’l BV v. Comm’n, 1978 E.C.R. 207. These legal
definitions of market power, monopoly power and dominance may not fully
comport with economic notions. For example, Carlton & Perloff state:
It is common practice to say that whenever a firm can profitably set
its price above its marginal cost without making a loss, it has monop-
oly power or market power. One might usefully distinguish between the
terms by using monopoly power to describe a firm that makes a profit if
it sets its price optimally above its marginal cost, and market power to
describe a firm that earns only the competitive profit when it sets its
price optimally above its marginal cost. However, people do not
always make this distinction, and generally use the two terms inter-
changeably, sometimes creating confusion.
DENNIS CARLTON & JEFFREY PERLOFF, MODERN INDUSTRIAL ORGANIZATION 93 (2005).
2
Discussions of the requisite market share required for monopoly power
or dominance commonly begin with Judge Hand’s statement in United States v.
Aluminum Co. of America that while a market share of ninety percent “is enough
to constitute a monopoly[,] it is doubtful whether sixty or sixty-four percent
would be enough; and certainly thirty-three per cent is not.” 148 F.2d 416, 424
(2d Cir. 1945. Fol owing Alcoa and American Tobacco, U.S. courts typical y have
required a dominant market share before inferring the existence of monopoly
power. The Fifth Circuit observed, “monopolization is rarely found when the
defendant’s share of the relevant market is below 70%.” Exxon Corp. v.
Berwick Bay Real Estate Partners, 748 F.2d 937, 940 (5th Cir. 1984) (per curiam).

---

D O M I N A N C E T H R E S H O L D S : 857
dence is a rebuttable presumption. Conclusions about monopoly
power or dominance can be made only after evaluating possible offset-
ting considerations and justifications, such as entry, and transaction or
practice-specific efficiencies, such as innovation, among others.
This article discusses an additional reason that caution may be
warranted before drawing conclusions about dominance or market
power from share evidence alone. It is argued that in markets with
relatively few transactions (that is, “thinly” traded markets), the
implications of observed high market shares are much less clear than
in more thickly traded markets.3 Simply put, it is entirely possible that
Similarly, the Tenth Circuit noted that to establish “monopoly power, lower
courts general y require a minimum market share of between 70% and 80%.”
Colo. Interstate Gas Co. v. Natural Gas Pipeline Co. of Am., 885 F.2d 683, 694
n.18 (10th Cir. 1989) (citation omitted). Likewise, the Third Circuit stated that “a
share significantly larger than 55% has been required to establish prima facie
market power” and held that a market share between seventy-five percent and
eighty percent of sales is “more than adequate to establish a prima facie case of
power.” United States v. Dentsply Int’l, Inc., 399 F.3d 181, 187 (3d Cir. 2005).
On the other hand, the European Commission’s Tenth Report on Competition
indicated a dominant position can generally be said to exist once a market
share on the order of forty to forty-five percent is reached. It also emphasized
the general unevenness in shares in reaching this conclusion:
Although this share does not in itself automatically give control of
the market, if there are large gaps between the position of the firm
concerned and those of its closest competitors and also other fac-
tors likely to place it at an advantage as regards competition, a
dominant position may well exist.
COMMISSION OF THE EUROPEAN COMMUNITIES, TENTH REPORT ON COMPETITION
POLICY, at 103, ¶ 150 (1981).
3
There are numerous industries that are of antitrust interest that are
thinly traded, including defense contracting, aerospace, school milk
contracting, public works contracting, and various commodities (agricultural
as well as chemicals, metals and other industrial supplies), although this may
be affected by how markets are defined. One example of a recent case where
issues of market thinness may have been relevant is Race Tires Am., Inc. v.
Hoosier Racing Tire Corp., 614 F.3d 57 (3d Cir. 2010). This case involved
exclusive equipment requirements adopted by motor sports sanctioning
bodies. The market involved “tires for dirt oval track racing in the United
States and Canada.” Id. at 62. The sanctioning bodies at issue in this market
required that racers use a specific tire type and brand in all of their races for
the season. There were three major manufacturers and three major

---

858 : THE ANTITRUST BULLETIN: Vol. 59, No. 4/Winter 2014
situations that appear to implicate a dominant firm actually reflect
pure random chance in a competitive process involving equally
matched or nondominant firms. In thinly traded markets with firms
that are equally matched in every way, it would not be unusual to
observe one firm with a market share exceeding generally accepted
thresholds for monopoly power or dominance, even though these
observed market shares are a result of a stochastic outcome.4 The goal
of this article is to discuss methods and theory to distinguish between
outcomes that show strong statistical evidence of dominance versus
those that merely reflect the random distribution of winnings among
nondominant firms.
The remainder of this article is organized as follows. Section II
develops the above contention in the context of a two-firm model
with “equally matched” firms. Probability theory is used to demon-
strate the likelihood that a firm in such a context can have an
observed market share consistent with monopoly power or domi-
nance. The analysis also discusses how this conclusion is directly
“customers” which were the sanctioning bodies. The plaintiff alleged that
between 2003 and 2007 the sanctioning companies had requested competitive
bids only seven times. The defendant, Hoosier Tire, had a seventy-nine
percent share of the market for dirt track racing tires. The court found that
notwithstanding the defendant’s high market share, the sanctioning bodies’
rule “creates more exciting races, ensures equal access to a uniform product,
tends to increase safety, and lowers the cost of tires” and...