In Re Insurance Brokerage Antitrust Litigation

PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT

No. 07-4046

IN RE: INSURANCE BROKERAGE
ANTITRUST LITIGATION (MDL No. 1663)

OptiCare Health Systems, Inc., Comcar Industries, Inc.,
Sunburst Hospitality Corporation, Robert Mulcahy,
Golden Gate Bridge, Highway and Transportation District,
Glenn Singer, Redwood Oil Company,
Omni Group of Companies, Bayou Steel Corporation,
Clear Lam Packaging, Inc., Cellect, LLC, Enclave, LLC,
Gateway Club Apartments, Ltd., Michigan Multi-King, Inc.,
City of Stamford, Belmont Holdings Corporation,
Tri-State Container Corporation,
Appellants

D.C. Civil Action No. 04-cv-5184
MDL No. 1663

Nos. 08-1455 & 08-1777
 IN RE: EMPLOYEE BENEFIT INSURANCE
BROKERAGE ANTITRUST LITIGATION (MDL 1663)

Maryann Waxman, on behalf of herself
and all others similarly situated,
Golden Gate Bridge, Highway and Transportation District,
Christopher Bare, David Boros, Cynthia Brandes,
Hans Fuson, Sharon Gehringer, Larry Hayes,
Brannen Henn, Robert H. Kimball, Wayne Moran,
Alicia A. Pombo, Clear Lam Packaging Inc.,
Connecticut Spring & Stamp Company,
City of Danbury, Connecticut,
Fire District of Sun City West,
Hollander Home Fashions Corporation,
Appellants

D.C. Civil Action No. 05-cv-1079
MDL No. 1663

On Appeal from the United States District Court
for the District of New Jersey
(Honorable Honorable Garrett E. Brown, Jr.)

Argued April 21, 2009
Before: SCIRICA, FISHER and
GREENBERG, Circuit Judges.

2
 (Filed August 16, 2010)

ELLEN MERIWETHER, ESQUIRE (ARGUED)
BRYAN L. CLOBES, ESQUIRE
Cafferty Faucher LLP
1717 Arch Street, Suite 3610
Philadelphia, Pennsylvania 19103

JOE R. WHATLEY, JR., ESQUIRE (ARGUED)
EDITH M. KALLAS, ESQUIRE
Whatley Drake & Kallas LLC
1540 Broadway, 37th Floor
New York, New York 10036

CHARLENE P. FORD, ESQUIRE
Whatley Drake & Kallas LLC
2001 Pennsylvaniark Place North, Suite 1000
Birmingham, Alabama 35203
Attorneys for Appellants

DANIEL J. LEFFELL, ESQUIRE
ANDREW C. FINCH, ESQUIRE
DAVID J. FRIAR, ESQUIRE
Paul Weiss Rifkind Wharton & Garrison LLP
1285 Avenue of the Americas
New York, New York 10019

3
KENNETH A. GALLO, ESQUIRE
Paul Weiss Rifkind Wharton & Garrison LLP
2001 K Street, N.W., Suite 600
Washington, D.C. 20006
Attorneys for Appellees,
American International Group, Inc.; American
International Specialty Lines Insurance Company;
Lexington Insurance Company; AIG Casualty Company
f/k/a Birmingham Fire Insurance Company of
Pennsylvania; American Home Assurance Company;
National Union Fire Insurance Company of Pittsburgh,
Pa.; National Union Fire Insurance Company of
Louisiana; American International Insurance Company;
The Insurance Company of the State of Pennsylvania;
AIU Insurance Company; Commerce and Industry
Insurance Company; New Hampshire Insurance
Company; The Hartford Steam Boiler Inspection and
Insurance Company; Illinois National Insurance Co.;
AIG Life Holdings (US), Inc. f/k/a American General
Corporation; AIG Excess Liability Insurance Company
Ltd. f/k/a Staff Excess Liability Company, Ltd.; AIG
Life Insurance Company; The United States Life
Insurance Company in the City of NewYork

SETH P. WAXMAN, ESQUIRE (ARGUED)
WILLIAM J. KOLASKY, ESQUIRE
JONATHAN E. NUECHTERLEIN, ESQUIRE
Wilmer Cutler Pickering Hale & Dorr LLP

4
1875 Pennsylvania Avenue, N.W.
Washington, D.C. 20006

PAUL A. ENGELMAYER, ESQUIRE
ROBERT W. TRENCHARD, ESQUIRE
Wilmer Cutler Pickering Hale & Dorr LLP
399 Park Avenue, 30th Floor
New York, New York 10022

ANDREA J. ROBINSON, ESQUIRE
JOHN J. BUTTS, ESQUIRE
Wilmer Cutler Pickering Hale & Dorr LLP
60 State Street
Boston, Massachusetts 02109
Attorneys for Appellees,
The Hartford Financial Services Group, Inc.; Hartford
Fire Insurance Co.; Twin City Fire Insurance Co.; Pacific
Insurance Co., Ltd.; Nutmeg Insurance Co.; The Hartford
Fidelity & Bonding Co.; Hartford Life and Accident
Insurance Company; Hartford Life Group Insurance
Company; Hartford Life Insurance Company

DONALD A. ROBINSON, ESQUIRE
LEDA DUNN WETTRE, ESQUIRE
Robinson Wettre & Miller LLC
One Newark Center, 19th Floor
Newark, New Jersey 07102

5
RICHARD C. GODFREY, ESQUIRE
LESLIE M. SMITH, ESQUIRE
DANIEL E. LAYTIN, ESQUIRE
ELIZABETH A. LARSEN, ESQUIRE
Kirkland & Ellis LLP
300 North LaSalle Street, Suite 2400
Chicago, Illinois 60654
Attorneys for Appellees,
Aon Corporation; Aon Broker Services, Inc.; Aon Risk
Services Companies, Inc.; Aon Risk Services, Inc. U.S.;
Aon Risk Services, Inc. of Maryland; Aon Risk Services,
Inc. of Louisiana; Aon Risk Services of Texas, Inc.; Aon
Risk Services, Inc. of Michigan; Aon Group, Inc.; Aon
Services Group, Inc.; Aon Re, Inc.; Affinity Insurance
Services, Inc.; Aon Re Global, Inc.; Aon Consulting, Inc.

LIZA M. WALSH, ESQUIRE
MARC D. HAEFNER, ESQUIRE
Connell Foley LLP
85 Livingston Avenue
Roseland, New Jersey 07068

H. LEE GODFREY, ESQUIRE
NEAL S. MANNE, ESQUIRE
JOHNNY CARTER, ESQUIRE
Susman Godfrey LLP
1000 Louisiana, Suite 5100
Houston, Texas 77002-5096

6
JEREMY S. BRANDON, ESQUIRE
Susman Godfrey LLP
901 Main Street, Suite 5100
Dallas, Texas 75202-3775
Attorneys for Appellees,
ACE Limited; ACE INA Holdings, Inc.; ACE USA, Inc.;
ACE American Insurance Co.; Westchester Surplus
Lines Insurance Co.; Illinois Union Insurance Co.;
Indemnity Insurance Co. of North America; ACE Group
Holdings, Inc.; ACE US Holdings, Inc.; Westchester Fire
Insurance Company; INA Corporation; INA Financial
Corporation; INA Holdings Corporation; ACE Property
& Casualty Insurance Co.; Pacific Employers Insurance
Co.

EAMON O'KELLY, ESQUIRE
JOHNS F. COLLINS, ESQUIRE
Dewey & LeBoeuf LLP
1301 Avenue of the Americas
New York, New York 10019

DAVID J. GRAIS, ESQUIRE
Grais & Ellsworth LLP
70 East 55th Street
New York, New York 10022
Attorneys for Appellees,
American Re Corporation; American Re-Insurance
Company; Munich-American Risk Partners; American

7
 Alternative Insurance Corporation

MICHAEL L. WEINER, ESQUIRE
PAUL M. ECKLES, ESQUIRE
Skadden Arps Slate Meagher & Flom LLP
Four Times Square
New York, New York 10036
Attorneys for Appellees,
AXIS Specialty Insurance Company; AXIS Surplus
Insurance Company; AXIS Capital Holdings Ltd.

MICHAEL L. McCLUGGAGE, ESQUIRE
BETH L. FANCSALI, ESQUIRE
Wildman Harrold Allen & Dixon LLP
225 West Wacker Drive, Suite 2800
Chicago, Illinois 60606
Attorneys for Appellees,
CNA Financial Corp.; The Continental Insurance Co.;
Continental Casualty Co.; American Casualty Co. of
Reading, PA

LAZAR P. RAYNAL, ESQUIRE
McDermott Will & Emery LLP
227 West Monroe Street, Suite 5200
Chicago, Illinois 60606
Attorney for Appellees,
Chicago Insurance Co.; Fireman’s Fund Insurance
Company; National Surety Corp.

8
PETER R. BISIO, ESQUIRE
Hogan Lovells US LLP
555 13th Street, N.W.
Washington, D.C. 20004
Attorney for Appellees,
The Chubb Corporation; Federal Insurance Company;
Executive Risk Indemnity Inc.; Vigilant Insurance
Company

LOUIS G. CORSI, ESQUIRE
Landman Corsi Ballaine & Ford P.C.
120 Broadway, 27th Floor
New York, New York 10271-0079
Attorney for Appellees,
Crum & Forster Holdings Corp.;
United States Fire Insurance Company

JOHN L. THURMAN, ESQUIRE
Farrell & Thurman PC
172 Tamarack Circle
Skillman, New Jersey 08558

9
ROBERT A. ALESSI, ESQUIRE
Cahill Gordon & Reindel LLP
Eighty Pine Street
New York, New York 10005-1702
Attorneys for Appellees,
Greenwich Insurance Company; Indian Harbor Insurance
Company; XL Capital Ltd.; X.L. America, Inc.; XL
Insurance America, Inc.

ALAN L. KILDOW, ESQUIRE
SONYA R. BRAUNSCHWEIG, ESQUIRE
JAROD M. BONA, ESQUIRE
DLA Piper US LLP
90 South Seventh Street, Suite 5100
Minneapolis, Minnesota 55402
Attorneys for Appellees,
Wells Fargo & Co.; Acordia, Inc.

JONATHAN M. WILAN, ESQUIRE
Hunton & Williams LLP
1900 K Street, N.W., Suite 1200
Washington, D.C. 20006

JOHN J. GIBBONS, ESQUIRE
MICHAEL R. GRIFFINGER, ESQUIRE
Gibbons P.C.
One Gateway Center
Newark, New Jersey 07102

10
 Attorneys for Appellee,
Hilb, Rogal & Hobbs Company

RICHARD C. PEPPERMAN II, ESQUIRE
Sullivan & Cromwell LLP
125 Broad Street
New York, New York 10004
Attorney for Appellees,
Willis Group Holdings Limited; Willis Group Limited;
Willis North America, Inc.; Willis of New York, Inc.;
Willis of Michigan, Inc.

KEVIN J. FEE, ESQUIRE
Kornstein Veisz Wexler & Pollard LLP
757 Third Avenue, 18th Floor
New York, New York 10017

BRIAN E. ROBISON, ESQUIRE
Vinson & Elkins LLP
Trammell Crow Center
2001 Ross Avenue, Suite 3700
Dallas, Texas 75201
Attorneys for Appellees,
Liberty Mutual Holding Company, Inc.; Liberty Mutual
Insurance Co.; Liberty Mutual Fire Insurance Co.;
Wausau Underwriters Insurance Co.; Employers
Insurance Co. of Wausau; Wausau Business Insurance
Co.; Wausau General Insurance Co.

11
MICHAEL J. GARVEY, ESQUIRE
PAUL C. CURNIN, ESQUIRE
DAVID ELBAUM, ESQUIRE
BRYCE L. FRIEDMAN, ESQUIRE
Simpson Thacher & Bartlett LLP
425 Lexington Avenue
New York, New York 10017
Attorneys for Appellees,
The Travelers Companies, Inc.; St. Paul Fire and Marine
Insurance Company; Gulf Insurance Company; St. Paul
Mercury Insurance Company; Travelers Casualty and
Surety Company of America; The Travelers Indemnity
Company; Athena Assurance Company; Travelers
Property Casualty Corp.

HENRY WEISBURG, ESQUIRE
Shearman & Sterling LLP
599 Lexington Avenue
New York, New York 10022
Attorney for Appellees,
Munich Reinsurance

MICHAEL M. MADDIGAN, ESQUIRE
PAUL B. SALVATY, ESQUIRE
O'Melveny & Myers LLP
400 South Hope Street, 15th Floor
Los Angeles, California 90071

12
SAMUEL P. MOULTHROP, ESQUIRE
Riker Danzig Scherer Hyland & Perretti LLP
Headquarters Plaza
One Speedwell Avenue
Morristown, New Jersey 07962
Attorneys for Appellees,
Life Insurance Company of North America; Connecticut
General Life Insurance Company

JAMES W. CARBIN, ESQUIRE
Duane Morris LLP
744 Broad Street, Suite 1200
Newark, New Jersey 07102
Attorney for Appellees,
MetLife, Inc.; Metropolitan Life Insurance Company;
Paragon Life Insurance Company; General American
Life Insurance Company; New England Life Insurance
Company; Citicorp Life Insurance Company; Travelers
Life and Annuity Company; Travelers Insurance
Company; Reinsurance Group of America, Inc.

EDWARD G. BIESTER III, ESQUIRE
JEFFREY S. POLLACK, ESQUIRE
Duane Morris LLP
30 South 17th Street
Philadelphia, Pennsylvania 19103
Attorneys for Appellees,
MetLife, Inc.; Metropolitan Life Insurance Company;

13
 Paragon Life Insurance Company

DOUGLAS S. EAKELEY, ESQUIRE
JOHN R. MIDDLETON, ESQUIRE
MATTHEW SAVARE, ESQUIRE
SCOTT L. WALKER, ESQUIRE
Lowenstein Sandler PC
65 Livingston Avenue
Roseland, New Jersey 07068
Attorneys for Appellees,
Prudential Financial, Inc.;
The Prudential Insurance Company of America

PATRICK W. SHEA, ESQUIRE (ARGUED)
Paul Hastings Janofsky & Walker LLP
75 East 55th Street
New York, New York 10022

STEVEN P. DEL MAURO, ESQUIRE
McElroy Deutsch Mulvaney & Carpenter LLP
Three Gateway Center
100 Mulberry Street
Newark, New Jersey 07102
Attorneys for Appellees,
The Unum Group Corporation; Unum Life Insurance
Company of America; Provident Life and Accident
Insurance Company

14
STEPHEN P. YOUNGER, ESQUIRE
LAURA J. WOOD, ESQUIRE
Patterson Belknap Webb & Tyler LLP
1133 Avenue of the Americas
New York, New York 10036
Attorneys for Appellees,
Universal Life Resources; ULR Insurance Services Inc.;
Benefits Commerce; Douglas P. Cox

RACHEL L. GERSTEIN, ESQUIRE
ROBERT H. PEES, ESQUIRE
Akin Gump Strauss Hauer & Feld LLP
One Bryant Park
New York, New York 10036
Attorneys for Appellees,
USI Holdings Corporation; USI Consulting Group, Inc.;
USI Services Corporation f/k/a USI Insurance Services
Corp.

KEVIN P. RODDY, ESQUIRE
Wilentz Goldman & Spitzer, P.A.
90 Woodbridge Center Drive, Suite 900
Woodbridge, New Jersey 07095
Attorney for Amicus Curiae-Appellants at 07-4046,
National Association of Shareholder and Consumer
Attorneys (NASCAT)

15
EUGENE R. ANDERSON, ESQUIRE
Anderson Kill & Olick, P.C.
1251 Avenue of the Americas
New York, New York 10020
Attorney for Amicus Curiae-Appellants at 07-4046,
United Policyholders

OPINION OF THE COURT

SCIRICA, Circuit Judge.

This appeal from orders of dismissal under Federal Rule
of Civil Procedure 12(b)(6) involves multiple putative class
actions alleging massive conspiracies throughout the insurance
industry. Plaintiffs are purchasers of commercial and employee
benefit insurance, and defendants are insurers and insurance
brokers that deal in those lines of insurance. According to
plaintiffs, defendants entered into unlawful, deceptive schemes
to allocate purchasers among particular groups of defendant
insurers. The complaints assert that conspiring brokers funneled
unwitting clients to their co-conspirator insurers, which were
insulated from competition; in return, the insurers awarded the
brokers contingent commission payments—concealed from the
insurance purchasers and surreptitiously priced into insurance
premiums—based on the volume of premium dollars steered
their way. As a result of this scheme, plaintiffs allege they paid

16
inflated prices for their insurance coverage and were generally
denied the benefits of a competitive market. The question on
appeal is whether plaintiffs have adequately pled either a per se
violation of § 1 of the Sherman Act (plaintiffs have foresworn
a full-scale rule-of-reason analysis) or a violation of the
Racketeer Influenced and Corrupt Organizations (RICO) Act.
Concluding they had not, the District Court dismissed the
complaints. We will affirm in large part, vacate in part, and
remand for further proceedings.

I. Procedural History and Plaintiffs’ Allegations

This litigation followed on the heels of a public
investigation and enforcement action. In October 2004, the New
York State Attorney General filed a civil complaint in state court
against insurance broker Marsh & McLennan (“Marsh”),
alleging “that Marsh had solicited rigged bids for insurance
contracts, and had received improper contingent commission
payments in exchange for steering its clients to a select group of
insurers.” In re Ins. Brokerage Antitrust Litig., Nos. 04-5184,
05-1079, 2006 WL 2850607, at *1 (D.N.J. Oct. 3, 2006) (citing
People v. Marsh & McLennan Cos., No. 04/403342 (N.Y. Sup.
Ct. Oct. 14, 2004)). The next month, a group of attorneys
general and state insurance departments began a broader
investigation of insurance-industry practices. Private parties
also filed numerous federal actions, which are the subject of this
appeal.

The private actions were transferred by the Judicial Panel

17
on Multidistrict Litigation to the United States District Court for
the District of New Jersey for consolidated pretrial proceedings.
In re Ins. Brokerage Antitrust Litig., 360 F. Supp. 2d 1371
(J.P.M.L. 2005); see 28 U.S.C. § 1407. The District Court
severed and realigned the actions into two consolidated
dockets—the first pertaining to claims regarding property and
casualty insurance (the “Commercial Case”), and the second
pertaining to claims regarding employee benefits insurance (the
“Employee Benefits Case”).

The plaintiffs in the Commercial Case are a
proposed class of businesses, individuals, and
public entities who, between August 26, 1994 and
September 1, 2005, engaged the services of the
Broker Defendants to obtain advice with respect
to the procurement or renewal of commercial
property and casualty insurance and entered into
or renewed an insurance policy with the Insurer
Defendants. The plaintiffs in the Employee
Benefits Case are both employers who utilized the
services of the Broker Defendants to obtain group
insurance coverage from the Insurer Defendants
for their employees as part of their employee
benefits plans and employees who obtained
insurance from the Insurer Defendants through
the employers’ benefits plans.

18
In re Ins. Brokerage Antitrust Litig., 579 F.3d 241, 249 (3d Cir.
2009) (affirming, inter alia, the District Court’s approval of
plaintiffs’ settlement agreements with two defendants in this
litigation).1

In accordance with the District Court’s restructuring,
plaintiffs filed a separate consolidated amended complaint in
each of the Commercial and Employee Benefits cases. Each
complaint alleged violations of the Sherman Act, 15 U.S.C. § 1,
and the RICO Act, 18 U.S.C. § 1962(c), (d), as well as
violations of various state-law antitrust statutes and common-
law duties. Shortly thereafter, defendants moved to dismiss the
Sherman Act and RICO claims in both cases under Federal Rule

1
This statement paraphrases the description of the proposed
plaintiff classes given by the District Court. See In re Ins.
Brokerage Antitrust Litig., 2007 WL 2892700, at *2 (D.N.J.
Sept. 28, 2007). The complaints, however, arguably define the
proposed classes to include not only those persons or entities
who bought insurance from a defendant insurer through a
defendant broker, but also those persons or entities who bought
insurance from any insurer through a defendant broker. See
Commercial Case Second Amended Complaint (Comm. SAC)
¶ 555; Employee Benefits Case Second Amended Complaint
(EB SAC) ¶ 585. This discrepancy is not relevant to our
disposition of this appeal.

19
of Civil Procedure 12(b)(6).2

On October 3, 2006, the District Court granted the
motions and dismissed the claims without prejudice. Ins.
Brokerage, 2006 WL 2850607. Defendants had asserted in their
moving papers that they were immune from Sherman Act
liability under the McCarran-Ferguson Act, 15 U.S.C. §§
1011–1015, which “provides a statutory antitrust exemption for
activities that (1) constitute the ‘business of insurance,’ (2) are
regulated pursuant to state law, and (3) do not constitute acts of
‘boycott, coercion or intimidation.’” Ticor Title Ins. Co. v. FTC,
998 F.2d 1129, 1133 (3d Cir. 1993) (quoting Group Life &
Health Ins. Co. v. Royal Drug Co., 440 U.S. 205, 219–20
(1979)); see 15 U.S.C. §§ 1012(b), 1013(b). The District Court
rejected this argument on the ground that defendants’ alleged
conduct was not part of the “business of insurance” within the
meaning of the Act. 2006 WL 2850607, at *7–10. But the court
nonetheless dismissed both the Sherman Act and RICO claims
because it found the complaints lacked the requisite factual
specificity.

In granting leave to amend, the District Court instructed
plaintiffs to file in each case a supplemental statement of

2
In the Employee Benefits Case only, plaintiffs also brought
claims under the Employee Retirement Income Security Act of
1974 (ERISA), 29 U.S.C. § 1132(a)(2), alleging defendants had
breached fiduciary duties imposed by the statute. These claims
are not before us. See infra note 3.

20
particularity for their federal antitrust claims and an amended
RICO case statement for their RICO claims. Plaintiffs did so,
and defendants again moved to dismiss. On April 5, 2007, the
District Court again granted the motions, but it once again
allowed plaintiffs an opportunity to amend their pleadings. In
re Ins. Brokerage Antitrust Litig., 2007 WL 1100449 (D.N.J.
Apr. 5, 2007) (antitrust claims); In re Ins. Brokerage Antitrust
Litig., 2007 WL 1062980 (D.N.J. Apr. 5, 2007) (RICO claims).
In response, plaintiffs filed a Second Amended Complaint
(“SAC”) in each of the Commercial and Employee Benefits
cases, as well as a Revised Particularized Statement (“RPS”)
and Amended RICO Case Statement (“ARCS”) augmenting the
Second Amended Complaint’s allegations. For a third time,
defendants moved to dismiss under Rule 12(b)(6). In orders
dated August 31 and September 28, 2007, the District Court
again dismissed the antitrust and RICO claims—this time with
prejudice. Applying the pleading standard set forth by the
Supreme Court in Bell Atlantic Corp. v. Twombly, 550 U.S. 544
(2007), which had been decided on May 21, 2007, the District
Court concluded that plaintiffs’ allegations in both the
Commercial and the Employee Benefits cases were insufficient
with respect to both the Sherman Act and RICO claims. In re
Ins. Brokerage Antitrust Litig., 2007 WL 2533989 (D.N.J. Aug.
31, 2007) (antitrust claims); In re. Ins. Brokerage Antitrust
Litig., 2007 WL 2892700 (D.N.J. Sept. 28, 2007) (RICO

21
claims). Plaintiffs filed a timely notice of appeal in each case.3

Plaintiffs’ pleadings are of a substantial volume. The
complaint in each case is more than 200 pages (including
attached exhibits), and to this total must be added the pages in

3
The District Court exercised jurisdiction under 28 U.S.C. §§
1331, 1367. We have jurisdiction under 28 U.S.C. § 1291. The
Sherman Act and RICO claims were the only federal causes of
action asserted in the Commercial complaint. Having dismissed
both claims in its August 31 and September 28 opinions, the
District Court declined to exercise supplemental jurisdiction
over the remaining state-law claims and dismissed the
Commercial complaint in its entirety. 2007 WL 2892700, at
*34; see 28 U.S.C. § 1367(c). The Employee Benefits
complaint also included claims that the insurer defendants had
breached their fiduciary duties under ERISA. The District Court
subsequently disposed of these ERISA claims when it granted
defendants’ motion for summary judgment, In re Ins. Brokerage
Antitrust Litig., 2008 WL 141498 (D.N.J. Jan. 14, 2008), after
which it declined to exercise supplemental jurisdiction over the
state-law claims and dismissed the Employee Benefits complaint
in its entirety, In re Ins. Brokerage Antitrust Litig., No. 05-1079
(D.N.J. Feb. 13, 2008).
Although plaintiffs originally appealed the District
Court’s summary judgment order regarding the ERISA claims,
they expressly waived that issue in their opening brief.
Plaintiffs’ Employee Benefits (EB) Br. 10.

22
the Revised Particularized Statements and Amended RICO Case
Statements. Significantly, the District Court allowed discovery
to proceed while the motions to dismiss were pending.
Plaintiffs’ amended pleadings were thus able to draw on
documents produced and depositions taken pursuant to these
discovery orders, as well as material unearthed in the course of
the public investigations.

As reflected by the length of this opinion—and of the
caption—this is extraordinarily complex litigation involving a
large swath of the insurance provider and brokerage industries,
elaborate allegations of misconduct, and challenging legal
issues. The District Court skillfully managed the consolidated
proceedings. We take particular note of the court’s thorough
treatment of defendants’ motions to dismiss, which comprised
five separate opinions examining three successive rounds of
pleadings. The court’s patient and meticulous analysis has
greatly aided our review.

A. Antitrust Claims

1. Broker-Centered Conspiracies

In each complaint, plaintiffs allege the existence of a
number of broker-centered antitrust conspiracies. As the name
suggests, at the center of each alleged conspiracy was a
defendant broker, who colluded with its defendant insurer-
partners to steer its clients, purchasers of insurance, to particular
insurers in exchange for the payment of contingent
commissions. In the Commercial Case, plaintiffs allege six such

23
conspiracies, centered on defendant brokers Marsh,4 Aon
Corporation, Wells Fargo & Company, HRH, Willis Group, and
Gallagher, respectively. In the Employee Benefits Case,
plaintiffs allege five broker-centered conspiracies, led
respectively by Marsh,5 Aon, Universal Life Resources,

4
While this appeal was pending, the District Court approved
a settlement agreement between the plaintiffs and the Marsh
Defendants. (In the Commercial Case, the Marsh Defendants
comprise Marsh & McLennan Companies, Inc.; Marsh Inc.;
Marsh USA, Inc.; Marsh USA Inc. (Connecticut); and Seabury
& Smith, Inc. In the Employee Benefits Case, the Marsh
Defendants comprise Marsh & McLennan Companies, Inc.;
Marsh Inc.; Marsh USA, Inc.; Marsh USA Inc. (Connecticut);
Mercer, Inc.; Mercer Human Resource Consulting LLC; Mercer
Human Resource Consulting of Texas, Inc.; and Seabury &
Smith, Inc.) See In re Ins. Brokerage Antitrust Litig., No. 04-
5184, 2009 WL 411877 (D.N.J. Feb. 17, 2009), appeal
docketed, No. 09-1821 (3d Cir. Mar. 30, 2009). Per the settling
parties’ joint motions, we dismissed the instant appeal as to the
Marsh Defendants without prejudice. See In re Ins. Brokerage
Antitrust Litig., Nos. 07-4046, 08-1455, 08-1777 (3d Cir. June
30, 2008).
5
See supra note 4.

24
Gallagher, and Willis Group.6

According to the complaints, the broker-centered
conspiracies proceeded in two stages. First, “[b]eginning in the
mid-to-late 1990s, each of the Broker Defendants,” in “a
dramatic change” from prior practice, “began to form so-called
‘strategic partnerships’ with certain insurance companies, to
which it would then allocate the bulk of its business.” Comm.
SAC ¶ 83. The broker and each of its co-conspiring insurers
agreed, “and each of the conspiring insurers horizontally
agreed,” that the broker “would ‘consolidate’ its business by
directing the bulk of its premium volume to its ‘strategic
partner’ co-conspirators, thereby eliminating hundreds of other

6
The defendant-broker names given here encompass related
and/or subsidiary companies, as detailed in Comm. SAC ¶¶
24–36 and EB SAC ¶¶ 34–45. The names of the defendant
insurers allegedly conspiring with each broker can also be found
in the complaints; in the interest of brevity, we will not
reproduce them here. See Comm. SAC ¶¶ 95, 157, 201, 236,
262, 326; EB SAC ¶¶ 106, 139, 175, 239, 271. The number of
insurers in each alleged broker-centered conspiracy ranges from
three to thirteen. (These numbers refer to parent entities and do
not include the subsidiary/related companies also named as
defendants. See Comm. SAC ¶¶ 37–63 (listing
subsidiary/related insurers in the Commercial Case); EB SAC ¶¶
46–57 (listing subsidiary/related insurers in the Employee
Benefits Case).)

25
insurers from competing equally with the conspiring insurers for
the majority” of the broker’s business. Id. ¶ 158. In the second
stage, the insurer-members of each conspiracy each agreed with
the broker, “and agreed horizontally among themselves, to
reduce or eliminate competition for that secured business among
the conspiring [‘strategic partner’] insurers.” Id.7

As alleged by plaintiffs, a major focus of the second stage
of the conspiracies was protecting the incumbent business of
each insurer. To maximize insurers’ retention of existing
customers, the conspiracies allegedly employed a variety of
“incumbent protection devices.” Specifically, plaintiffs aver
that brokers facilitated the non-competitive allocation of
customers to insurers by giving insurers “last looks” and “first
looks” on bids.8 The complaints also assert that each insurer in
each broker-centered conspiracy knew the identity of the

7
The quoted language is drawn from the description of the
Aon-centered conspiracy in the Commercial complaint but is
generally applicable to all of the alleged broker-centered
conspiracies. See Comm. SAC ¶¶ 66–67; EB SAC ¶¶ 76–77.
8
Plaintiffs do not specifically define these terms, but from
context we infer that a “last look” affords a bidder the ability to
make the final bid with knowledge of all previous bids, and a
“first look” allows a bidder the opportunity to bid without
competition (for example, guaranteeing a sale to the bidder if it
can match a certain price).

26
broker’s other “strategic partners.” The brokers also revealed to
each insurer detailed information about the arrangements
between the broker and its other insurer-partners, including
information about the size of the contingent commissions those
partners were paying to the broker, and even the amount of
premium volume steered by the broker to the other insurers.
These facts, plaintiffs contend, evince the existence of an
agreement between the insurers and the broker—and among the
insurers themselves—to reap inflated profits by stifling
competitive bidding and protecting incumbent business, in
violation of § 1 of the Sherman Act.

These incumbent protection devices, plaintiffs claim,
were common to all of the broker-centered conspiracies. In the
Commercial Case only, plaintiffs also allege that insurers in the
Marsh-centered conspiracy acceded to broker requests to
provide “false” bids that were intentionally higher than the bids
of the insurer to which the broker wished to award the business.
For example, the complaint relates a statement by a former
employee of a defendant insurer that his employer had agreed to
“provide[] losing quotes” to its broker-partner in exchange for,
among other things, the broker’s “getting ‘quotes from other
[insurance] carriers that would support the [employer, at least
when it was the incumbent carrier] as being the best price.’” Id.
¶ 109. The employee of another insurer allegedly stated that
“she provided protective quotes when the broking plan called
for it ‘[t]o show, to pretend to show competition where there is
none.’” Id. ¶ 119. This employee was allegedly told by the

27
broker that the insurer “should provide protective quotes so that
[it] would not face competition on its own renewals.” Id. This
bid-rigging behavior facilitated the customer allocation scheme
by deceiving insurance customers into believing they were
receiving the best possible price in a competitive market.
According to plaintiffs, insurers were willing to assist co-
conspiring insurers in this way because they expected to be the
beneficiary of such bid rigging where their own incumbent
business was concerned.

2. Global Conspiracy

In addition to the broker-centered conspiracies, each
complaint alleges a “global conspiracy” among all of the
defendants: “[W]hile engaging in their separate ‘hub and spoke’
schemes [i.e., the broker-centered conspiracies] to create supra-
competitive premiums and contingent commissions, each of the
Broker ‘hubs’ simultaneously agreed horizontally not to
compete with each other by disclosing any competing broker’s
contingent commission arrangements, or the consequent
premium price impact of those arrangements, in an effort to win
those brokers’ customers’ business.” Id. ¶ 354. Although each
broker, plaintiffs claim, knew that the other brokers were using
contingent commission arrangements to obtain outsized profits,
each “also knew that exposing another broker’s contingent
commission arrangements to the other broker’s customers would
lead to retaliation, thereby threatening the first broker’s own
contingent commission scheme and supra-competitive profits.”
Id. ¶ 355. “Therefore,” plaintiffs allege, the brokers “agreed

28
horizontally” to maintain a mutually beneficial silence. Id.
¶ 362. Plaintiffs further allege that the defendant insurers were
“complicit[]” in this horizontal agreement among the brokers,
id. ¶ 353, and that they also agreed “horizontally with each
other[] not to disclose the Broker-Centered Conspiracies and
resulting supra-competitive premiums to the brokers’
customers,” id. ¶ 359.

As evidence of this asserted “global” agreement in the
Commercial Case, plaintiffs point to allegations that each
broker-centered conspiracy operated in a similar way and that
the brokers incorporated similar standardized confidentiality
provisions into their respective contingent commission
agreements with insurers, which prohibited disclosing the terms
of the contingent commission agreements to insurance
customers. Furthermore, plaintiffs allege that the brokers’
membership in the Council of Insurance Agents & Brokers
(CIAB), a trade association, “afforded them many opportunities
to exchange information and allowed Defendants to adopt
collective policies towards nondisclosure of rival brokers’
contingent commissions.” Id. ¶ 364.

Plaintiffs in the Employee Benefits Case also rely on
these types of allegations to support their claim of a global
conspiracy. They find additional support, however, in the
similar way in which insurers, at the alleged behest of the
brokers, accounted for the expense of contingent commissions
on Schedule A of Form 5500, a document that must, under
ERISA, be filed with the Internal Revenue Service and the

29
Department of Labor. According to plaintiffs, instead of
reporting the commissions as “a variable, case-specific cost,”
insurers treated them “improperly as a non-reportable fixed cost
(overhead).” EB SAC ¶ 305. This reporting technique allegedly
yielded two advantages to defendants. First, they “were enabled
to evade their disclosure requirements under ERISA and mislead
their clients.” Id. Second, by classifying contingent
commissions as a fixed cost spread across all lines of an
insurer’s business, “the Insurer Defendants artificially raised the
price of all lines of insurance, rather than substantially raising
the cost of insurance” obtained through the co-conspiring
brokers, which would have rendered that insurance blatantly
uncompetitive with insurance obtained through other, non-
conspiring brokers. Id. ¶ 306. Not only, plaintiffs allege, did
defendants adopt a similar approach to accounting for the
contingent commission agreements, but employees of the
defendants also sometimes exchanged information about how
they completed Form 5500. Plaintiffs claim these allegations
support an inference of an agreement not to disclose contingent
commissions properly in order to conceal the existence of
defendants’ anticompetitive practices.

B. RICO Claims

Plaintiffs contend that defendants’ alleged customer
allocation schemes also violated the RICO statute. In the
Commercial Case, plaintiffs assert the existence of six RICO
enterprises, which correspond to the six broker-centered
conspiracies identified in the antitrust claims. “Alternatively,

30
Plaintiffs allege that CIAB is a legal entity which constitutes a
RICO enterprise . . . .” Comm. SAC ¶ 512. According to the
complaint, the defendants utilized these enterprises to engage in
a pattern of racketeering activity consisting of numerous acts of
mail and wire fraud that served to conceal and misrepresent
defendants’ customer allocation schemes.

The Employee Benefit complaint alleges similar
predicate acts of racketeering and adds allegations that
defendants misrepresented information reported on Form 5500
and otherwise violated ERISA through their use of contingent
commissions. Here, plaintiffs allege the existence of five RICO
enterprises congruent with the five alleged broker-centered
antitrust conspiracies.

II. Discussion

We exercise plenary review of the District Court’s orders
granting defendants’ motions to dismiss under Federal Rule of
Civil Procedure 12(b)(6). See Gelman v. State Farm Mut. Auto.
Ins. Co., 583 F.3d 187, 190 (3d Cir. 2009). This Rule authorizes
dismissal of a complaint for “failure to state a claim upon which
relief can be granted.” Fed. R. Civ. P. 12(b)(6). Under Rule
8(a)(2), a complaint need present “only ‘a short and plain
statement of the claim showing that the pleader is entitled to
relief,’ in order to ‘give the defendant fair notice of what the . . .
claim is and the grounds upon which it rests.’” Twombly, 550
U.S. at 555 (quoting Conley v. Gibson, 355 U.S. 41, 47 (1957))
(omission in Twombly); see Fed. R. Civ. P. 8(a)(2). To comply

31
with this general pleading standard, the complaint, “construe[d]
. . . in the light most favorable to the plaintiff,” Gelman, 583
F.3d at 190 (quoting Phillips v. County of Allegheny, 515 F.3d
224, 233 (3d Cir. 2008)), must contain “‘enough factual matter
(taken as true) to suggest’ the required element[s]” of the claims
asserted, Phillips, 515 F.3d at 234 (quoting Twombly, 550 U.S.
at 556).

A. Antitrust Claims

1. Plausibility Under Twombly

a. Legal Standards

Section 1 of the Sherman Act provides: “Every contract,
combination in the form of trust or otherwise, or conspiracy, in
restraint of trade or commerce among the several States, or with
foreign nations, is declared to be illegal.” 15 U.S.C. § 1. As we
have explained, this statutory language imposes two essential
requirements on an antitrust plaintiff.9 “First, the plaintiff must
show that the defendant was a party to a ‘contract, combination

9
In addition to the following two requirements, the plaintiffs
in any antitrust case “must prove antitrust injury, which is to say
(1) injury of the type the antitrust laws were intended to prevent
and (2) that flows from that which makes defendants’ acts
unlawful.” A.D. Bedell Wholesale Co. v. Phillip Morris Inc.,
263 F.3d 239, 247 (3d Cir. 2001) (quoting Brunswick Corp. v.
Pueblo Bowl-O-Mat, 429 U.S. 477, 489 (1997) (emphasis
omitted)).

32
. . . or conspiracy.’” Toledo Mack Sales & Serv., Inc. v. Mack
Trucks, Inc., 530 F.3d 204, 218 (3d Cir. 2008). Instead of
assigning each of these last three terms a distinct meaning,
courts have interpreted them collectively to require “some form
of concerted action,” In re Baby Food Antitrust Litig., 166 F.3d
112, 117 (3d Cir. 1999) (internal quotation marks omitted), in
other words, a “‘unity of purpose or a common design and
understanding or a meeting of minds’ or ‘a conscious
commitment to a common scheme,’” In re Flat Glass Antitrust
Litig., 385 F.3d 350, 357 (3d Cir. 2004) (quoting Monsanto Co.
v. Spray-Rite Serv. Corp., 465 U.S. 752, 764 (1984)). Put more
succinctly, “[t]he existence of an agreement is the hallmark of
a Section 1 claim.” Baby Food, 166 F.3d at 117; see InterVest,
Inc. v. Bloomberg, L.P., 340 F.3d 144, 159 (3d Cir. 2003)
(“Unilateral activity by a defendant, no matter the motivation,
cannot give rise to a section 1 violation.”).10

10
“Congress used th[e] distinction between concerted and
independent action to deter anticompetitive conduct and
compensate its victims, without chilling vigorous competition
through ordinary business operations. . . . [U]nlike independent
action, ‘[c]oncerted activity inherently is fraught with
anticompetitive risk’ insofar as it ‘deprives the marketplace of
independent centers of decisionmaking that competition
assumes and demands.’” Am. Needle, Inc. v. NFL, 130 S. Ct.
2201, 2209 (2010) (quoting Copperweld Corp. v. Independence
Tube Corp., 467 U.S. 752, 768–69 (1984)).

33
 In addition to demonstrating the existence of a
“conspiracy,” or agreement, “the plaintiff must show that the
conspiracy to which the defendant was a party imposed an
unreasonable restraint on trade.” 11 Mack Trucks, 530 F.3d at
218; see Flat Glass, 385 F.3d at 356 (“Despite its broad
language, Section 1 only prohibits contracts, combinations or
conspiracies that unreasonably restrain trade.”). “[T]he usual
standard” applied to determine whether a challenged practice
unreasonably restrains trade is the so-called “rule of reason.”
Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S.
877, 882 (2007). Under this standard, “the factfinder weighs all
of the circumstances of a case in deciding whether a restrictive
practice should be prohibited.” Mack Trucks, 530 F.3d at 225
(internal quotation marks omitted). Significantly, under a rule-
of-reason analysis, the plaintiff “bears the initial burden of
showing that the alleged [agreement] produced an adverse,
anticompetitive effect within the relevant geographic market.”
Gordon v. Lewistown Hosp., 423 F.3d 184, 210 (3d Cir. 2005).
Because of “the difficulty of isolating the [actual] market effects
of challenged conduct,” United States v. Brown Univ., 5 F.3d
658, 668 (3d Cir. 1993), successful attempts to meet this burden
typically include a demonstration of defendants’ market power,
as “a judgment about market power is [a] means by which the

11
“The question whether an arrangement is a contract,
combination, or conspiracy is different from and antecedent to
the question whether it unreasonably restrains trade.” Am.
Needle, 130 S. Ct. at 2206.

34
effects of the [challenged] conduct on the market place can be
assessed,” NCAA v. Bd. of Regents of Univ. of Okla., 468 U.S.
85, 110 n.42 (1984) (quoting the Solicitor General’s “correct[]”
observation). Cf. FTC v. Ind. Fed’n of Dentists, 476 U.S. 447,
460–61 (1986) (“Since the purpose of the inquiries into market
definition and market power is to determine whether an
arrangement has the potential for genuine adverse effects on
competition, proof of actual detrimental effects, such as a
reduction of output, can obviate the need for an inquiry into
market power, which is but a surrogate for detrimental effects.”
(internal quotation marks omitted)). If the plaintiff carries this
burden, the court will need to decide whether the
anticompetitive effects of the practice are justified by any
countervailing pro-competitive benefits.12 See Eichorn v. AT&T
Corp., 248 F.3d 131, 143 (3d Cir. 2001) (describing an analysis
in which courts “balance the effect of the alleged anti-
competitive activity against its competitive purposes within the

12
In the event a genuinely disputed issue of fact exists
regarding the reasonableness of the restraint, the determination
is for the jury. See Arizona v. Maricopa County Med. Soc’y,
457 U.S. 332, 343 (1982) (“[T]he rule of reason requires the
factfinder to decide whether under all the circumstances of the
case the restrictive practice imposes an unreasonable restraint on
competition.”); 11 Herbert Hovenkamp, Antitrust Law ¶ 1909b
(2d ed. 2005) (“[O]nce the court decide[s] that the rule of reason
should apply, disputed factual questions about reasonableness
should be left to the jury.”).

35
relevant product and geographic markets”); see also Leegin, 551
US. at 886 (“In its design and function the rule [of reason]
distinguishes between restraints with anticompetitive effect that
are harmful to the consumer and restraints stimulating
competition that are in the consumer’s best interest.”).

Judicial experience has shown that some classes of
restraints have redeeming competitive benefits so rarely that
their condemnation does not require application of the full-
fledged rule of reason. Paradigmatic examples are “horizontal
agreements among competitors to fix prices or to divide
markets.” Leegin, 551 U.S. at 886 (citations omitted). Once a
practice has been found to fall into one of these classes, it is
subject to a “per se” standard. As the Supreme Court has
explained, these practices

are ordinarily condemned as a matter of law under
an “illegal per se” approach because the
probability that these practices are anticompetitive
is so high; a per se rule is applied when “the
practice facially appears to be one that would
always or almost always tend to restrict
competition and decrease output.” In such
circumstances a restraint is presumed
unreasonable without inquiry into the particular
market context in which it is found.

NCAA, 468 U.S. at 100 (quoting Broad. Music, Inc. v. Columbia
Broad. Sys., Inc., 441 U.S. 1, 19–20 (1979)); see Brown Univ.,

36
5 F.3d at 670 (“Per se rules of illegality are judicial constructs,
and are based in large part on economic predictions that certain
types of activity will more often than not unreasonably restrain
competition.” (internal citation omitted)). Under the per se
standard, plaintiffs are relieved of the obligation to define a
market and prove market power. See Copperweld Corp. v.
Independence Tube Corp., 467 U.S. 752, 768 (1984) (citing N.
Pac. Ry. Co. v. United States, 356 U.S. 1, 5 (1958)); Rossi v.
Standard Roofing, Inc., 156 F.3d 452, 464–65 (3d Cir. 1998); 11
Herbert Hovenkamp, Antitrust Law ¶ 1910a (2d ed. 2005); see
also 7 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law
¶ 1509c, at 403–04 (2d ed. 2003) (“Little is lost when the court
condemns a restraint that was harmless because the defendants
lacked power but that was socially useless in any event.”). Once
a defendant’s practice has been found to fall into one of the
recognized classes, it is “conclusively presumed to unreasonably
restrain competition.” Flat Glass, 385 F.3d at 356 (internal
quotation marks omitted). 13

13
When evaluating tying arrangements, in which a firm
“sell[s] one good (the tying product) on the condition that the
buyer also purchase another, separate good (the tied product),”
Town Sound & Custom Tops, Inc. v. Chrysler Motors Corp., 959
F.2d 468, 475 (3d Cir. 1992) (en banc), courts have applied a
modified version of the per se standard. Unlike the “truly per se
rules” explicated above, in which no inquiry is made into market
structure, actual anticompetitive effects, or possible
justifications, “[t]he ‘per se’ rule against tying goes only

37
 While pleading exclusively per se violations can lighten
a plaintiff’s litigation burdens, it is not a riskless strategy. If the
court determines that the restraint at issue is sufficiently
different from the per se archetypes to require application of the
rule of reason, the plaintiff’s claims will be dismissed. E.g.,
AT&T Corp. v. JMC Telecom, LLC, 470 F.3d 525, 531 (3d Cir.
2006); see also Texaco Inc. v. Dagher, 547 U.S. 1, 7 n.2 (2006)
(declining to conduct a rule of reason analysis where plaintiffs
“ha[d] not put forth a rule of reason claim”). See generally 11
Hovenkamp, supra, ¶ 1910b (discussing the cost-benefit
analysis involved in deciding whether to pursue an exclusively
per se theory of liability).

Some restraints of trade are “highly suspicious” yet

halfway . . .: the inquiry into tying product market structure . . .
is still required, but if the defendant is found to have market
power there, the plaintiff is, in theory, relieved of proving actual
harm to competition and of rebutting justifications for the tie-
in.” Id. at 477; see U.S. Healthcare, Inc. v. Healthsource, Inc.,
986 F.2d 589, 593 n.2 (1st Cir. 1993) (stating that tying might
better be described as a “quasi” per se offense, “since some
element of [market] power must be shown and defenses are
effectively available”) (citing Eastman Kodak Co. v. Image
Technical Servs., Inc., 504 U.S. 451 (1992)). See generally 7
Areeda & Hovenkamp, supra, ¶ 1510a (explicating various
different meanings of “per se” language in antitrust
jurisprudence).

38
“sufficiently idiosyncratic that judicial experience with them is
limited.” 11 Hovenkamp, supra, ¶ 1911a. Per se condemnation
is inappropriate, but at the same time, the “inherently suspect”
nature of the restraint obviates the sort of “elaborate industry
analysis” required by the traditional rule-of-reason standard.
Gordon, 423 F.3d at 210. Courts have devised a “quick look”
approach for these cases. See Dagher, 547 U.S. at 7 n.3; Cal.
Dental Ass’n. v. FTC, 526 U.S. 756, 770 (1999) (stating that a
“‘quick-look’ analysis” is appropriate where “an observer with
even a rudimentary understanding of economics could conclude
that the arrangements in question would have an anticompetitive
effect on customers and markets”); see also 11 Hovenkamp,
supra, ¶ 1911a (“What [the ‘quick-look’] term is intended to
connote is that a certain class of restraints, while not
unambiguously in the per se category, may require no more than
cursory examination to establish that their principal or only
effect is anticompetitive.”). Under a quick look analysis, which
is essentially an abbreviated form of the rule of reason, Cal.
Dental, 526 U.S. at 770, “competitive harm is presumed and the
defendant must set forth some competitive justification for the
restraints,” Gordon, 423 F.3d at 210. If no plausible
justification is forthcoming, the restraint will be condemned.
Brown Univ., 5 F.3d at 669. “If the defendant offers sound
procompetitive justifications, however, the court must proceed
to weigh the overall reasonableness of the restraint using a full-

39
scale rule of reason analysis.” Id.14

Here, plaintiffs abjure “a full-scale rule of reason
analysis.” They claim instead that defendants’ behavior was per
se unlawful, or that, at the very least, it is susceptible to
condemnation under a “quick look” analysis. Plaintiffs do not

14
As the above discussion ought to make clear, the respective
analyses conducted under the rule of reason, per se, and quick
look standards are not categorically different. In every case,
“the essential inquiry” is “whether or not the challenged restraint
enhances competition.” Cal. Dental, 526 U.S. at 780 (internal
quotation marks omitted). Under a traditional rule-of-reason
analysis, a court requires “actual market analysis,” id. at 779–80,
and carefully balances all of the factors bearing on that ultimate
question. In applying per se or quick look analysis, courts make
judgments based on judicial experience with certain types of
restraints and market contexts, without demanding such
extensive inquiry into the market in which the specific restraint
at issue operates. But “there is often no bright line separating”
the three standards. Id. at 779 (quoting NCAA, 468 U.S. at 104
n.26). “What is required . . . is an enquiry meet for the case,
looking to the circumstances, details, and logic of a restraint.
The object is to see whether the experience of the market has
been so clear, or necessarily will be, that a confident conclusion
about the principal tendency of a restriction will follow from a
quick (or at least quicker) look, in place of a more sedulous
one.” Id. at 781.

40
dispute that in order to succeed under either of these approaches,
they need to show the existence of a horizontal agreement, that
is, an agreement between “competitors at the same market
level.” In re Pharmacy Benefits Managers Antitrust Litig., 582
F.3d 432, 436 n.5 (3d Cir. 2009); see also Bus. Elecs. Corp. v.
Sharp Elecs. Corp., 485 U.S. 717, 730 (1988) (“Restraints
imposed by agreement between competitors have traditionally
been denominated as horizontal restraints, and those imposed by
agreement between firms at different levels of distribution as
vertical restraints.”). Under the Supreme Court’s jurisprudence,
virtually all vertical agreements now receive a traditional rule-
of-reason analysis. See Leegin, 551 U.S. 877; see also Gordon,
423 F.3d at 210 (rejecting quick look analysis and applying rule
of reason where restraint was vertical).15 In the factual context

15
In Leegin, the Supreme Court overruled its earlier holding
that vertical price-fixing agreements were subject to per se
condemnation. 551 U.S. at 881–82 (overruling Dr. Miles Med.
Co. v. John D. Park & Sons Co., 220 U.S. 373 (1911)). Earlier,
in Continental T.V., Inc. v. GTE Sylvania Inc., the Court had
ruled that non-price vertical restraints must be analyzed under
the traditional rule of reason rather than a per se standard. 433
U.S. 36 (1977); see Leegin, 551 U.S. at 901. Tying
arrangements, however, appear to remain an exception to the
general rule that vertical restraints are reviewed under the full-
scale rule of reason. See supra note 13; Sheridan v. Marathon
Petroleum Co., 530 F.3d 590, 593–94 (7th Cir. 2008)
(explaining that despite a series of Supreme Court decisions

41
of this case, a horizontal agreement means an agreement among
the insurers in the broker-centered conspiracies, and an
agreement among either the brokers or the insurers in the global
conspiracy. Agreements between brokers and insurers, on the
other hand, are vertical and would have to be analyzed under the
traditional rule of reason, which plaintiffs have disclaimed.16

subjecting various vertical restraints to the rule of reason,
including Leegin, tying is still reviewed under a modified per se
standard).
16
Although plaintiffs’ First Amended Complaints (FAC)
expressly pled a rule-of-reason claim in the alternative, see, e.g.,
Comm. FAC ¶ 530; EB FAC ¶ 454, their Second Amended
Complaints omit any reference to the rule of reason, and their
moving papers and appellate arguments make clear they are
alleging exclusively per se violations. In their initial motions to
dismiss, defendants contended that the First Amended
Complaints had not adequately defined a market or pled anti-
competitive effects and had thus failed to state a claim under the
rule of reason. In response, plaintiffs did not assert that they
had, in fact, met these requirements; they argued only that
“where plaintiffs allege per se claims,” these requirements do
not apply. Plaintiffs’ Memorandum of Law in Opposition to
Defendants Motions to Dismiss 43 n.26, filed in the District
Court as No. 04-5184, Dkt. Entry # 344. In a subsequent
submission, plaintiffs explicitly stated that the allegations in
their complaints were “subject to per se antitrust analysis, not

42
 Plaintiffs’ obligation to show the existence of a
horizontal agreement is not only an ultimate burden of proof but
also bears on their pleadings. “[A] plaintiff’s obligation to
provide the ‘grounds’ of his ‘entitle[ment] to relief’ requires
more than labels and conclusions, and a formulaic recitation of
the elements of a cause of action will not do.” Twombly, 550
U.S. at 555 (quoting Fed. R. Civ. P. 8(a)(2)). Because Federal
Rule of Civil Procedure 8(a)(2) “requires a ‘showing,’ rather
than a blanket assertion, of entitlement to relief,” courts
evaluating the viability of a complaint under Rule 12(b)(6) must
look beyond conclusory statements and determine whether the
complaint’s well-pled factual allegations, taken as true, are
“enough to raise a right to relief above the speculative level.”

evaluation under the rule of reason.” Plaintiffs’ Reply Brief in
Support of Motion for Class Certification 1, filed in the District
Court as No. 04-5184, Dkt. Entry # 506. Plaintiffs have never
disputed the District Court’s determination that “[b]ecause
Plaintiffs have alleged the Section 1 claim as a per se violation,
even at the pleading stage Plaintiffs must set forth sufficient
facts evidencing a horizontal conspiracy involving market or
customer allocation in order for their claim to survive a motion
to dismiss.” 2007 WL 1100449, at *10; see also Defendants’
Comm. Br. 10 (stating that on appeal, “[a]s in the district court,
Plaintiffs have abandoned any argument that [the complaints]
state[] a claim under the rule of reason”). Plaintiffs argue only
that they have, in fact, adequately pled such horizontal
conspiracies.

43
Twombly, 550 U.S. at 555 & n.3. The test, as authoritatively
formulated by Twombly, is whether the complaint alleges
“enough fact[] to state a claim to relief that is plausible on its
face,” id. at 570, which is to say, “‘enough fact to raise a
reasonable expectation that discovery will reveal evidence of
illegal[ity],’” Arista Records, LLC v. Doe 3, 604 F.3d 110, 120
(2d Cir. 2010) (quoting Twombly, 550 U.S. at 556) (alteration in
Arista Records).17

17
Twombly affirms that Rule 8(a)(2) requires a statement of
facts “suggestive enough” (when assumed to be true) “to render
[the plaintiff’s claim to relief] plausible,” that is, “enough fact
to raise a reasonable expectation that discovery will reveal
evidence of illegal” conduct. Twombly, 550 U.S. at 556. Iqbal,
which reiterated and applied Twombly’s pleading standard,
endorses this understanding. See Iqbal, 129 S. Ct. at 1949–51.
Although Fowler v. UPMC Shadyside, 578 F.3d 203 (3d Cir.
2009), stated that Twombly and Iqbal had “repudiated” the
Supreme Court’s earlier decision in Swierkiewicz v. Sorema
N.A., 534 U.S. 506 (2002), see Fowler, 578 F.3d at 211, we are
not so sure. Clearly, Twombly and Iqbal inform our
understanding of Swierkiewicz, but the Supreme Court cited
Swierkiewicz approvingly in Twombly, see 550 U.S. at 555–56,
and expressly denied the plaintiffs’ charge that Swierkiewicz
“runs counter” to Twombly’s plausibility standard, id. at 569–70.
As the Second Circuit has observed, Twombly “emphasized that
its holding was consistent with [the Court’s] ruling in
Swierkiewicz that ‘a heightened pleading requirement,’ requiring

44
 As we have recognized, this plausibility standard is an
interpretation of Federal Rule of Civil Procedure 8. Phillips,
515 F.3d at 234; see Twombly, 550 U.S. at 557 (stating that the
plausibility standard “reflects the threshold requirement of Rule
8(a)(2) that the ‘plain statement’ possess enough heft to ‘sho[w]
that the pleader is entitled to relief.’” (alteration in original)).
Twombly’s importance to the case before us, however, goes
beyond its formulation of the general pleading standard.
Twombly is also an essential guide to the application of that
standard in the antitrust context, for in Twombly the Supreme
Court also had to determine whether a Sherman Act claim
alleging horizontal conspiracy was adequately pled.18

the pleading of ‘specific facts beyond those necessary to state
[a] claim and the grounds showing entitlement to relief,’ was
‘impermissibl[e].’” Arista Records, 604 F.3d at 120 (quoting
Twombly, 550 U.S. at 570 (alterations in Arista Records). In
any event, Fowler’s reference to Swierkiewicz appears to be
dicta, as Fowler found the complaint before it to be adequate.
578 F.3d at 212; see also id. at 211 (“The demise of
Swierkiewicz, however, is not of significance here.”).
18
As the Supreme Court has noted, “[c]ontext matters in
notice pleading,” Phillips, 515 F.3d at 232, and what suffices to
withstand a motion to dismiss necessarily depends on
substantive law and the elements of the specific claim asserted.
See Iqbal, 129 S. Ct. at 1950 (“Determining whether a complaint
states a plausible claim for relief [so as to satisfy the Twombly

45
 The Twombly plaintiffs had alleged that defendant
telephone companies had “entered into a contract, combination
or conspiracy to prevent competitive entry in their respective
local telephone and/or high speed internet service markets and

standard] will . . . be a context-specific task . . . .”); see also id.
at 1947 (“In Twombly, the Court found it necessary first to
discuss the antitrust principles implicated by the complaint.
Here too we begin by taking note of the elements [the] plaintiff
must plead to state [his discrimination] claim . . . .” (internal
citation omitted)). The touchstone of Rule 8(a)(2) is whether a
complaint’s statement of facts is adequate to suggest an
entitlement to relief under the legal theory invoked and thereby
put the defendant on notice of the nature of the plaintiff’s claim.
See Twombly, 550 U.S. at 565 n.10 (noting that “a defendant
seeking to respond to plaintiffs’ conclusory allegations in the §
1 [of the Sherman Act] context would have little idea” how to
answer). Some claims will demand relatively more factual
detail to satisfy this standard, while others require less. See
Arista Records, 604 F.3d at 120 (stating that the Supreme
Court’s recent pleading decisions “require factual amplification
[where] needed to render a claim plausible” (internal quotation
marks omitted) (alteration in original)). As discussed below, the
question of the sufficiency of the complaint in Twombly turned
largely on the doctrinal fact that “antitrust law limits the range
of permissible inferences from ambiguous evidence in a § 1
case.” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475
U.S. 574, 588 (1986); see Twombly, 550 U.S. at 554–57.

46
ha[d] agreed not to compete with one another and otherwise
allocated customers and markets to one another.” Twombly, 550
U.S. at 551 (internal quotation marks omitted). The Court
found, however, that this sort of “wholly conclusory statement
of claim,” id. at 561, was insufficient to plead an entitlement to
relief. Id. at 564 & n.9; see id. at 556–57 (“Without more, . . .
a conclusory allegation of agreement at some unidentified point
does not supply facts adequate to show illegality.”). The Court
therefore proceeded to examine the entirety of the complaint’s
allegations, in order to determine whether the complaint
contained “enough factual matter (taken as true) to suggest that
an agreement was made,” in other words, “enough to render a
§ 1 conspiracy plausible.” Id. at 556.

In conducting this inquiry, the Court looked to well-
settled jurisprudence establishing what is necessary to satisfy the
conspiracy requirement of a § 1 claim at various post-pleading
stages of litigation. Id. at 554 (citing Theatre Enters., Inc. v.
Paramount Film Distrib. Corp., 346 U.S. 537 (1954) (affirming
denial of directed verdict); Monsanto Co. v. Spray-Rite Serv.
Corp., 465 U.S. 752 (1984) (same); Matsushita Elec. Indus. Co.
v. Zenith Radio Corp., 475 U.S. 574 (1986) (addressing whether
the record evidence of agreement was sufficient to withstand a
motion for summary judgment)). The crux of this case law is
that evidence of parallel conduct by alleged co-conspirators is
not sufficient to show an agreement. Indeed, “[e]ven ‘conscious
parallelism,’ a common reaction of ‘firms in a concentrated
market [that] recogniz[e] their shared economic interests and

47
their interdependence with respect to price and output decisions’
is ‘not in itself unlawful.’” Id. at 553–54 (quoting Brooke Group
Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227
(1993)) (alterations in Twombly).19 Parallel conduct is, of
course, consistent with the existence of an agreement; in many
cases where an agreement exists, parallel conduct—such as
setting prices at the same level—is precisely the concerted

19
In a highly concentrated market, “any single firm’s price
and output decisions will have a noticeable impact on the market
and on its rivals,” such that when any firm in that market “is
deciding on a course of action, any rational decision must take
into account the anticipated reaction of the other firms.” Flat
Glass, 385 F.3d at 359 (internal quotation marks omitted); see
6 Areeda & Hovenkamp, Antitrust Law ¶ 1429 (2d ed. 2003).
According to this “theory of interdependence . . . firms in a
concentrated market may maintain their prices at
supracompetitive levels, or even raise them to those levels,
without engaging in any overt concerted action.” Flat Glass,
385 F.3d at 359. Although this oligopolistic behavior, or
“conscious parallelism,” is often adverse to consumer interests,
courts have nonetheless found that it is not, without more,
sufficient evidence of a § 1 violation, both because it is not an
agreement within the meaning of the Sherman Act, and because
it is resistant to judicial remedies. Id. at 359–60. But see
Richard A. Posner, Antitrust Law 51–100 (2d ed. 2001) (arguing
that “conscious parallelism,” or “tacit collusion,” should
sometimes suffice to prove a § 1 violation).

48
action that is the conspiracy’s object. But as the Supreme Court
has long recognized, parallel conduct is “just as much in line
with a wide swath of rational and competitive business strategy
unilaterally prompted by common perceptions of the market.”
Id. at 554; see Matsushita, 475 U.S. at 594 (warning that
“mistaken inferences” of conspiracy from ambiguous
circumstantial evidence may “chill the very conduct the antitrust
laws are designed to protect”); see also supra note 10. In order
“to avoid deterring innocent conduct that reflects enhanced,
rather than restrained, competition,” Flat Glass, 385 F.3d at 357,
and in order to enforce the Sherman Act’s requirement of an
agreement, the Supreme Court has required that “a § 1 plaintiff’s
offer of conspiracy evidence must tend to rule out the possibility
that the defendants were acting independently,” Twombly, 550
U.S. at 554; see also Matsushita, 475 U.S. at 597 n.21
(“[C]onduct that is as consistent with permissible competition as
with illegal conspiracy does not, without more, support even an
inference of conspiracy.”).

Some courts have denominated these facts, the presence
of which may indicate the existence of an actionable agreement,
as “plus factors.” Flat Glass, 385 F.3d at 360. Although
“[t]here is no finite set of such criteria . . .[,] [w]e have
identified . . . at least three such plus factors: (1) evidence that
the defendant had a motive to enter into a price fixing
conspiracy; (2) evidence that the defendant acted contrary to its
interests; and (3) ‘evidence implying a traditional conspiracy.’”
Id. (quoting Petruzzi’s IGA Supermarkets, Inc. v. Darling-

49
Delaware Co., 998 F.2d 1224, 1244 (3d Cir. 1993)). As we
have cautioned, however, care must be taken with the first two
types of evidence, each of which may indicate simply that the
defendants operate in an oligopolistic market, that is, may
simply restate the (legally insufficient) fact that market behavior
is interdependent and characterized by conscious parallelism.
Id. at 360–61; see 6 Areeda & Hovenkamp, Antitrust Law ¶
1434c1 (2d ed. 2003); see also Baby Food, 166 F.3d at 135
(“[E]vidence of action that is against self-interest or motivated
by profit must go beyond mere interdependence.”).20 The third
factor, “evidence implying a traditional conspiracy,” consists of
“non-economic evidence ‘that there was an actual, manifest
agreement not to compete,’” which may include “‘proof that the
defendants got together and exchanged assurances of common

20
In fact, “in actual practice, most courts rely on the absence
of motivation or offense to self-interest to preclude a conspiracy
inference” from ambiguous evidence or mere parallelism. 6
Areeda & Hovenkamp, supra, ¶ 1434c2; see, e.g., Matsushita,
475 U.S. at 596–97 (“[I]f petitioners had no rational economic
motive to conspire, and if their conduct is consistent with other,
equally plausible explanations, the conduct does not give rise to
an inference of conspiracy.”); Southway Theatres v. Georgia
Theatre Co., 672 F.2d 485, 494 (5th Cir. Unit B 1982) (The
“basic rule” is “that the inference of a conspiracy is always
unreasonable when it is based solely on parallel behavior that
can be explained as the result of the independent business
judgment of the defendants.”).

50
action or otherwise adopted a common plan even though no
meetings, conversations, or exchanged documents are shown.’”
Flat Glass, 385 F.3d at 361 (quoting In re High Fructose Corn
Syrup Antitrust Litig., 295 F.3d 651, 661 (7th Cir. 2002); 6
Areeda & Hovenkamp, supra, ¶ 1434b); see 6 Areeda &
Hovenkamp, supra, ¶ 1416, at 103 (referring generally to “an
overt act more consistent with some pre-arrangement for
common action than with independently arrived-at decisions”).

One important question raised by Twombly is what is the
relationship between this summary judgment (and directed
judgment) jurisprudence governing the kind of evidentiary facts
necessary to support a finding of conspiracy, on the one hand,
and the “antecedent” issue, Twombly, 550 U.S. at 554, of a § 1
plaintiff’s pleading burden, on the other. We think Twombly
aligns the pleading standard with the summary judgment
standard in at least one important way: Plaintiffs relying on
circumstantial evidence of an agreement must make a showing
at both stages (with well-pled allegations and evidence of
record, respectively) of “something more than merely parallel
behavior,” id. at 560, something “plausibly suggest[ive of] (not
merely consistent with) agreement,” id. at 557. See id. at 557
n.5 (noting that a plaintiff’s pleadings must cross the line
“between the factually neutral and the factually suggestive”).
“Hence, when allegations of parallel conduct are set out in order
to make a § 1 claim, they must be placed in a context that raises
a suggestion of a preceding agreement, not merely parallel
conduct that could just as well be independent action.” Id. at

51
557. Put differently, allegations of conspiracy are deficient if
there are “obvious alternative explanation[s]” for the facts
alleged. Id. at 567.21

A corollary of this proposition is that plaintiffs relying on

21
Although Twombly’s articulation of the pleading standard
for § 1 cases draws from summary judgment jurisprudence, the
standards applicable to Rule 12(b)(6) and Rule 56 motions
remain distinct. In expounding this distinction, some judges and
commentators have opined that “[e]ven in those contexts in
which an allegation of [conspiracy based on] parallel conduct
will not suffice to take an antitrust plaintiff’s case to the jury, it
will sometimes suffice to overcome a motion to dismiss and
permit some discovery, perhaps leaving the issue for later
resolution on a motion for summary judgment.” Starr v. Sony
BMG Music Entm’t, 592 F.3d 314, 329 (2d Cir. 2010)
(Newman, J., concurring). One of Twombly’s formulations of
the plausibility pleading standard—calling for “enough fact to
raise a reasonable expectation that discovery will reveal
evidence of illegal agreement,” 550 U.S. at 556—appears to
support this view. See also supra note 17. In any case, a claim
of conspiracy might appear plausible in light of the well-pled
facts in the complaint, only to appear deficient at the summary
judgment stage, when (1) the plaintiff can no longer rely on
mere allegations but must adduce evidence, and (2) the
defendant’s uncontroverted evidence is also added to the
picture.

52
parallel conduct must allege facts that, if true, would establish
at least one “plus factor,” since plus factors are, by definition,
facts that “tend[] to ensure that courts punish concerted
action—an actual agreement—instead of the unilateral,
independent conduct of competitors.” Flat Glass, 385 F.3d at
360 (internal quotation marks omitted); accord Lum v. Bank of
Am., 361 F.3d 217, 230 (3d Cir. 2004) (describing plus factors
as “circumstances under which . . . the inference of rational
independent choice [is] less attractive than that of concerted
action” (quoting Bogosian v. Gulf Oil Corp., 561 F.2d 434, 446
(3d Cir. 1977)).22

22
Twombly did not explicitly use the term “plus factor” in
formulating its pleading standard. But the Court did note that
the lower-court decision under review had held that “plus
factors are not required to be pleaded to permit an antitrust
claim based on parallel conduct to survive dismissal.” Twombly,
550 U.S. at 553 (quoting 425 F.3d 99, 114 (2d Cir. 2005)
(emphasis in original)). The basis for the Court of Appeals’
conclusion was that parallel conduct alone was sufficient to
plead a § 1 conspiracy, as long as the court could conceive of
some set of facts “that would permit a plaintiff to demonstrate
that the particular parallelism asserted was the product of
collusion rather than coincidence.” 425 F.3d at 114, rev’d, 550
U.S. 544. In reversing, the Supreme Court expressly rejected
that premise and retired the “no set of facts” language from
Conley v. Gibson, 355 U.S. 41, 45–46 (1957), on which the
Court of Appeals had relied. Twombly, 550 U.S. at 561–63. By

53
 It bears noting that, consistent with summary judgment
analysis, plus factors need be pled only when a plaintiff’s claims
of conspiracy rest on parallel conduct. Allegations of direct
evidence of an agreement, if sufficiently detailed, are
independently adequate. See Twombly, 550 U.S. at 564
(distinguishing “independent allegation[s] of actual agreement”
from “descriptions of parallel conduct”).23 But this does not

repudiating this premise, the Supreme Court necessarily rejected
the proposition that plaintiffs may plead conspiracy on the basis
of mere parallelism—and thus necessarily required the pleading
of plus factors. As Twombly put it, “[a] statement of parallel
conduct . . . needs some . . . further circumstance,” or “further
factual enhancement,” to plead a plausible § 1 claim. 550 U.S.
at 557. Moreover, as discussed below, the crucial deficiency in
the Twombly complaint was that the plaintiffs could not
demonstrate what we have identified as an important plus factor,
see Petruzzi’s, 998 F.2d at 1244, namely that the defendants’
alleged parallel conduct was contrary to their self-interest.
Accordingly, although a plaintiff still need not plead specific
evidence, see supra note 17, Twombly abrogates our earlier
statements, see, e.g., Lum, 361 F.3d at 230, that a theory of
agreement resting on parallel conduct need not plead facts that,
if true, would constitute plus factors.
23
Courts devised the requirement of “plus factors” in the
context of offers of proof of an agreement that rest on parallel
conduct, i.e., circumstantial evidence. On appeals from

54
summary judgment, we have stated that direct evidence of a
conspiracy, such as a document or conversation explicitly
m anifestin g th e existe nce of th e agreem ent in
question—“evidence that is explicit and requires no inferences
to establish the proposition or conclusion being asserted,” Baby
Food, 166 F.3d at 118—obviates the need for such a showing.
Rossi, 156 F.3d at 466 (citing Petruzzi’s, 998 F.2d at 1233); see
also Cosmetic Gallery, Inc. v. Schoeneman Corp., 495 F.3d 46,
52 (3d Cir. 2007) (providing examples of direct evidence of
conspiracy). “This is because when the plaintiff has put forth
direct evidence of conspiracy, the fact finder is not required to
make inferences to establish facts, and therefore the Supreme
Court’s concerns over the reasonableness of inferences in
antitrust cases evaporate.” Rossi, 156 F.3d at 466 (citing
Petruzzi’s, 998 F.2d at 1233); see, e.g., Mack Trucks, 530 F.3d
at 222 (noting that a statement by a vice president of the
defendant was “direct evidence of collusion, which, if believed,
requires no further inference”); see also Golden Bridge Tech.,
Inc. v. Motorola, Inc., 547 F.3d 266, 272 (5th Cir. 2008)
(implying the same distinction between the treatment of direct
and circumstantial evidence), cert. denied, 129 S. Ct. 2055
(2009); Williamson Oil Co. v. Philip Morris USA, 346 F.3d
1287, 1300 (11th Cir. 2003) (same). Put differently, direct
evidence of conspiracy, if credited, removes any ambiguities
that might otherwise exist with respect to whether the parallel
conduct in question is the result of independent or concerted

55
mean that a § 1 claim will be considered adequately pled
because of the bare possibility that discovery might unearth
direct evidence of an agreement. The Court of Appeals’ opinion
in Twombly had pointed to that possibility as a ground for
denying dismissal. 425 F.3d at 114. But the Supreme Court
expressly rejected this reasoning, stating that “this approach to
pleading would dispense with any showing of a ‘reasonably
founded hope’ that a plaintiff would be able to make a case.”
Twombly, 550 U.S. at 562 (quoting Dura Pharm., Inc. v.
Broudo, 544 U.S. 336, 347 (2005)). After Twombly, if a
plaintiff expects to rely exclusively on direct evidence of
conspiracy, its complaint must plead “enough fact to raise a
reasonable expectation that discovery will reveal” this direct
evidence. Id. at 556. And if the plaintiff alternatively expects
to rest on the circumstantial evidence of parallel behavior, the
complaint’s statement of facts must place the alleged behavior
in “a context that raises a suggestion of a preceding agreement,
not merely parallel conduct that could just as well be

action.
Twombly noted that no such direct allegations appeared
in the complaint before it. See 550 U.S. at 565 n.11 (observing
that plaintiffs do not “directly allege illegal agreement” but
rather “proceed exclusively via allegations of parallel conduct”);
see also id. at 565 n.10 (“Apart from identifying a seven-year
span in which the § 1 violations were supposed to have
occurred . . ., the pleadings mentioned no specific time, place,
or person involved in the alleged conspiracies.”).

56
independent action.” Id. at 557.24 In other words, regardless of
whether the plaintiff expects to prove the existence of a
conspiracy directly or circumstantially, it must plead “enough
fact to raise a reasonable expectation that discovery will reveal
evidence of illegal agreement.” Id. at 556.25

24
Sometimes, of course, discovery will uncover both direct
and circumstantial evidence of agreement. We do not imply that
a plaintiff must commit to a single method of proof at the
pleading stage, but merely that a plaintiff must put forth some
statement of facts suggestive of unlawful conspiracy. “[O]nce
a claim has been stated adequately, it may be supported by
showing any set of [evidentiary] facts consistent with the
allegations in the complaint.” Twombly, 550 U.S. at 563.
25
Twombly thus abrogates our earlier holdings that § 1
plaintiffs can survive a motion to dismiss without alleging facts
supporting a plausible inference of conspiracy. See, e.g.,
Bogosian, 561 F.2d at 446. Bogosian correctly observed that
“[i]t is not necessary to plead evidence.” Id. at 446; accord id.
at 458 (Aldisert, J., dissenting); see also supra note 17. But we
think the opinion is at odds with Twombly insofar as it absolves
plaintiffs of the obligation “to plead the facts upon which the[ir]
claim is based.” Id. at 446 (majority opinion). Bogosian’s
formulation of the pleading standard appears to have derived
from the view that a complaint is sufficient so long as “it does
not appear to a certainty that plaintiffs can prove no set of facts
which . . . would entitle them to reach the jury,” id., that is, it

57
 Because Twombly dismissed the antitrust claim before it,
the Court did not provide specific examples of allegations that
would satisfy its plausibility standard. Nonetheless, the Court
did point in general terms to “parallel behavior that would
probably not result from chance, coincidence, independent
responses to common stimuli, or mere interdependence unaided

appears to reflect precisely the pervasive misapprehension of
Federal Rule of Procedure 8(a)(2) that led the Twombly Court to
“retire” the oft cited language from Conley v. Gibson. See
Twombly, 550 U.S. at 560–63. Based on this pre-Twombly
understanding of “the precept that the complaint be liberally
construed,” Bogosian found it sufficient that the complaint
provided a statement of alleged consciously parallel conduct by
the defendants, along with the unelaborated assertion that the
defendants had entered into a “combination.” Bogosian, 561
F.2d at 445–46. The opinion did not examine whether the
allegation of concerted action was plausible in light of the
context in which the parallel conduct was situated, instead
deferring until after discovery the question of whether such
conduct might in fact be perfectly consistent with each
defendant’s independent self-interest. Id. at 446. Twombly, we
think, clearly demands more scrutiny of a § 1 complaint. As the
dissent in Bogosian maintained, “an allegation of consciously
parallel behavior, without more, [does] not state a Sherman Act
claim,” id. at 459 (Aldisert, J., dissenting), and a plaintiff cannot
merely assert that the defendants’ actions were concerted
without alleging facts plausibly suggesting an agreement.

58
by an advance understanding among the parties.” 550 U.S. at
556 n.4 (citing 6 Areeda & Hovenkamp, supra, ¶ 1425, at
167–85). More significantly, the shortcomings identified in the
T w om bly com plain t p ro vide an im portant— albeit
negative—gloss on the governing standard.

The Twombly plaintiffs proffered two basic theories of
anticompetitive collusion. First, they charged that the defendant
regional telephone companies (ILECs) conspired to “inhibit the
growth of upstart” competitors (CLECs). 550 U.S. at 550.
Second, they asserted that the ILECs agreed not to compete with
one another so as to preserve the preexisting regional monopoly
each enjoyed. Id. at 551.

At the outset of its analysis, the Court remarked that the
complaint’s sufficiency would “turn[] on the suggestions raised
by [defendants’ alleged] conduct when viewed in light of
common economic experience.” Id. at 565. Under this lens, the
complaint’s first theory immediately revealed its inadequacy
because “nothing in the complaint intimate[d] that the resistance
to the upstart[ CLECs] was anything more than the natural,
unilateral reaction of each ILEC intent on keeping its regional
dominance. . . . [T]here [was] no reason to infer that the
companies had agreed among themselves to do what was only
natural anyway . . . .” Id. at 566. A rudimentary economic
analysis also fatally undermined the complaint’s second charge,
namely that the ILECs agreed not to enter one another’s
markets. The Court recognized that “[i]n a traditionally
unregulated industry with low barriers to entry, sparse

59
competition among large firms dominating separate
geographical segments of the market could very well signify
illegal agreement.” Id. at 567. But in the telecommunications
industry at issue in Twombly, monopoly had been “the norm . . .,
not the exception.” Id. at 568. Noting that “[t]he ILECs were
born in that world, doubtless liked the world the way it was, and
surely knew the adage about him who lives by the sword,” the
Court found that “a natural explanation for the noncompetition
alleged is that the former Government-sanctioned monopolists
were sitting tight, expecting their neighbors to do the same
thing.” Id. In fact, “the complaint itself” bolstered this
conclusion. Id. Not only did it “not allege that competition
[against other ILECs] as CLECs was potentially any more
lucrative than other opportunities being pursued by the ILECs
during the same period,” but “the complaint [was] replete with
indications that any CLEC faced nearly insurmountable barriers
to profitability owing to the ILECs’ flagrant resistance to the
network sharing requirements” of federal law. Id. In short, both
“common economic experience” and the complaint’s own
allegations showed that each defendant ILEC was independently
motivated to behave in the ways alleged. Accordingly, neither
of plaintiffs’ theories successfully pled a § 1 conspiracy because
in each case, defendants’ parallel conduct “was not only
compatible with, but indeed was more likely explained by,
lawful, unchoreographed free-market behavior.” Ashcroft v.
Iqbal, 129 S. Ct. 1937, 1950 (2009) (summarizing Twombly).

In sum, Twombly makes clear that a claim of conspiracy

60
predicated on parallel conduct should be dismissed if “common
economic experience,” or the facts alleged in the complaint
itself, show that independent self-interest is an “obvious
alternative explanation” for defendants’ common behavior. For
our present purposes, we find this guidance sufficient.

b. Assessing the Sufficiency of Plaintiffs’ Pleadings

As the Supreme Court has instructed, we begin by
identifying the complaints’ bare assertions that the insurers or
brokers entered into horizontal agreements. See, e.g., Comm.
SAC ¶ 158 (“[T]he Insurers members of the . . . Broker-
Centered Conspiracy all agreed with [the Broker], and agreed
horizontally among themselves, to reduce or eliminate
competition for [the Broker’s] secured business among the
conspiring insurers.”); id. ¶ 354 (“[T]he Broker ‘hubs’
simultaneously agreed horizontally not to compete with each
other . . . .”). Because these conclusory averments do not
“show[]” but merely “assert[]” plaintiffs’ entitlement to relief,
Twombly, 550 U.S. at 555 n.3, they cannot carry plaintiffs’
pleading burden. See id. at 556–57 (“Without more, . . . a
conclusory allegation of agreement at some unidentified point
does not supply facts adequate to show illegality.”); cf. Howard
Hess Dental Labs. Inc. v. Dentsply Int’l, Inc., 602 F.3d 237,
254–55 (3d Cir. 2010) (holding that it was inadequate for the
complaint to state in “a conclusory manner” that “Defendants,
each with all of the others, have entered into two interrelated
conspiracies” and that “every Dealer knew that every other
Dealer agreed, or would agree, to th[e] same [allegedly

61
unlawful] plan” (emphasis omitted)). Accordingly, we must
examine the entirety of the complaints’ factual allegations and
determine whether, taken as true, they support a plausible
inference of horizontal conspiracy.

i. The Broker-Centered Conspiracies

(a) Conspiracies Not Involving Bid Rigging

As the District Court recognized, plaintiffs’ “broker-
centered conspiracies” are alleged as hub-and-spoke
conspiracies, with the broker as the hub and its insurer-partners
as the spokes. This type of conspiracy has “a long history in
antitrust jurisprudence.” Dentsply Int’l, 602 F.3d at 255 (citing
Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939))
(discussing general hub-and-spoke model). “[T]he critical issue
for establishing a per se violation with the hub and spoke system
is how the spokes are connected to each other.” Total Benefits
Planning Agency, Inc. v. Anthem Blue Cross & Blue Shield, 552
F.3d 430, 436 (6th Cir. 2008). Here, the District Court found
plaintiffs had not adequately alleged the existence of a “wheel”
or “rim” (that is, a horizontal agreement) connecting the insurer-
spokes. 2007 WL 2533989, at *17; see Dickson v. Microsoft
Corp., 309 F.3d 193, 203 (4th Cir. 2002) (“A rimless wheel
conspiracy is one in which various defendants enter into
separate agreements with a common defendant, but where the
defendants have no connection with one another, other than the
common defendant’s involvement in each transaction.”) (citing
Kotteakos v. United States, 328 U.S. 750, 755 (1946)); cf.

62
Dentsply Int’l, 602 F.3d at 255 (concluding that “even assuming
the Plaintiffs have adequately identified the hub (Dentsply) as
well as the spokes (the Dealers), . . . the amended complaint”
fails to allege adequately “an agreement among the Dealers
themselves”).

Plaintiffs’ allegations in support of horizontal conspiracy
in the broker-centered schemes fall into two different categories.
First, plaintiffs assert that the very nature of the contingent
commission agreements between the broker and each of its
insurer-partners implies an agreement among the brokers.
Second, plaintiffs rely on specific details about the operation of
the customer steering schemes, particularly the “devices” used
to ensure that a particular piece of business was placed with the
designated insurer. With the exception of the bid rigging
alleged in the Marsh-centered commercial conspiracy, we agree
with the District Court that plaintiffs’ allegations do not give
rise to a plausible inference of horizontal conspiracy.

Contrary to plaintiffs’ arguments, one cannot plausibly
infer a horizontal agreement among a broker’s insurer-partners
from the mere fact that each insurer entered into a similar
contingent commission agreement with the broker. As the
District Court concluded, the first stage of the alleged broker-
centered conspiracies—the consolidation of the groups of
insurers to which each broker referred business—evinces
nothing more than a series of vertical relationships between the
broker and each of its “strategic partners.” 2007 WL 2533989,
at *15.

63
 According to the complaints, the defendant brokers
decided to consolidate the pool of insurers to which they
referred business in order to improve efficiency and extract
higher commissions from each of their insurer-partners. As
defendants point out, “[o]nce a broker decided to organize its
business in this fashion, each insurer had sound, independent
business reasons to pay contingent commissions to become and
remain a ‘preferred insurer.’ Paying such commissions helped
the insurer to compete for and retain a larger share of its
partners’ business than if it had no such vertical relationships.”
Defendants’ EB Br. 38. In short, the obvious explanation for
each insurer’s decision to enter into a contingent commission
agreement with a broker that was consolidating its pool of
insurers was that each insurer independently calculated that it
would be more profitable to be within the pool than without.
The complaints themselves reinforce this conclusion with their
portrait of a concentrated brokerage market, in which a handful
of brokers controlled the majority of client business, and an
unconcentrated, more competitive market of insurers vying for
premium dollars. Comm. SAC ¶¶ 70–76; EB SAC ¶¶ 67–73.
According to plaintiffs’ own account, “[t]he Insurer Defendants
are thus largely dependent on the Broker Defendants to assure
access to business and protect their market share.” EB SAC
¶ 73; accord Comm. SAC ¶ 76; see also id. ¶ 73 (“The close
bond between broker and client gives brokers tremendous
influence, and often decisive control, over the placement of their
clients’ insurance business.”). Given this economic landscape,
each insurer had an obvious incentive to enter into the “strategic

64
partnerships” offered by the defendant brokers, irrespective of
the actions of its competitors.

Refusing to concede this point, plaintiffs argue that the
parallel decisions of insurers to join the broker-centered
conspiracies plausibly imply a horizontal agreement among the
insurers because “an insurer would not pay enormous contingent
commissions in order to access premium volume if its major
rivals were getting the same access for free.” Plaintiffs’ EB
Reply Br. 8 (emphasis omitted). This contention is implausible.
Although each insurer would be motivated to achieve the best
deal possible with the broker—and would doubtless like to
obtain terms as least as favorable as those negotiated by other
insurers—the determinative consideration would be whether the
insurer is better off paying contingent commissions for
privileged access to the broker’s clients than it would be saving
those payments and foregoing the broker’s assistance in winning
and retaining business. Especially in light of the market
dynamics alleged by plaintiffs, the obvious explanation for the
decision of the defendant insurers to enter into contingent
commissions agreements with the consolidating brokers is that
each insurer found that the benefits justified the costs. In fact,
the complaints relate incidents in which insurers who were
reluctant to conform to the contingent commission demands of
a broker nonetheless did so when faced with the prospect of
losing their privileged access to the broker’s book of business.
See, e.g., Comm. SAC ¶¶ 135–140; EB SAC ¶¶ 163–168; id. ¶¶
214–226. These anecdotes only strengthen the obvious

65
conclusion that no horizontal agreement was necessary to induce
the insurers to become “strategic partners.”

Moreover, plaintiffs’ argument proves too much. If the
parallel decisions by several insurers to pay contingent
commissions imply a horizontal agreement, then it is difficult to
see why parallel decisions to pay standard commissions (that is,
a fixed percentage of each policyholder’s premium payment)
would not also imply an agreement. For that matter, plaintiffs’
logic would divine a horizontal agreement from virtually any
parallel expenditures for marketing services, on the mistaken
ground that a firm would not pay for advertising, for example,
in the absence of an agreement with its competitors to enter into
similar contracts with the advertising company. Cf. Twombly,
550 U.S. at 566 (noting that “resisting competition is routine
market conduct,” and that “if alleging parallel decisions to resist
competition were enough to imply an antitrust conspiracy,
pleading a § 1 violation against almost any group of competing
businesses would be a sure thing”).26 The District Court

26
Plaintiffs distinguish contingent commissions from
advertising costs on two grounds, neither of which is relevant.
First, plaintiffs stress that unlike advertising, which is
procompetitive, the customer allocation schemes allegedly
linked to the contingent commission payments were antagonistic
to competition. This response, however, misunderstands the
thrust of the advertising analogy. Even assuming defendants’
practices unreasonably restrained trade, plaintiffs’ § 1 claims

66
correctly found that the brokers’ alleged consolidation of the
insurers with which they did business did not plausibly imply an
agreement susceptible to per se condemnation.

Plaintiffs seek to bolster the inference of horizontal
agreement with allegations of information-sharing among the
members of each putative broker-centered conspiracy. Plaintiffs
assert numerous instances, for example, in which a broker
communicated the details of its contingent commission

must also plausibly suggest that these practices were the product
of an agreement among the insurers. The advertising analogy
illustrates plaintiffs’ failure to satisfy this element of their
pleading burden; parallel conduct, such as the payment of
contingent commissions, does not plausibly imply the existence
of an agreement when each defendant had a strong, independent
motive to engage in that conduct.
Second, plaintiffs allege that contingent commission
agreements were not customary before the brokers’ decisions in
the 1990’s to consolidate their pool of insurers, and that the
insurers received no additional benefits in exchange for these
payments. Even if that is true, however—and the complaints’
assertions of increasing premium revenue by defendant insurers
during the proposed class period suggest otherwise—the point
is that, once the brokers had undertaken that consolidation,
insurers had much to lose if they did not become a “strategic
partner,” which provided each of them with an independent
business reason to pay brokers contingent commissions.

67
agreement with one insurer-partner to other insurer-partners, in
violation of confidentiality provisions forbidding such
disclosures. In plaintiffs’ view, these alleged disclosures helped
defendants to police the broker-centered conspiracies by
assuring each conspiring insurer that none of the other insurer-
partners was “cheating” by taking more than the allegedly
agreed-upon share of premium volume.

But there is a significant obstacle to plaintiffs’ attempts
to infer a horizontal agreement from this sharing of information.
The complaints allege only that the brokers made the
disclosures; there are no allegations that any insurer ever
horizontally disclosed to its competitors the details of its vertical
agreement with a broker. Furthermore, there are obvious
reasons for each broker to share this information with its
insurer-partners, reasons that have nothing to do with
preexisting agreements of any kind. The details of commission
agreements with other insurers, for example, could be a
powerful tool for a broker attempting to negotiate a more
favorable agreement with a particular insurer-partner. Either
match the “market” price for my premium volume, a broker
might threaten, or I will transfer your share of my business to
other, higher-commission-paying insurer-partners. This tactic
would seem to be an effective way for brokers to exploit the
leverage that, according to the complaints, they enjoyed over the
insurers. And in fact, the complaints show that brokers used the
information in precisely this way. See, e.g., EB SAC ¶ 126
(recounting an incident in which a broker “reveals [to a

68
particular insurer] that the bonus compensation arrangement it
was seeking from [that particular insurer] had been agreed to by
the other conspiring Insurers, and that [the particular insurer]
should offer terms like those put forth by another Insurers
[sic]”). Just as a manufacturer’s practice of informing each of
its distributors of the identities of its other distributors—as well
as the prices they paid and the volume of product they
received—would not plausibly imply a horizontal agreement
among the distributors, the disclosure of information alleged
here fails to plausibly suggest a conspiracy among the insurers.
It is true that if a horizontal conspiracy of the sort asserted by
plaintiffs existed, the exchange of information alleged could
conceivably serve the “policing” function plaintiffs describe.
But it does not follow that this disclosure of information
plausibly implies such a conspiracy; it is at least equally
consistent with unconcerted action.27

27
Plaintiffs contend that “[i]t strains credulity to insist that an
insurer, which repeatedly and systematically receives
confidential information about a rival’s contingent commission
arrangements and premium volume, would not expect and
understand that its rivals were being provided with the same
information about its business.” Plaintiff’s EB Reply Br. 13.
But the allegation that insurers knew that the brokers would
disclose the details of their vertical agreements to other insurer-
partners does not imply that insurers intended that the
information be so disclosed, let alone that they had entered into
a horizontal agreement with other insurers. Plaintiffs’ reliance

69
 The manufacturer analogy highlights a basic fallacy that
undergirds much of plaintiffs’ argumentative strategy. Plaintiffs
repeatedly insist that

when [an] insurer knows that it is buying
competitive protections for its incumbent business
and it knows that other insurers are not getting a
real opportunity on its incumbent business, and it
knows that there are other partners of the broker
who have the same competitive protections
bought with the same contingent commissions, it
is a fair inference . . . that this describes . . . a
horizontal conspiracy.

Tr. of Oral Arg. 15–16. “Competitive protections” sound
vaguely sinister, but what insurers were allegedly buying was a
portion of the client business controlled by the broker.

on United States v. Container Corp. of America, 393 U.S. 333
(1969), is thus inapposite. In Container Corp., the Supreme
Court found that the exchange among competitors of
information about the prices they charged to customers
constituted a horizontal conspiracy to limit price competition in
violation of the Sherman Act. The disclosure of information
alleged here, by contrast, is vertical and, unlike the exchange in
Container Corp., does not give rise to an inference of harm to
competition. See id. at 337; see also id. at 338 (“Price is too
critical, too sensitive a control to allow it to be used even in an
informal manner to restrain competition.”).

70
Whatever portion of that business one insurer buys is, of course,
a portion unavailable to other insurers. Each contract between
an insurer and the broker is, in this sense, a restraint of trade, but
only in the way that every contract is a restraint of trade. See
Bd. of Trade v. United States, 246 U.S. 231, 238 (1918) (“Every
agreement concerning trade, every regulation of trade, restrains.
To bind, to restrain, is of their very essence.”); cf. Am. Needle,
Inc. v. NFL, 130 S. Ct. 2201, 2208 (2010) (“[E]ven though, read
literally, § 1 would address the entire body of private contract,
that is not what the statute means.” (internal quotation marks
omitted)). A similar restraint occurs when a manufacturer signs
a contract with a distributor, agreeing to sell the distributor a
certain percentage of the manufacturer’s product. This
arrangement alone does not signify an agreement to
unreasonably restrain trade, let alone a horizontal agreement to
unreasonably restrain trade. Nor would an inference of
horizontal conspiracy arise from the fact that each distributor
knows which of its competitors have purchased the remaining
portions of the manufacturer’s product, as well as the specific
terms of the other deals. Here, plaintiffs claim the brokers’
ability to “guarantee” insurers certain amounts of premium
volume depended on deceiving their clients into believing that
the brokers had solicited competitive bids from the insurers, and
that in a given transaction, the insurer recommended by the
broker was the one who had made the most attractive offer.
These allegations of fraud, however, involve only the manner in
which the brokers obtained the “product” they sold to insurers;

71
they do not make the sales themselves an antitrust violation.28
Contrary to plaintiffs’ contentions, the allegations that each
insurer knew about the “competitive protections” purchased by
the other insurer-partners manifestly do not “describe[] . . . a
horizontal conspiracy” to unreasonably restrain trade.

Plaintiffs maintain that this conclusion is at odds with the
holdings in two hub-and-spoke-conspiracy cases, Interstate
Circuit, Inc. v. United States, 306 U.S. 208 (1939), and Toys
“R” Us, Inc. v. FTC, 221 F.3d 928 (7th Cir. 2000). In Interstate
Circuit, a theater chain company, Interstate, wrote to each of
eight movie distributors, asking them to meet certain conditions
in exchange for the theater company’s “continued exhibition of
the distributors’ films in its . . . first-run theatres” at a prescribed
price of admission. 306 U.S. at 216–17. The conditions
operated to restrict the terms under which the distributors could
license their films to subsequent-run theatres, Interstate’s
competitors. Although there was no evidence of any direct
communications among the eight distributors, the letter sent to
each distributor listed all eight distributors as addressees; in
other words, “from the beginning each of the distributors knew
that the proposals were under consideration by the others.” Id.

28
See infra note 31 and accompanying text. We discuss
below plaintiffs’ argument that the specific means allegedly
used to steer clients, e.g., first looks, last looks, and the
solicitation of intentionally uncompetitive bids, imply a
horizontal agreement among the insurers.

72
at 222. Each distributor accepted Interstate’s proposed terms.
The district court found this evidence proved concerted action
by the distributors in violation of § 1 of the Sherman Act, and
the Supreme Court affirmed. Plaintiffs cite Interstate Circuit for
the proposition that an actionable horizontal conspiracy does not
require direct communication among the competitors.

We do not dispute this principle, but it does not relieve
plaintiffs of the obligation to “allege facts plausibly suggesting
‘a unity of purpose or a common design and understanding, or
a meeting of minds in an unlawful arrangement.’” Dentsply
Int’l, 602 F.3d at 254 (quoting Copperweld, 467 U.S. at 771).
Key to Interstate Circuit’s conspiracy finding was its
determination that each distributor’s decision to accede to
Interstate’s demands would have been economically self-
defeating unless the other distributors did the same: “Each was
aware that . . . without substantially unanimous action . . . there
was risk of a substantial loss of the business and good will . . . .”
Interstate Circuit, 306 U.S. at 222. In the absence of common
action, agreeing to Interstate’s demands would have meant
reducing output (specifically, surrendering the distributor’s
share of the subsequent-run theater business) with no reasonable
prospect of countervailing benefits; only collective conduct by
the distributors could exert the market power necessary to
increase profits in the first-run arena. The Court stated that it
would “tax[] credulity to believe that the several distributors
would, in the circumstances, have accepted and put into
operation with substantial unanimity such far-reaching changes

73
in their business methods without some understanding that all
were to join, and we reject as beyond the range of probability
that it was the result of mere chance.” Id. at 223.

As noted, however, in the circumstances alleged here, the
rationality of each insurer’s decision to enter into a “strategic
partnership” with the broker does not presuppose concerted
action. The advantages of the partnership to the insurers flowed
from the broker’s control of its clients’ business, not the market
power of the insurers. If anything, an insurer here would prefer
that fewer of its competitors participate in the scheme, as it
would then enjoy that much more of the broker’s steered
business. See, e.g., Comm. SAC ¶ 242 (noting that one of
broker HRH’s insurer-partners preferred that HRH have only
three other partners, whereas HRH wanted four). The
opportunity to become a broker’s “strategic partner” was an
opportunity for the insurer to increase output, not reduce it.

Toys “R” Us is likewise distinguishable. There, Toys
“R” Us, a toy retailer, invited manufacturers to stop selling toys
to wholesale toy clubs, which competed with Toys “R” Us. The
manufacturers did so. The Court of Appeals for the Seventh
Circuit affirmed the FTC’s finding of § 1 conspiracy among the
manufacturers. The court acknowledged that the “agreements
between [Toys “R” Us] and the various manufacturers were, of
course, vertical agreements,” 221 F.3d at 932, which could not
in themselves constitute a per se violation. But the court
determined that the FTC’s finding of a horizontal agreement
among the manufacturers was warranted under the

74
circumstances. The evidence showed that the manufacturers
were “reluctan[t] to give up a new, fast-growing, and profitable
channel of distribution,” id. (quoting FTC opinion); they “were
in effect being asked by [Toys “R” Us] to reduce their output
. . ., and as is classically true in such cartels, they were willing
to do so only if [Toys “R” Us] could protect them against
cheaters,” id. at 936. In fact, the FTC had direct evidence, in the
form of statements by the manufacturers’ executives, that each
manufacturer agreed to Toys “R” Us’s proposal on the explicit
condition that its competitors do the same. Id. As the Seventh
Circuit noted, Toys “R” Us “is a modern equivalent of the old
Interstate Circuit decision.” Id. at 935. In both cases, the
evidence clearly indicated that the defendants would not have
undertaken their common action without reasonable assurances
that all would act in concert.

Here, the parallel vertical agreements are of a different
sort. Interstate and Toys “R” Us solicited exclusive-dealing
agreements from movie distributors and toy manufacturers,
respectively, in an attempt to exploit the latters’ collective
market power. Plaintiffs here do not allege that the insurers
possessed market power (as noted, plaintiffs instead emphasize
the brokers’ market power, see Comm. SAC ¶ 76; EB SAC ¶
73), nor that each broker wanted its insurer-partners to deal
exclusively with it (the complaints show that some insurers had
contingent commission agreements with multiple brokers 29 ).

29
See supra note 6.

75
Instead, plaintiffs’ allege that brokers demanded contingent
commissions in exchange for given amounts of broker-
controlled business. And the complaints show that each insurer
had an incentive to pay these commissions based solely on the
brokers’ ability to guarantee delivery of premium volume. Each
insurer’s share of the market thus depended on its ability to gain
the broker’s favor, not on the choices of its competitors.

Plaintiffs’ attempt to compare their allegations with the
facts of Interstate Circuit and Toys “R” Us is thus misguided.
If anything, the fundamentally different factual contexts in those
cases reinforce our view that the alleged information-sharing by
the brokers here does not plausibly support a claim of horizontal
conspiracy. 3 0 We believe the alleged contingent

30
Plaintiffs place special emphasis on the alleged
information-sharing in the HRH- and Wells Fargo/Acordia-
centered commercial conspiracies, but these allegations do not
overcome the basic deficiency we have just described. Plaintiffs
allege that HRH allocated its book of business among three
insurers and assert that “[t]he number of [insurers] to which
HRH allocated its business was discussed among and agreed to
by the three chosen insurers.” Comm. SAC ¶ 242. When we
search for additional information about this putative agreement,
we find mostly allegations common to the other broker-centered
conspiracies, namely that each insurer-partner knew the
identities of the others and the details of their similar contingent
commission agreements with the broker. Plaintiffs’ pleadings

76
commission agreements between brokers and insurers—which
form the backbone of plaintiffs’ alleged “broker-centered
conspiracies”—find a more apt analogue in the facts of NYNEX
Corp. v. Discon, Inc., 525 U.S. 128 (1998). In NYNEX, plaintiff

suggest that one insurer wanted HRH to have one fewer insurer-
partner than HRH originally had in mind. See Comm. RPS ¶
281 (“During its negotiations with HRH, [the insurer] was aware
of the existence of other proposed carrier partners and expressed
concern that HRH was considering consolidating its business
with four Insurers rather than only three, which [the insurer]
preferred.”). But this vertical effort to persuade HRH (with
apparent success) to exclude the participation of a competitor
hardly implies horizontal conspiracy among the insurers. (It
also stands in stark contrast to the hub-and-spoke conspiracies
found in Interstate Circuit and Toys “R” Us, in which each
firm’s motivation to enter into the vertical agreement was
contingent on all of its competitors’ doing the same.) To the
contrary, the obvious alternative explanation for the insurer’s
behavior is a desire to maximize its piece of HRH’s guaranteed-
premium-volume pie.
Similarly, the allegations in the Well Fargo/Acordia
conspiracy indicate only that the insurer-partners knew one
another’s identities, and knew that each was benefitting in
similar ways from the broker’s ability to steer business. They do
not imply that any insurer-partner’s agreement with Wells
Fargo/Acordia was dependent on the conduct of its competitors.

77
Discon alleged that Material Enterprises, a NYNEX subsidiary,
had switched its purchase of certain services from Discon to
AT&T Technologies, one of Discon’s competitors, despite the
fact that Discon was the less expensive servicer. According to
Discon, the transaction was part of a fraudulent scheme in which
Material Enterprises “could pass the higher prices on to New
York Telephone, which in turn could pass those prices on to
telephone consumers in the form of higher regulatory-agency-
approved telephone service charges. At the end of the year,
Material Enterprises would receive a special rebate from AT&T
Technologies, which Material Enterprises would share with its
parent, NYNEX.” Id. at 132. The scheme allegedly allowed
New York Telephone, a lawful monopoly, to circumvent
regulatory restrictions on the telephone services charges it could
impose on consumers, to the profit of the participating entities.
Discon alleged that Material Enterprises refused to choose it as
the service provider, despite its lower price, because it refused
to go along with the scheme. The Supreme Court “concede[d]
Discon’s claim that the petitioners’ behavior hurt consumers by
raising telephone service rates” but refused to apply a rule of per
se condemnation to this vertical restraint, noting further that the
consumer injury “naturally flowed” not so much from a less
competitive market for removal services as from New York
Telephone’s exercise of its lawfully held market power
“combined with a deception worked upon the regulatory
agency.” Id. at 136. Rather than § 1 of the Sherman Act, the
Court suggested, a more appropriate remedy might be found in
“other laws, for example ‘unfair competition’ laws, business tort

78
laws, or regulatory laws, [which] provide remedies for various
competitive practices thought to be offensive to proper standards
of business morality.” Id. at 137 (internal quotation marks
omitted).

Here, too, the “strategic partnerships” alleged by
plaintiffs imply only a vertical restraint. Furthermore, the
complaints show that the injury to purchasers of insurance
“naturally flowed” primarily from the nature of the broker-client
relationship and the ability it afforded brokers to deceive clients
about the quality and competitive status of the bids received
from insurers. Contingent commission agreements were the
means by which the brokers converted this power into profit,
ultimately at their clients’ expense; contingent commissions
were the “rebate” insurers paid to brokers. But none of the
allegations examined to this point give reason to believe that the
broker-centered schemes were underwritten by horizontal
agreements among the insurer-partners. Purchasers may have
some cause of action against the defendants for their alleged
deception and unfair trade practices, see id. (listing possible
legal remedies), but plaintiffs’ allegations of parallel contingent-
commissions-for-guaranteed-premium-volume agreements
between each broker and its insurer-partners do not adequately
plead a per se violation of § 1 of the Sherman Act.31

31
Hovenkamp’s discussion of NYNEX is also relevant to this
case:
[T]he allegations in [NYNEX] contained an

79
 The gravamen of plaintiffs’ allegations lies in what the
District Court described as the second stage of the asserted
schemes: the operation of the “incumbent protection rackets”
within each broker-centered conspiracy. Even if the parallel

element of fraud, but many thousands of contracts
have exchanged exclusivity for kickbacks or some
deception on consumers or third parties. An
agreement giving a waste removal or towing
company an exclusive right to the buyer’s
business in exchange for a secret rebate or
kickback does not injure competition simply
because of the fraud. Such a holding would cross
the line from antitrust to consumer protection.
And while protecting consumers from such
schemes is certainly a worthy goal of legal policy
generally, it is not an antitrust goal.
11 Hovenkamp, supra, ¶ 1902d, at 223. Here, too, the basic
scheme alleged by plaintiffs is one in which defendant brokers
exchanged exclusivity (premium volume) for kickbacks
(contingent commissions). To be sure, here the brokers dealt
“exclusively” with multiple parties—the exclusive dealing
involved individual insurance policies (most notably those
already placed with a particular insurer and up for possible
renewal), rather than a broker’s entire roster of clients—but this
difference does not materially alter the basic exclusivity-for-
kickbacks model. It merely presents multiple, parallel
instantiations of that model.

80
decisions to become strategic partners of the broker do not in
themselves bespeak a horizontal agreement, plaintiffs contend
their allegations about the “devices” used to conduct the
customer-steering schemes suffice to meet the Twombly
threshold.

According to the complaints, several of the devices that
allegedly facilitated the schemes are common to all of the
broker-centered conspiracies. For instance, plaintiffs allege that
brokers often afforded insurer-partners “first looks” and “last
looks” in bidding on policies. Once again, however, the
practices identified by plaintiffs are strictly vertical in nature.
On the complaint’s own account, first and last looks were
techniques utilized by brokers to ensure that a given client’s
policy was placed (or remained) with a designated insurer-
partner. See, e.g., Comm. SAC ¶ 88 (“Broker Defendants
shielded their insurer partners from normal competition by
agreeing not to bid renewals competitively, or by limiting the
circumstances under which renewals could be marketed. Broker
Defendants also routinely promised to provide competitive
advantages to Insurer partners, by disclosing other carriers’ bids,
providing first or last looks, and other methods.”). The
complaints describe “[t]he close bond between broker and
client,” which “gives brokers tremendous influence, and often
decisive control, over the placement of their clients’ insurance
business. Given the high degree of financial investment and
trust placed in their broker, clients will rarely if ever seek quotes
from insurers other than those recommended by the broker.” Id.

81
¶ 73. In other words, the complaints themselves provide
obvious reasons to conclude that the brokers were able to steer
clients to preferred insurers without the need for any agreement
among the insurers. Whatever the vices of these steering
techniques, they do not give rise to a plausible inference of
horizontal conspiracy.

Also insufficient are two allegations of certain “bid
manipulation” within the broker-centered conspiracies in the
Employee Benefits Case. In the first example, the complaint
asserts only that a broker unilaterally refused to submit an
insurer’s bid to the client. In the second, a broker successfully
persuaded one of its insurer-partners not to withdraw a bid the
insurer had come to view as unacceptably low. If the insurer
had withdrawn the bid, another, non-partner insurer would have
become a “finalist,” an outcome the broker wished to avoid. To
allay the insurer-partner’s concerns, the broker assured it that it
would not end up winning the contract because another insurer
had submitted an even lower bid. Shortly afterward, the broker
placed a large account with the insurer-partner. Neither
example provides a plausible basis for inferring anything more
than vertical agreements between brokers and individual
insurers.

In the Employee Benefits Case, plaintiffs allege that
defendant insurers used similar strategies to evade their
obligation to report contingent commission payments on Form
5500. But the asserted fact that the insurers intended to violate
their reporting obligations, and that they all adopted the same

82
deceptive reporting model, does not plausibly suggest a
horizontal agreement. If anything, the allegations suggest that
each insurer would be independently motivated to evade the
requirement, and that each had access to the same effective
model of how to accomplish this deception. Cf. In re Elevator
Antitrust Litig., 502 F.3d 47, 51 (2d Cir. 2007) (observing that
“similarities in contractual language . . . do not constitute
‘plausible grounds to infer an agreement’” because “[s]imilar
contract language can reflect the copying of documents that may
not be secret”). The insurers would be disinclined to expose
their competitors’ reporting violations for fear of calling
attention to their own self-interested deception. Cf. Twombly,
550 U.S. at 568 (finding that the failure of the defendants to
compete in one another’s regions was most plausibly explained
by the fact that the defendants “liked the world the way it was,
and surely knew the adage about him who lives by the sword”).

In sum, the allegations discussed thus far do not provide
“plausible grounds to infer” a horizontal agreement. Id. at 556.
This does not mean that defendants’ alleged treatment of
insurance purchasers was praiseworthy—or even lawful—but
that it fails to plead a per se violation of § 1 of the Sherman Act.
Plaintiffs have pled facts showing that brokers deceptively
steered their clients to preferred insurer-partners in order to
obtain contingent commission payments from those partners, but
this in itself is insufficient to plausibly imply a horizontal
conspiracy.

83
 (b) Bid-Rigging Allegations

There is, however, one notable exception to this
conclusion. In the Marsh-centered commercial conspiracy,
plaintiffs provide detailed allegations of bid rigging by the
insurer-partners.32 According to these allegations, insurers
furnished purposefully uncompetitive sham bids on policies in
order to facilitate the steering of business to other insurer-
partners, on the understanding that the other insurers would later
reciprocate. Bid rigging—or more specifically, as alleged in this
case, bid rotation 33 —is quintessentially collusive behavior

32
Apart from the multiple, detailed incidents of bid rigging in
the Marsh-centered commercial conspiracy, plaintiffs appear to
allege one incident of bid rigging in each of the Willis-centered
and Gallagher-centered commercial conspiracies. Comm. SAC
¶¶ 275, 336. In their briefs and at oral argument, however,
plaintiffs’s bid-rigging discussion appears to be limited to Marsh
and its insurer-partners. See, e.g., Tr. of Oral Arg. 12 (affirming
that “[t]he specific instances of bid rigging . . . occurred with
respect to the Marsh broker centered conspiracies [sic]”).
33
See United States v. Heffernan, 43 F.3d 1144, 1146 (7th
Cir. 1994) (contrasting bid rotation, in which “for each job the
competitors agree which of them shall be the low bidder, and the
others submit higher bids to make sure the designated bidder
wins,” with identical bidding, in which the competitors all agree
to bid the same price).

84
subject to per se condemnation under § 1 of the Sherman Act.
See United States v. All Star Indus., 962 F.2d 465, 469–73 (5th
Cir. 1992); see also United States v. Heffernan, 43 F.3d 1144,
1147 (7th Cir. 1994) (citing United States v. Portsmouth Paving
Corp., 694 F.2d 312, 317 (4th Cir. 1982)) (noting that bid
rotation may be especially anticompetitive because it
“eliminate[s] all competition rather than just price
competition”); 12 Hovenkamp, Antitrust Law ¶ 2006, at 77 (2d
ed. 2005) (“[B]id-rigging and bid rotation schemes are really
nothing more than output or market share agreements.”).34 This
point does not quite end our inquiry, as plaintiffs do not seek to
hold defendants liable for a bid-rigging conspiracy, but instead
proffer the alleged bid rigging as circumstantial evidence of a
“broader” agreement. Accordingly, we must assess the bid-
rigging allegations, like the other alleged circumstantial

34
As one treatise explains:
A strong inference of coordinated behavior arises
when a participant actively seeks to lose a bid.
Deliberate sacrifice of a contract implies an
unusual confidence that the winning party will
return the favor. Moreover, spurious bidding
indicates an awareness of wrongdoing coupled
with a desire to hide it by simulating normal
bidding. A spurious bid is almost always
anticompetitive . . . .
6 Areeda & Hovenkamp, supra, ¶ 1420b, at 140.

85
evidence discussed above, to determine whether, if true, they
plausibly imply the existence of the horizontal agreement on
which plaintiffs’ claim is predicated (and if so, whether that
agreement is subject to per se condemnation). For the reasons
that follow, we believe the bid-rigging behavior does plausibly
suggest concerted action by the insurers; it proffers “enough fact
to raise a reasonable expectation that discovery will reveal
evidence of illegal agreement,” Twombly, 550 U.S. at
556—more specifically, a horizontal agreement among the
insurers not to compete for one another’s incumbent business.

The District Court did not find the bid-rigging allegations
sufficient to imply any sort of horizontal agreement among
Marsh’s insurer-partners, even one to rig bids. The court
appears to have believed that because Marsh, the broker, was the
one who directed the insurers to provide sham bids, the bid
rigging was not indicative of an agreement among insurers but
simply reflected the desire by individual insurers to propitiate
Marsh in order to ensure that Marsh would continue to steer
premium volume their way. See 2007 WL 2533989, at *16–17
(acknowledging that “Plaintiffs presented a panoply of facts . . .
which allege that certain actions were taken by the Insurer
Defendants at the request of the Broker Defendants, such as . . .
protective bidding and bid-rigging,” but concluding that “[t]he
fact that Broker Defendants demanded or expected certain
behavior from the Insurer Defendants does not necessarily
amount to a horizontal agreement amongst the Defendant

86
Insurers.”).35

We agree that plaintiffs’ allegations portray a conspiracy
masterminded and directed by defendant broker Marsh, but this
fact does not make implausible the inference of a horizontal
agreement among the insurers. If the defendant insurers
supplying sham bids were truly indifferent as to whether
Marsh’s other insurer-partners would ever reciprocate, then the
bid rigging might not plausibly imply a horizontal agreement.36
On a motion to dismiss, however, we must assume the truth of
the complaint’s statement of facts, and the complaint here sets
forth a plausible basis for inferring that each bid-rigging
defendant’s decision not to compete was conditioned on an
expectation of reciprocity from its competitors—and not based
purely on independent motivation or broker Marsh’s behavior,
as the District Court concluded. See Comm. SAC ¶ 109

35
We note that, under Twombly, the test is not whether
plaintiffs’ allegations necessarily amount to an unlawful
horizontal agreement, but rather whether they plausibly
imply—that is, “raise a reasonable expectation that discovery
will reveal evidence of”—such an agreement. 550 U.S. at 556.
36
This aspect of the District Court’s reasoning as to why the
bid rigging does not imply a horizontal agreement is more fully
developed in its evaluation of the RICO claims. See 2007 WL
2892700, at *21. Accordingly, the bulk of our analysis on this
point occurs in Section II.B.2.a.i. infra.

87
(quoting statement by a former employee of a defendant insurer
to the effect that the Insurer had agreed to “provide[] losing
quotes” to its broker-partner in exchange for, among other
things, the broker’s “getting ‘quotes from other [insurance]
carriers that would support the [Insurer, at least when it was the
incumbent carrier] as being the best price’”).

The fact that Marsh, an entity vertically oriented to the
insurers, appears to be a sine qua non of the alleged horizontal
agreement is not necessarily an obstacle to plaintiffs’ claim. As
one of our sister courts of appeals has written, “defendants
cannot escape the per se rule [for certain horizontal restraints of
trade] simply because their conspiracy depended upon the
participation of a ‘middle-man’, even if that middleman
conceptualized the conspiracy, orchestrated it . . . and collected
most of the booty.” All Star, 962 F.2d at 473.

The conspiracy alleged in All Star has some striking
similarities with the broker-centered conspiracy alleged here. In
All Star, a criminal prosecution for antitrust conspiracy in the
specialty pipe industry, the government’s theory was that
defendant Texas Pipe Bending Company (TPB), which
performed fabrication jobs on a cost-plus basis, coordinated a
bid-rigging scheme among defendant pipe distributors. The
distributor(s) designated to win a particular bid would be
protected by higher bids submitted by the other bidders, and the
winning distributors rebated some portion of their sales
revenue—which was significantly inflated over the price that
would have prevailed in competitive bidding—to TPB. Id. at

88
467–68. In both All Star and (as alleged) this case, competitors
agreed to submit intentionally uncompetitive bids in order to
dictate the firm to which a particular contract would be awarded,
as well as (by implication if not design) the price of that
contract. This conduct plausibly implies a horizontal
conspiracy, and the fact that here it was the broker, Marsh, that
allegedly designated the winner and solicited the sham bids does
not alter that conclusion. Marsh may have been an essential
conduit and coordinator, but the insurers’ agreement to provide
protective bids to one another was also instrumental to the
operation of the asserted broker-centered conspiracy. Even if
the broker could have allocated customers on its own, without
enlisting the assistance of other insurer-partners, the alleged
willingness of those partners not only to refrain from competing
with one another, but also actively to assist in the deceptive
steering practices, plausibly suggests that customer allocation
could be the result not only of vertical collusion, but also of a
horizontal agreement among the insurers.37 The anticompetitive
danger inherent in insurers’ alleged concerted efforts to rotate
bids is not necessarily mitigated by the fact that the broker

37
As noted, it may be more precise to say that allegations of
brokers’ unilateral acts of fraud against their clients, while
undeniably asserting a form of consumer injury, do not plead an
injury to competition, which is adequately alleged in the Marsh-
centered scheme only by virtue of the well-pled horizontal
agreement among the insurer-spokes. See supra note 31 and
accompanying text.

89
managed the details of each bid, nor by the likelihood that the
horizontal collusion would not have occurred without the
broker’s involvement.

On appeal, defendants do not dispute that the bid-rigging
allegations plausibly imply a horizontal agreement among the
insurers. For several reasons, however, they contend this
agreement is insufficient to support plaintiffs’ antitrust claims.
Defendants do not deny that plaintiffs have set forth
particularized allegations of unlawful bid rigging, but they
contend that plaintiffs have no standing to challenge this activity
because plaintiffs do not assert that the bids were rigged on any
of the policies they purchased. Plaintiffs, in turn, insist that this
argument misses the point, since their claim is not that
defendants engaged in an actionable bid-rigging conspiracy; as
noted, the alleged horizontal agreement on which they base their
§ 1 claim is not an agreement to rig bids. Instead, they complain
of a “broader scheme” of “incumbent protection,” and the
incidents of bid rigging are alleged as evidence of this “broader
scheme.” Tr. of Oral Arg. 70.38

38
At oral argument, counsel for plaintiffs explained: “[T]he
defendants take a lot of time talking about how we can’t win in
a big [sic] rigging scheme because we didn’t allege a bid rigging
scheme. And that’s right. We have [instead] alleged an
agreement among these participants in the Marsh broker-
centered conspiracies . . . to protect each other’s incumbent
business.” Tr. of Oral. Arg. 72; see also Letter from Plaintiffs

90
 To evaluate the merit of this argument—that is, to
determine whether the bid-rigging allegations satisfy Twombly’s
pleading standard—it is necessary to identify the scope of this
“broader scheme” with precision. This imperative derives from
the requisite elements of a claim under § 1 of the Sherman Act.
As noted, since plaintiffs have elected to forego a rule-of-reason
analysis, they must adequately plead (1) a horizontal agreement
among insurers (2) to engage in an unreasonable restraint of
trade.39 Plaintiffs might be able to allege some sort of horizontal

to the District Court, No. 04-5184, Dkt. Entry # 669, at 2
(“[P]laintiffs do not allege that defendants are liable under the
antitrust laws because they engaged in ‘bid-rigging.’ Instead,
the theory of the Complaint is that defendants are liable under
the antitrust laws because they participated in a conspiracy to
allocate customers, using, on some occasions, bid-rigging, last
looks and other manipulative devices as overt acts to achieve the
conspiracies’ end.”).
39
Furthermore, because of the way plaintiffs have pled their
claim, plaintiffs must plead a type of horizontal restraint that can
be deemed unreasonable without evaluation of market power.
See Leegin, 551 U.S. at 886 (“Restraints that are per se unlawful
include horizontal agreements among competitors to fix prices
or to divide markets.” (internal citations omitted)); cf. R.C. Dick
Geothermal Corp. v. Thermogenics, Inc., 890 F.2d 139, 162 (9th
Cir. 1989) (en banc) (Norris, J., dissenting) (citing NCAA, 468
U.S. 85) (noting that the Supreme Court has “recognized a

91
agreement among defendants, the object of which would
nonetheless not amount to an unreasonable restraint of trade.
Alternatively, they might be able to allege that defendants
engaged in activity unreasonably restraining trade, but
nonetheless fail to plead that this conduct was the product of an
agreement. In both cases, plaintiffs would have failed to plead
a § 1 claim. Accordingly, we must define the object of the
horizontal agreement alleged in the complaint. See generally 6
Areeda & Hovenkamp, supra, ¶ 1409, at 54 (noting the
importance of “ask[ing] precisely (1) who was in agreement
with whom, and (2) about what?”).

Having reviewed the complaint, we believe it asserts two
different conceptions of this horizontal agreement. According
to the broader of the two conceptions, Marsh’s insurer-partners
agreed that Marsh would deliver to each insurer an amount of
premium volume necessary to trigger the payment of a
contingent commission under the vertical agreement between
Marsh and that insurer. See Comm. SAC ¶ 130 (“[Premium]
volume threshold commitments reflected a tacit agreement
among the conspiring parties that Marsh was guaranteeing the
delivery of a specified minimum amount of premium volume.”).
Reading the complaint in the light most favorable to plaintiffs,

caveat to the per se rule against horizontal restraints on
competition[,] holding that some horizontal relationships have
unique aspects that can create procompetitive justifications for
particular horizontal restraints”).

92
we find such a horizontal agreement implausible. Given the
context presented by plaintiffs, it is not plausible that the
insurers agreed among themselves that a third party, the broker,
would guarantee delivery of differing amounts of premium
volume to each of them. Perhaps such a claim would be
coherent if the insurers had power to extract such guarantees
from the broker, but the complaint demonstrates in abundant
detail that it was Marsh who held the reins. Plaintiffs note that
the contingent commission thresholds were established in
vertical agreements between the broker and each insurer, and
they recount stories of insurers who balked at Marsh’s demands
and refused to continue to pay contingent commissions, only to
relent and agree to resume payments after Marsh steered a
significant volume of business away from them. At the same
time, however, plaintiffs incongruously assert that the
contingent commission thresholds in Marsh’s contracts with
each of its insurer-partners were somehow the product of an
agreement among all of the insurers. This attempt to bootstrap
vertical contracts into horizontal conspiracy is at odds with both
“common economic experience,” Twombly, 550 U.S. at 565, and
the complaint’s own factual allegations, cf. id. at 568.

The complaint also posits a narrower agreement among
Marsh’s insurer-partners, namely, an agreement not to compete
for other partners’ incumbent business. See, e.g., Comm. SAC
¶ 89 (“[T]he Broker Defendants orchestrated a horizontal
agreement among rival Insurers not to compete for each others’
[sic] customers.”). Unlike the previous alleged agreement, this

93
one is not necessarily incompatible with the complaint’s account
of a market in which Marsh pulled most of the strings and called
most of the shots. The complaint alleges that Marsh prepared
broking plans “when an account was up for renewal. The
broking plans assigned the business to a specific insurer at a
target price and outlined the coverage. . . . If the incumbent
Insurer hit the ‘target’, it would get the business . . . .” Id. ¶ 117.
An agreement by the insurers not to compete with the incumbent
designated by Marsh would obviously facilitate Marsh’s
placement goals. That the bid-rigging allegations refer not to
closed, bilateral agreements in which insurers X and Y each help
the other win a specific account, but rather to open-ended
agreements in which insurer X provides “protection” of Y’s
“renewal” or “incumbent” account in exchange for an assurance
of similar assistance from some other insurer (not necessarily Y)
plausibly supports the inference that the bid rigging was in
service of a broader agreement not to compete for one another’s
incumbent business. As we have seen, plaintiffs allege that the
customer allocation schemes employed other mechanisms that
do not appear to have entailed a horizontal agreement among the
insurers, but this does not alter the fact that the bid-rigging
allegations plausibly imply a “broader” horizontal non-
competition agreement designed to aid the posited (broader still)
customer allocation scheme instigated by Marsh.

Nonetheless, one might reasonably ask (especially in
light of the allegations involving the other broker-centered
schemes) whether the insurers had an opportunity to compete in

94
the first place—that is, an opportunity other than that afforded
by Marsh’s solicitations of sham bids. An agreement not to
compete necessarily presupposes the existence of an opportunity
to compete, and if the only opportunities for insurers to compete
were Marsh’s requests for rigged bids, 40 then the alleged bid
rigging could not imply a “broader” horizontal agreement not to
compete for incumbent business. And in fact, certain allegations
in the complaint might be read to suggest that the solicitation of
rigged bids provided the only opportunity for insurers to
compete, that Marsh would either steer clients to the target
insurers on its own, or, in the rare cases when clients required it
to show them bids from multiple insurers,41 would solicit sham

40
The complaint shows how in providing these intentionally
non-competitive bids, the insurers necessarily passed up the
opportunity to compete. According to the complaint, one
insurer who was dissatisfied by Marsh’s protection of its own
incumbent business contemplated supplying competitive bids in
response to Marsh’s request for non-competitive offers. “If we
can not get proper protection,” the insurer stated, “we will go
hard after [another insurer’s incumbent business] that we feel
[Marsh is] protecting. We will no longer provide [Marsh] with
protective quotes for [that insurer] but will put out quotes that
[Marsh] will be forced to release . . . .” Comm. SAC ¶ 107.
41
See Comm. SAC ¶ 73 (“Given the high degree of financial
investment and trust placed in their broker, clients will rarely if
ever seek quotes from insurers other than those recommended

95
bids from other insurer-partners. See, e.g., id. ¶ 109 (“Marsh
would protect the incumbent of an excess casualty risk by not
sending submissions on that risk out to competition, or by
getting quotes from other carriers that would support the
incumbent as being the best price.” (internal quotation marks
omitted)).

In reviewing a motion to dismiss, however, we “construe
the complaint in the light most favorable to the plaintiff.”
Phillips, 515 F.3d at 233 (internal quotation marks omitted).42

by the broker.”).
42
As the Supreme Court reiterated in Iqbal, the Twombly
standard does not impose a “probability requirement.” Iqbal,
129 S. Ct. at 1949 (quoting Twombly, 550 U.S. at 556); it does
not require as a general matter that the plaintiff plead facts
supporting an inference of defendant’s liability more compelling
than the opposing inference. Twombly requires the plaintiff to
plead only enough “factual content [to] allow[] the court to draw
[a] reasonable inference that the defendant is liable for the
misconduct alleged.” Id. (emphasis added). Accordingly, “[i]t
remains an acceptable statement of the standard [for reviewing
a motion to dismiss under Rule 12(b)(6)] . . . that courts accept
all factual allegations as true, construe the complaint in the light
most favorable to the plaintiff, and determine whether, under
any reasonable reading of the complaint, the plaintiff may be
entitled to relief.” Phillips, 515 F.3d. at 233 (internal quotation

96
Accordingly, we do not interpret the complaint as disavowing
the possibility of opportunities to compete beyond those
afforded by Marsh’s bid-rigging requests. In any case,
defendants themselves have not advanced a no-other-
opportunity-to-compete argument in support of their motion to
dismiss. They may, of course, raise this objection at a
subsequent stage of the proceedings.

Defendants argue that plaintiffs have alleged only
“isolated episodes” of bid rigging. Defendants’ Comm. Br. 43.
To the extent defendants object that the allegations of bid
rigging within the Marsh-centered commercial conspiracy
cannot support claims of horizontal agreements within other
alleged broker-centered conspiracies, their point is well-taken.
But insofar as defendants contend that the bid-rigging
allegations do not adequately support the more general
allegation of an agreement among the defendant insurers to
allocate customers in the Marsh-centered commercial
conspiracy, we reject their argument for the reasons given. At
this stage of the litigation, Rule 8(a)(2) requires plaintiffs to
plead only “enough fact to raise a reasonable expectation that
discovery will reveal evidence of illegal agreement,” Twombly,

marks omitted). As noted, of course, Twombly makes clear that
in the specific context of a claim under § 1 of the Sherman Act,
it is unreasonable to infer an agreement from allegations of
parallel conduct that are equally consistent with independently
motivated behavior. See Twombly, 550 U.S. at 556–57.

97
550 U.S. at 556,43 in this case an agreement among the Marsh
partner-insurers not to compete for renewal business. We find
that the complaint satisfies this standard with respect to those
participants in the asserted Marsh-centered commercial
conspiracy who allegedly engaged in bid rigging.44

43
As the Supreme Court explained:
In applying the[] general standards [of Rule
8(a)(2)] to a § 1 claim, we hold that stating such
a claim requires a complaint with enough factual
matter (taken as true) to suggest that an agreement
was made. Asking for plausible grounds to infer
an agreement does not impose a probability
requirement at the pleading stage; it simply calls
for enough fact to raise a reasonable expectation
that discovery will reveal evidence of illegal
agreement.
Twombly, 550 U.S. at 556.
44
The number of defendants alleged to have engaged in bid
rigging appears to be slightly smaller than the number of
defendants alleged to be participants in the Marsh-centered
commercial conspiracy. Compare Comm. SAC ¶ 95 (naming
“AIG, ACE, CNA, Chubb, Crum & Forster, Hartford, Liberty
Mutual, Travelers, Zurich, Fireman’s Fund, Munich, XL and
Axis” as defendant insurers in the Marsh broker-centered
conspiracy), with Plaintiffs’ Comm. Br. 78 n.17 (claiming that
the defendant insurers that engaged in bid rigging are “AIG,

98
ACE, Axis, Chubb, XL, Munich/AmRe, Liberty Mutual, St.
Paul Travelers, Fireman’s Fund, and Zurich”), and Comm. RPS
¶¶ 27–56 (detailing bid-rigging allegations).
Our disposition must also take account of the fact that
although the complaint’s narrative of wrongdoing speaks
primarily (if not exclusively) in terms of parent entities or
corporate groups, subsidiary corporate entities are also named as
individual defendants. See Comm. SAC ¶¶ 37–63 (stating that
the use of the parent or group entity name is meant to
incorporate the subsidiaries by reference). Defendants contend
that the bid-rigging allegations are limited to a single line of
commercial insurance, namely excess casualty. Plaintiffs appear
to concede this point. See Plaintiffs’ Comm. Reply Br. 11
(referring to the “Marsh Excess Casualty conspiracy”). As
noted, without the bid-rigging allegations, plaintiffs have not
stated “enough factual matter . . . to suggest that an agreement
was made” among the insurers. Twombly, 550 U.S. at 556.
Accordingly, any subsidiary entities not alleged to have dealt in
excess casualty (and thus not alleged to have engaged in bid
rigging) must be dismissed, as the complaint fails to plausibly
imply that they entered into a horizontal agreement to
unreasonably restrain trade.
Plaintiffs argue that subsidiary companies “act[] at the
common direction of the parent[],” Plaintiffs’ Comm. Reply Br.
12, and that “in reality a parent and a wholly owned subsidiary
always have a unity of purpose or a common design,” Plaintiffs’

99
EB Reply Br. 35 (quoting Copperweld, 467 U.S at 771) (internal
quotation marks omitted). Emphasizing these features of the
parent-subsidiary relationship, the Supreme Court held in
Copperweld that parents and subsidiaries could not conspire for
purposes of § 1 of the Sherman Act. 467 U.S. at 776; see Am.
Needle, 130 S. Ct. at 2212 (noting that an “agreement” is
cognizable under § 1 only if it “joins together ‘independent
centers of decisionmaking’” (quoting Copperweld, 467 U.S. at
769)). Contrary to plaintiffs’ suggestion, however, it does not
follow from Copperweld that subsidiary entities are
automatically liable under § 1 for any agreements to which the
parent is a party. As a matter of well-settled common law, a
subsidiary is a distinct legal entity and is not liable for the
actions of its parent or sister corporations simply by dint of the
corporate relationship. See 1 William Meade Fletcher,
Cyclopedia of Law of Private Corporations § 33, at 89 (perm.
ed. rev. vol. 2006) (observing that “the mere fact that there
exists a parent-subsidiary relationship between two corporations
[does not] make the one liable for the torts of its affiliates”); see
also Burks v. Lasker, 441 U.S. 471, 478 (1979) (“Congress has
never indicated that the entire corpus of state corporation law is
to be replaced simply because a plaintiff’s cause of action is
based upon a federal statute.”). As plaintiffs allege no other
basis for imputing § 1 liability to defendant entities that are not
plausibly alleged to be directly liable—that is, are not plausibly
alleged to have themselves entered into unlawful

100
 Defendants attempt to resist this conclusion with a
number of different arguments, but after due consideration we
find none have merit. According to defendants, the scheme
alleged by plaintiffs is incoherent. To illustrate its
implausibility, defendants contrast it with the conspiracy at issue
in Petruzzi’s IGA Supermarkets, Inc. v. Darling-Delaware Co.,
998 F.2d 1224 (3d Cir. 1993). The Petruzzi’s plaintiff alleged
a conspiracy to allocate customers in the fat and bone rendering
industry. Id. at 1228. More specifically, the plaintiff claimed
that although the defendant rendering companies would compete
for new accounts, once an account was won the non-incumbent
defendants would not compete over renewal business and
sometimes “put forward sham bids.” Id. at 1228–29. If any
defendant violated the agreement, the remaining conspirators
would purportedly punish it through predatory pricing. Id.
Given the circumstances of the industry, we found that the
plaintiff’s theory of conspiracy was not only “not implausible,”
but made “perfect economic sense.” Id. at 1232.

Defendants contend that at least two salient features
distinguish the Petruzzi’s conspiracy from the one alleged here.
First, in Petruzzi’s the method for allocating business was
transparently obvious. Each conspirator could easily ascertain
which member of the scheme was entitled to a given
account—namely, the incumbent holder of the account. Here,
defendants argue, there is no way for an insurer to know with

agreements—the antitrust claims against these entities must fail.

101
which conspirator a given policy should be placed. Plaintiffs
propose that the allocation was structured not by particular
policies but by premium volume, but defendants insist that such
a basis of allocation would be unworkable in light of the various
contingent commission incentives detailed in the complaint. In
addition to contingent commission payments triggered by a
threshold volume of incumbent business retained, the
contractual agreements between the brokers and insurers also
provided for commission payments based on the overall volume
of premium steered to an insurer, growth in volume over a
particular benchmark (such as the previous year’s level), and the
quality of the premium volume (i.e., premiums for policies
requiring relatively small indemnification payments for covered
losses). Defendants contend that these multifarious incentives
would often conflict with the alleged scheme’s posited goal of
incumbent protection. For example, a broker’s placement of a
given policy with incumbent insurer X might bring the broker
that much closer to the negotiated contingent commission
threshold for premium volume renewed with that broker. But
placement of that same policy with another insurer might trigger
a contingent commission payment for overall premium volume
or volume growth—and that commission payment might be
larger than the one negotiated with the incumbent. “It defies
credulity,” defendants insist, “to assert, as Plaintiffs do, that . . .
insurers agreed to join conspiracies in which they agreed to
allow brokers to unilaterally decide who got what business
based on what was most profitable for the brokers.”
Defendants’ Comm. Br. 51.

102
 Second, defendants contend that while the scheme in
Petruzzi’s included an obvious mechanism for the conspirators
to discipline deviant members, the conspiracy alleged here is
“hardly a scheme of market allocation that the insurers could
enforce.” Tr. of Oral Arg. 43. According to defendants, since
virtually all of the power to steer insurance purchasers belonged
to the brokers, who operated under the competing incentives
created by the variegated contingent commission agreements,
there could be no feasible mechanism to enforce a customer
allocation scheme.

We agree with defendants that the scheme alleged by
plaintiffs appears a good deal more complex than the one in
Petruzzi’s. And as noted, we agree that based on the facts
alleged, it is implausible to claim that the defendant insurers
came to an agreement together and instigated an arrangement
whereby each would receive whatever volume of premium
happened to be prescribed by each’s contingent commission
agreement with Marsh. But as also noted, a narrower horizontal
agreement not to compete for one another’s incumbent business
does not appear incompatible with the larger picture painted by
the complaint, in which Marsh was the dominant force.

The complaint also provides a coherent mechanism for
disciplining recalcitrant insurers. Consistent with the
complaint’s general narrative of broker power, it was Marsh that
did the enforcing. In a vivid illustration of this enforcement
potential, the complaint recounts the following alleged statement
from a high-ranking Marsh executive:

103
 [I]f an alternative [i.e., a non-incumbent insurer
from which Marsh has solicited a sham bid]
quotes below [the incumbent insurer’s target bid]
then they have made a conscious decision to quote
below [the incumbent insurer] and pull [the
incumbent] down. If that happens, then . . . we
will put this guy in open competition on every
acct. and CRUCIFY him. Further, we must make
sure [the] incumbent [or another insurer] keep[s]
this [account] and NOT give it to the alternative
and reward them.

Comm. SAC ¶ 118 (emphasis omitted). According to the
complaint, insurers who breached the non-competition
agreement would not only find themselves deprived of the
conspiracy’s protection, but their renewal business would be
specifically targeted for transfer.

Although we acknowledge that the hub-and-spoke
conspiracy alleged by plaintiffs has a more prominent vertical
dimension than most, if not all, other examples found in the case
law—owing to the relative power of broker Marsh and the
relative dependence of its insurer-partners—we believe the
complaint contains enough well-pled factual matter to suggest
a plausible horizontal agreement among the insurers not to
compete for renewal business. On the complaint’s own account,
the conspiracy was instigated, coordinated, and policed by
Marsh, but this does not belie the alleged horizontal agreement.
On the contrary, Marsh’s influence could create a powerful

104
incentive for exactly such an agreement: join and enjoy renewals
at inflated premium rates and without threat of competition, or
remain outside the “strategic partnership” and be denied access
to Marsh’s large and loyal clientele. To be sure, the complaint
suggests that Marsh could be a tough master, threatening at
times to transfer business to another insurer in order to coerce a
more lucrative contingent commission agreement. And in some
cases, as defendants suggest, Marsh may even have steered
renewal business away from an incumbent insurer-partner in
order to realize a more profitable commission offered by another
partner.45 If so, however, this would show only that Marsh, and

45
We find the complaint somewhat ambiguous on this
question. Plaintiffs allege that under the customer allocation
scheme, “each conspiring insurer would be permitted to keep its
own incumbent business.” Comm. SAC ¶ 96. But as
defendants point out, the alleged contingent commission
agreements tied commissions to factors other than incumbent
business, which might motivate Marsh to transfer business away
from incumbents. Plaintiffs contend that Marsh only used new
business and business transferred from non-partner insurers to
satisfy these thresholds. More problematic for plaintiffs’ claim
of guaranteed incumbent protection may be the complaint’s
statement that Marsh “grouped its preferred insurers into three
tiers, classified as A, B, and C tiers, based on how much they
were paying in contingent commissions. Tiers A and B were the
more preferred markets to which the bulk of premium was
allocated.” Comm. SAC ¶ 101. It is unclear from the

105
not the insurers, had the negotiating power to set the terms of
participation in the scheme. It does not make implausible the
inference, created by the bid-rigging allegations, that insurer-
partners agreed not to compete for one another’s renewal
business. As we have noted, plaintiffs’ allegations paint a
conspiracy in which the hub, Marsh, held an unusual amount of
power and may even have been able economically to “coerce”
the insurers into the non-competition agreement. Defendants
have failed, however, to show why this feature would preclude
per se condemnation of the horizontal agreement. See 6 Areeda
& Hovenkamp, supra, ¶ 1408c (“[S]ociety prefers that coerced
parties seek the protection of public authorities rather than help
create a cartel.”).

Defendants next argue that “even if there were
agreements that could have existed among the insurers,” the
vertical element of the hub-and-spoke conspiracy would defeat
plaintiffs’ claim. Tr. of Oral Arg. 43. In defendants’ view,
“horizontal restraints that are ancillary to vertical arrangements,
in other words horizontal agreements that exist to facilitate the
vertical ones, are judged under the rule of reason which the
plaintiffs have disclaimed.” Id. (citing United States v.
Addyston Pipe & Steel Co., 85 F. 271 (6th Cir. 1898), modified

complaint’s brief description whether incumbent business from
lower-tier insurers would sometimes be transferred to higher-tier
insurers or whether the “premium” mentioned came only from
new or non-insurer-partner-held accounts.

106
and aff’d, 175 U.S. 211 (1899)). Any horizontal agreement
among the brokers, defendants contend, “would plainly have
been ancillary to the agreements that those insurers had with the
brokers.” Id. Therefore, the horizontal agreements cannot be
condemned per se.

Defendants’ contention draws on a fundamental principle
of antitrust law but misapplies it here. It is well settled that
“ancillary” restraints of trade are less suspicious than “naked”
ones, and that to qualify for per se condemnation, a restraint
must be of the naked horizontal type. Polk Bros., Inc. v. Forest
City Enters., Inc., 776 F.2d 185, 188–89 (7th Cir. 1985).
Ancillary restraints are “those that are part of a larger endeavor
whose success they promote.” Id. at 189; see 11 Hovenkamp,
supra, ¶ 1904 (“To say that a restraint is ‘ancillary’ is to
conclude that it is an essential or at least an important part of
some arrangement that has potentially redeeming virtues.”). By
contrast, a “naked” restraint is one that is not an integral part of
an arrangement with redeeming competitive virtues. See Polk
Bros., 776 F.2d at 188 (describing naked restraints as “those in
which the restriction on competition is unaccompanied by new
production or products”); 11 Hovenkamp, supra, ¶ 1904. The
quintessential example of an ancillary restraint is a restrictive
agreement that is an integral part of a joint venture. An
agreement by two competing manufacturers to price a product
identically, for instance, would be ancillary if manufacture of
the product were a collaborative effort between the two firms
and the pricing agreement could reasonably be viewed as a

107
necessary condition of the joint venture, which increased output.

As this example indicates, “[d]etermining ancillarity
requires [courts] to consider first, whether any aspect of the
defendants’ association contains a significant promise of
integration or cooperation yielding an increase in output.
Second, some determination must be made whether the
challenged agreement is an essential part of this arrangement,
whether it is important but perhaps not essential, or whether it
is completely unnecessary.” 11 Hovenkamp, supra, ¶ 1908b, at
253 (footnote omitted); see MLB Props., Inc. v. Salvino, Inc.,
542 F.3d 290, 339 (2d Cir. 2008) (Sotomayor, J., concurring in
the judgment) (“[A] restraint that is unnecessary to achieve a
joint venture’s efficiency-enhancing benefits may not be
justified based on those benefits. . . . In contrast, where a
restraint is reasonably necessary to achieve a joint venture’s
efficiency-enhancing purposes (i.e., ancillary), it will be
analyzed under the rule of reason as part of the joint venture
because the effects of that restraint are not so plainly
anticompetitive as to make a per se or quick-look approach
appropriate.”); see also 11 Hovenkamp, supra, ¶ 1908b, at 253
(“[T]he ‘essentiality’ query . . . considers whether the challenged
restraint is an inherent feature of the joint venture at all, or
simply an unnecessary, output-limiting appendage.”); cf. NCAA,
468 U.S. at 114 (denying the NCAA’s claim that its television
plan enhanced the competitiveness of college football television
rights because in light of the district court’s findings, “it cannot
be said that the [challenged] agreement on price is necessary to

108
market the product at all” (internal quotation marks omitted)).
Contrary to defendants’ argument, then, a restraint is not
automatically deemed ancillary simply because it “facilitates” a
procompetitive arrangement.

Defendants are unable to identify among plaintiffs’
allegations any procompetitive venture to which the insurers’
alleged horizontal agreement not to compete for incumbent
business could reasonably be deemed integral. Defendants
stress that vertical “preferred provider” agreements, used here
by the brokers to consolidate the insurers with which they did
business, have consistently been found by courts to have
competitive benefits. But defendants cannot explain why a non-
competition agreement among those providers is an essential or
reasonably necessary component of those agreements; the
benefits of preferred provider agreements do not depend on such
a horizontal restraint of trade, as other provider agreements well
illustrate.46

46
See, e.g., Stop & Shop Supermarket Co. v. Blue Cross &
Blue Shield of R.I., 373 F.3d 57, 61–64 (1st Cir. 2004)
(recognizing potential consumer benefits of, and refusing to
apply the per se rule to, a vertical “exclusive dealing”
arrangement in which an insurer’s pharmacy manager created a
“closed network” of pharmacies providing prescription benefits
to the insurer’s subscribers); Kartell v. Blue Shield of Mass.,
Inc., 749 F.2d 922, 924, 932 (1st Cir. 1984) (Breyer, J.)
(rejecting physicians’ antitrust challenge to health insurer’s

109
 Defendants also contend that the insurers competed with
one another to win a favored position with Marsh. See supra
note 45 (discussing Marsh’s classification of its insurer-partners
into hierarchical tiers). Without doubt, according to the
complaint’s allegations, Marsh’s ability to guarantee
competition-free access to business—in part by enforcing a non-
competition agreement among its insurer-partners—motivated
the insurers to pay Marsh larger commissions in order to receive
a larger slice of the competition-free premium pie. We fail to
see, however, how this kind of rivalry among the insurers would
increase output in the market for insurance.47

preferred provider organization, which allegedly put pressure on
doctors to join “because of the large number of [patient]
subscribers,” and emphasizing that there was no “evidence of a
horizontal conspiracy”); Quality Auto Body, Inc. v. Allstate Ins.
Co., 660 F.2d 1195, 1203–04 (7th Cir. 1981) (describing
procompetitive benefits of insurance companies’ contracting
with a specific set of “preferred” auto repair shops for repair
work at prescribed rates).
47
The facts in the seminal case of Addyston Pipe offer a
useful comparison. There, a cartel of pipe manufacturers rigged
bids for pipes sold principally to municipalities. The winning
bidder would pay a “bonus” to the other bidders. First, the
amount of the winning bid was determined, and then the bidders
conducted a competitive bid among themselves to determine
who was willing to pay the largest bonus to the others. The

110
 Furthermore, defendants’ argument proves too much. If
all “horizontal agreements that exist to facilitate . . . vertical
ones,” Tr. of Oral Arg. 43, must be tested by the rule of reason,
then per se condemnation of hub-and-spoke conspiracies would
appear to be impossible. In all hub-and-spoke conspiracies, the
horizontal agreement among the spokes supports the agreements
between the hub and each spoke, and vice versa. See, e.g.,
Interstate Circuit v. United States, 306 U.S. 208 (1939); Toys
“R” Us, Inc. v. FTC, 221 F.3d 928 (7th Cir. 2000); United
States v. All Star Indus., 962 F.2d 465 (5th Cir. 1992).

winner of that competition obtained the right to submit the
winning bid in the rigged auction. The structure of the scheme
meant that the most efficient or lowest-cost manufacturer—and
thus the one able to pay the largest bonus to the others—would
generally win the rigged bid, just as in a competitive market.
But the conspiracy ensured that the price of the winning bid was
supracompetitive and output-reducing. Addyston Pipe, 85 F. at
274. The court did not find that the competition among the
manufacturers within the bounds of the bid-rigging scheme
redeemed the restraint.
Toys “R” Us provides another illustration. Suppose the
toy manufacturers in that case competed with one another over
the amount of product Toys “R” Us would buy from each. This
competition for Toys “R” Us’s business would not alter the
basic fact that the horizontal agreement to sell exclusively to
Toys “R” Us reduced output.

111
 Although we have found that the bid-rigging allegations
suffice to plead a § 1 claim for purposes of Federal Rule of Civil
Procedure 8(a)(2), defendants insist that plaintiffs must
surmount not only this general requirement, but also the
heightened pleading standard set forth in Rule 9(b). That Rule
provides that “[i]n alleging fraud or mistake, a party must state
with particularity the circumstances constituting fraud or
mistake.” Fed. R. Civ. P. 9(b).48 Defendants contend the
antitrust claims here “sound in fraud” and argue that the
complaints fail to satisfy Rule 9(b)’s particularity requirement.
Defendants’ Comm. Br. 31. Plaintiffs, on the other hand,
maintain Rule 9(b) is inapplicable to the alleged antitrust
conspiracies.

In Lum v. Bank of America, we stated that Rule 9(b)
requires fraud to be “pled with particularity in all claims based
on fraud.” 361 F.3d 217, 220 (3d Cir. 2004). Accordingly,
although we acknowledged that “antitrust claims generally are
not subject to the heightened pleading requirement of Rule
9(b),” we found that standard applicable to a complaint in which
“the an titrust claim [w as] based on fraud— on
misrepresentations in the information given to consumers and on
misrepresentations in the information given to . . . independent
financial publications.” Id.; see id. at 228 (“Because plaintiffs

48
The Rule nonetheless allows “[m]alice, intent, knowledge,
and other conditions of a person’s mind [to] be alleged
generally.” Id.

112
allege that the defendant[ banks] accomplished the goal of their
conspiracy [to set an artificially high floor on interest rates by
agreeing to raise the prime rate] through fraud, the Amended
Complaint is subject to Rule 9(b).”). Citing Lum, the District
Court here agreed with defendants that Rule 9(b) applied to
plaintiffs’ allegations of antitrust conspiracy, as they were
“predicated on fraud.” 2007 WL 1100449, at *8. As both
plaintiffs and defendants appear to agree,49 however, the District
Court’s final dismissal order rested only on Twombly’s general
pleading standard; the court did not appear to make a separate
determination as to whether plaintiffs’ allegations satisfied the
particularity requirement of Rule 9(b).50 Insofar as we find that

49
See Plaintiffs’ Comm. Br. 14 n.3; Defendants’ Comm. Br.
31.
50
In its initial October 3, 2006 opinion, the District Court
found that plaintiffs’ First Amended Complaint did “not
specifically identify the entities which allegedly conspired with
each Broker Defendant” in the alleged broker-centered
conspiracies. 2006 WL 2850607, at *13. Granting leave to
amend, the court instructed plaintiffs to file a “supplemental
statement of particularity” setting forth, “with the degree of
particularity required under Rule 9(b), the identity of the
conspirators and the role of each Defendant in the alleged
conspiracies.” Id. at *14. Reviewing these revised pleadings in
its second dismissal order filed on April 5, 2007, the court
pronounced itself “satisfied . . . with the level of specificity

113
plaintiffs have satisfied the Twombly standard with respect to
defendants alleged to have engaged in bid rigging in the asserted
Marsh-centered commercial conspiracy, we will remand for the
District Court to determine in the first instance the extent, if any,
to which Rule 9(b) applies to those § 1 claims, and whether
plaintiffs pleadings are sufficiently particularized.51 We express

contained in the Particularized Statements which identify the
majority of the conspirators and their roles in the conspiracy.”
2007 WL 1100449, at *15. But the court found that these
allegations were insufficient to show “that the conduct alleged,
i.e., the consolidation of the insurance markets and the steering
of certain customers based on contingent commission payments,
constitutes a per se illegal horizontal customer or market
allocation scheme.” Id. at *18. It is unclear whether Rule 9(b)
played a role in this determination. In its final antitrust opinion,
filed on August 31, 2007, and dismissing plaintiffs’ claims with
prejudice, the District Court appeared to apply only Twombly’s
plausibility standard. See, 2007 WL 2533989, at *18–19.
Having determined that both complaints failed to satisfy this
general standard, the court had no occasion to test plaintiffs’
allegations against the heightened pleading requirements of Rule
9(b).
51
The District Court’s Rule 9(b) analysis should be directed
to the specific antitrust conspiracy we have found adequately
pled for purposes of Rule 8(a)(2)—namely, a horizontal
agreement among certain of Marsh’s insurer-partners not to

114
no opinion on these issues here.

ii. The Global Conspiracy

Overlaying the broker-centered conspiracies, plaintiffs
aver, was a “global conspiracy.” In this alleged scheme, the
defendant brokers, “with the complicity of the Defendant
Insurers,” EB SAC ¶ 301, agreed “to conceal from the general
public and other brokers [i.e., non-conspiring brokers]” the
existence of the broker-centered conspiracies and the details of
the contingent commission agreements. Id. ¶ 314. Plaintiffs
contend that this “agreement not to disclose the Contingent
Commission agreements and resulting profits was a naked
horizontal restraint of informational output that directly affected
the price of insurance.” Id. ¶ 303.

The District Court concluded that the complaints’ factual
allegations fail to plausibly imply horizontal non-disclosure
agreements among the defendant brokers or the defendant
insurers. 2007 WL 2533989, at *19. We agree. Plaintiffs
explain that defendants engaged in similar broker-centered
schemes, which were all structured by similar contingent
commission agreements. Plaintiffs further note that these
schemes “were very successful and yielded enormous profits,”
and that “[t]he Broker and Insurer Defendants were thus heavily
invested in their Broker-Centered schemes during the Class
Periods and did not want to risk losing their resulting profits by

compete for incumbent business.

115
disclosing their schemes to each others’ [sic] clients.” EB SAC
¶ 313. In other words, no broker could expose its competitors’
contingent commission agreements without drawing unwelcome
attention to its own golden-egg-laying goose. Having just
cogently explained why each broker had ample independent
motive to avoid disclosure, however, the complaints
discordantly conclude: “Therefore [the defendants] agreed not
to [disclose].” Id. (emphasis added); accord Comm. SAC ¶¶
355–56. We cannot credit this ipse dixit, which is in conflict
with its own premises.

Plaintiffs contend that “[i]n a truly competitive
environment, brokers could utilize information about another
broker’s charging of supra-competitive premiums through
inclusion of Contingent Commissions . . . to compete for that
broker’s business. An economically rational broker would
maximize its opportunity to increase market share by telling its
rival’s customers they are paying too much for their insurance.”
EB SAC ¶ 315. But this argument fails, much like the Twombly
plaintiffs’ contention that the defendant ILECs’ reluctance to
challenge one another’s regional monopolies bespoke
agreement. See Twombly, 550 U.S. at 567–69. Reaping
“enormous profits” from their own furtive use of contingent
commission agreements, the brokers had no desire to upset the
apple cart.52 See id. at 568; see also id. (noting that the

52
The Employee Benefits complaint “alleges that the brokers
knew, through industry studies and other means, that disclosure

116
complaint “does not allege that competition . . . was potentially
any more lucrative” than the defendants’ behavior during the
relevant period).

Nor do plaintiffs’ other proffered “plus factors” plausibly
imply a horizontal agreement among the brokers. The
Commercial complaint alleges that the defendant brokers
“issued substantially similar purported ‘disclosure’ statements
modeled after the CIAB’s position statement” advising brokers
on how to respond to questions regarding contingent
commissions. Comm. SAC ¶ 452. According to plaintiffs,
these statements misleadingly disguised the existence and effect
of the contingent commission agreements. But neither
defendants’ membership in the CIAB, nor their common
adoption of the trade group’s suggestions, plausibly suggest
conspiracy. Cf. Twombly, 550 U.S. at 567 n.12 (rejecting the
contention that the defendants’ common membership in a trade

of [the contingent commission] arrangements would cause a
decrease in commission income (which was almost exclusively
profit to the brokers) of between 5% to 25%.” Plaintiffs’ EB Br.
20. But this fact does nothing to strengthen the inference that
the brokers’ similar silence on contingent commissions was the
product of an agreement. The “obvious alternative explanation”
remains: each broker decided, perhaps on the basis of the same
industry studies, that disclosure was not in its best interest, just
as each ILEC in Twombly decided that competition with the
other regional monopolies would not benefit its bottom line.

117
union, combined with parallel conduct, plausibly suggests
conspiracy); Elevator Antitrust Litig., 502 F.3d at 51 (finding
that allegations that the defendants used similar contractual
language did not plausibly imply conspiracy because “similar
contract language can reflect the copying of documents that may
not be secret”). While these allegations indicate that the brokers
had an opportunity to conspire, they do not plausibly imply that
each broker acted other than independently when it decided to
incorporate the CIAB’s proposed approach as the best means of
protecting its lucrative arrangements from hostile scrutiny. See
Petruzzi’s, 998 F.2d at 1242 n.15 (“Proof of opportunity to
conspire, without more, will not sustain an inference that a
conspiracy has taken place.” (internal quotation marks omitted)).

Even if we read the complaint to assert that the defendant
brokers collaborated in crafting these allegedly misleading
disclosures (insofar as these defendants allegedly “control the
affairs of . . . CIAB,” Comm. SAC ¶ 515, which produced the
“position statement” allegedly incorporated into defendants’
disclosures to clients), this still would be insufficient to show a
horizontal agreement not to disclose one another’s contingent
commissions. If proven, this allegation would plausibly show
that defendants agreed to work together to determine the best
way of disguising activity in which each engaged. But this
allegation would not plausibly imply that the decision to
disguise that activity (namely, the alleged use of contingent
commissions as part of a scheme to steer customers to particular
insurers) was itself the product of an agreement—not, at least,

118
in the face of the complaint’s many allegations showing that
each defendant had ample independent motive to conceal its
own contingent commission arrangements. A contrary holding
would be tantamount to finding that any collaborate effort to
refine a “pernicious industry practice,” In re Ins. Brokerage
Antitrust Litig., 2007 WL 2892700, at *24 (so describing the
conduct alleged by plaintiffs), plausibly suggests a conspiracy
among all industry participants not to reveal the fact that other
participants engage in the same practice. Where, as here, the
“obvious alternative explanation” for such an industry practice
is that each member of the industry believes its profits would
suffer without the practice, it is not plausible to infer that each
member’s decision not to expose its competitors’ use of the
practice—that is, not to engage in mutually assured
destruction—is the product of an agreement.53

53
Plaintiffs describe the alleged “Global Conspiracy” as “the
Broker Defendants’ agreement not to disclose or advertise
truthful pricing information and to limit consumer information
about price.” Comm. SAC ¶ 358. This language, however,
elides the significant difference between this case and those in
which defendants are alleged to have agreed to refrain from
disclosing information about their own practices. Cf. Cal.
Dental Ass’n v. FTC, 526 U.S. 756, 759–62 & n.1 (1999)
(describing FTC’s allegations that dentists agreed not to engage
in advertising about pricing discounts or quality of service).
When each defendant would be expected to have an independent
motive to disclose information about its own product or services

119
 In the Employee Benefits Case, plaintiffs allege that
“Defendants executed substantially similar disclosure policies
regarding contingent compensation matters, including failing to
disclose contingent compensation information to ERISA plan
administrators on Form 5500s, as required by governmental
regulations.” EB SAC ¶ 324. Plaintiffs also allege instances in
which defendants exchanged information about how they
accounted for, and reported, this compensation. These
allegations, like the other allegations of shared information and
similar disclosure practices, are insufficient. They imply only
that each defendant had a similar motive to obfuscate the
structure of the brokers’ compensation, and that they sought the
most effective means to achieve this obfuscation.54 They do not

(such as when that information would presumably enhance the
value of that product or services to potential customers), the fact
that defendants made parallel decisions not to do so conceivably
raises a suspicion of agreement. Here, by contrast, the
allegation is that defendants agreed not to disclose unflattering
information about their competitors’ practices. But given that
each defendant is alleged to have engaged in the same practice,
so that such disclosure would inevitably be self-defeating, the
inference that the lack of disclosure is the result of agreement is
implausible.
54
Plaintiffs do not contend that the exchange of information
about reporting techniques was itself unlawful, but argue only
that this communication evinces a horizontal agreement to report

120
provide a “reason to infer that the [defendants] had agreed
among themselves to do what was only natural anyway.”
Twombly, 550 U.S. at 566.

Finally, plaintiffs point to the similar nature of each
broker-centered conspiracy, as well as the allegedly similar
confidentiality agreements the brokers inserted into the vertical
contracts with each of their partner insurers. But these
allegations of parallel conduct do not qualify under Twombly as
a basis for a plausible inference of horizontal agreement among
the brokers or insurers. Having reviewed the entirety of the
Global Conspiracy pleadings, we concur with the District
Court’s conclusion: “While Plaintiffs present facts to support the
possibility of inadequate disclosures by the brokers to the
insureds, the Complaints are bereft of allegations to demonstrate
that this was more than brokers adopting sub-par disclosure
methods to protect their own, lucrative agreements.” 2007 WL
2533989, at *19. Plaintiffs’ attack on the pervasive use of
contingent commissions to exploit insurance brokers’ power
over their clients—and the use of similar techniques to disguise

the brokers’ compensation improperly on Form 5500. Notably,
some of the particular exchanges detailed by plaintiffs actually
appear to undermine the inference of an agreement not to
disclose. See, e.g., EB RPS ¶ 63 (noting an e-mail from one
insurer to another expressing “surprise[]” that the addressee had
not been reporting its commission payments “given our
conversation on this topic earlier this year”).

121
this activity—may allege a “pernicious industry practice,” but
they do not plausibly imply an industry-wide conspiracy.

2. The McCarran-Ferguson Act

Defendants argue that whether or not plaintiffs have
adequately pled the elements of their Sherman Act claims under
the Federal Rules of Civil Procedure, the conduct alleged in the
complaints is exempt from federal antitrust regulation under the
McCarran-Ferguson Act, 15 U.S.C. §§ 1011–1015. Section 2(b)
of the McCarran-Ferguson Act provides:

No Act of Congress shall be construed to
invalidate, impair, or supersede any law enacted
by any State for the purpose of regulating the
business of insurance, or which imposes a fee or
tax upon such business, unless such Act
specifically relates to the business of insurance:
Provided, That . . . the Sherman Act . . . shall be
applicable to the business of insurance to the
extent that such business is not regulated by State
law.

15 U.S.C. § 1012(b). Section 3(b) of the Act provides that
“[n]othing contained in this chapter shall render the said
Sherman Act inapplicable to any agreement to boycott, coerce,
or intimidate, or act of boycott, coercion, or intimidation.” 15
U.S.C. § 1013(b).

The second, proviso clause of Section 2(b)—which is the

122
clause relevant to this appeal—“provides a statutory antitrust
exemption for activities that (1) constitute the ‘business of
insurance,’ (2) are regulated pursuant to state law, and (3) do not
constitute acts of ‘boycott, coercion or intimidation.’” Ticor
Title Ins. Co. v. FTC, 998 F.2d 1129, 1133 (3d Cir. 1993)
(quoting Group Life & Health Ins. Co. v. Royal Drug Co., 440
U.S. 205, 219–20 (1979)). Rejecting defendants’ arguments, the
District Court concluded that the conduct alleged by plaintiffs
is not part of the “business of insurance” and is thus not exempt
from plaintiffs’ antitrust claims. See Royal Drug, 440 U.S. at
210 (noting that if the challenged agreements “are not the
‘business of insurance’ within the meaning of § 2(b) of the
McCarran-Ferguson Act . . . then the Agreements are not
exempt from examination under the antitrust laws”). We agree.

The Supreme Court observed in Royal Drug that “the
[Act] does not define the ‘business of insurance.’” 440 U.S. at
211. But the Court noted that, insofar as the Act limits the scope
of federal antitrust statutes, its interpretation is subject to the
“well settled” rule “that exemptions from the antitrust laws are
to be narrowly construed.” Id. at 231. Furthermore, the Court
thought it significant “that the statutory language in question
here does not exempt the business of insurance companies from
the scope of the antitrust laws. The exemption is for the
‘business of insurance,’ not the ‘business of insurers’ . . . .” Id.
at 210–11. The mere fact that it is the conduct of insurance
companies that is challenged here is, therefore, not dispositive.

In Royal Drug itself, the plaintiffs, independent

123
pharmacies, challenged agreements between Blue Shield, a
health insurer, and three “participating pharmacies.” Under the
agreements, Blue Shield’s policyholders could purchase
prescription drugs from the participating pharmacies at a price
of $2 per prescription, and Blue Cross would reimburse the
pharmacy for the cost of acquiring the drug prescribed. By
contrast, policyholders who selected a non-participating
pharmacy were required to pay the full retail price charged by
the pharmacy and could then seek reimbursement from Blue
Shield for 75% of the difference between that price and $2. The
independent pharmacies asserted these agreements violated § 1
of the Sherman Act, while Blue Shield contended it was exempt
from the antitrust laws under the McCarran-Ferguson Act.

The Court held that the agreements did not constitute the
“business of insurance.” “The fallacy of the [defendants’]
position,” the Court explained, “is that they confuse the
obligations of Blue Shield under its insurance policies, which
insure against the risk that policyholders will be unable to pay
for prescription drugs during the period of coverage, and the
agreements between Blue Shield and the participating
pharmacies, which serve only to minimize the costs Blue Shield
incurs in fulfilling its underwriting obligations. . . . The
Pharmacy Agreements . . . do not involve any underwriting or
spreading of risk, but are merely arrangements for the purchase
of goods and services by Blue Shield.” Id. at 213–14. It is true,
the Court conceded, that these business arrangements with third-
party providers could affect Blue Shield’s costs, which could in

124
turn affect the premiums it charges. Id. at 214. But this
relationship to the “business of insurance” was too attenuated.
Defendants may have been able to demonstrate that the
Pharmacy Agreements lowered Blue Shield’s expenses, and that
these savings might be passed on to policyholders in the form of
lower premiums. “But, in that sense, every business decision
made by an insurance company has some impact on its
reliability, its ratemaking, and its status as a reliable insurer.”
Id. at 216–17. If the “business of insurance” were interpreted so
expansively, “almost every business decision of an insurance
company could be included in the [term]. Such a result would
be plainly contrary to the statutory language, which exempts the
‘business of insurance’ and not the ‘business of insurance
companies.’” Id. at 217. Another factor that influenced the
Court’s conclusion was that “[t]he Pharmacy Agreements are
not ‘between insurer and insured.’ They are separate contractual
arrangements between Blue Shield and pharmacies engaged in
the sale and distribution of goods and services other than
insurance.” Id. at 216.

Looking back on its decision in Royal Drug, the Court
later distilled three criteria for determining whether particular
conduct constitutes the “business of insurance”: “first, whether
the practice has the effect of transferring or spreading a
policyholder’s risk; second, whether the practice is an integral
part of the policy relationship between the insurer and the
insured; and third, whether the practice is limited to entities
within the insurance industry.” Union Labor Life Ins. Co. v.

125
Pireno, 458 U.S. 119, 129 (1982). The Court added, however,
that “[n]one of these criteria is necessarily determinative in
itself.” Id.

In Pireno, the challenged conduct was a health insurer’s
use of a professional association’s peer review committee to
examine chiropractors’ statements and charges and render an
opinion on the necessity of treatments and the reasonableness of
charges paid for them. The Court held that the use of the
association did not implicate the transfer of risk because “[p]eer
review takes place only after the risk has been transferred by
means of the policy, and then it functions only to determine . . .
whether the insured’s loss falls within the policy limits,” that is,
whether the insured’s loss is, under the terms of the policy,
among the risks that has been transferred to the insurer. Id. at
130 (internal quotation marks omitted). Furthermore, the Court
found that the insurer’s relationship with the peer review
committee “is not an integral part of the policy relationship
between insurer and insured” because “the challenged
arrangement . . . is obviously distinct from [the insurer’s]
contracts with its policyholders.” Id. at 131. “Finally, as
respects the third . . . criterion,” the Court concluded “it is plain
that the challenged peer review practices are not limited to
entities within the insurance industry.” Id. at 132. Accordingly,
the challenged practice was not the “business of insurance”
under the McCarran-Ferguson Act and so was not entitled to
exemption from federal antitrust law. Id. at 134.

The Supreme Court’s analysis in Royal Drug and Pireno

126
was informed by an extensive inquiry into the Act’s legislative
history. “The law was enacted in 1945 in response to [the
Supreme Court’s] decision in United States v. South-Eastern
Underwriters Assn., 322 U.S. 533” (1944). Royal Drug, 440
U.S. at 217. The defendants in South-Eastern Underwriters had
been charged with, inter alia, conspiring to fix insurance rates
and commissions, but the district court had dismissed the
indictment on the strength of a long line of jurisprudence
suggesting “that the insurance industry was not a part of
interstate commerce subject to [federal] regulation under the
Commerce Clause.” Royal Drug, 440 U.S. at 217; see, e.g.,
Paul v. Virginia, 75 U.S. (8 Wall.) 168, 183 (1868) (“Issuing a
policy of insurance is not a transaction of commerce.”). The
Supreme Court reversed, “holding that the business of insurance
is interstate commerce, and that the Congress which enacted the
Sherman Act had not intended to exempt the insurance industry
from its coverage.” Royal Drug, 440 U.S. at 217.

As the Court has explained, the primary purpose of the
McCarran-Ferguson Act was “to preserve state regulation of the
activities of insurance companies, as it existed before the South-
Eastern Underwriters case.” Id. at 218 n.18; see Stephens v.
Nat’l Distillers & Chem. Corp., 69 F.3d 1226, 1231 n.5 (2d Cir.
1995) (“[S]tate insurance legislation is exempt from the
restrictions of the Dormant Commerce Clause as a result of the
McCarran-Ferguson Act.”). This purpose is embodied in the
first clause of section 2(b). “[T]he applicability of the antitrust
laws to the insurance industry” was only “[a] secondary

127
concern.” Royal Drug, 440 U.S. at 218. Although the House of
Representatives initially approved a bill that would have entirely
exempted the insurance industry from federal antitrust laws, the
Act as passed created only a “partial exemption from those laws.
Perhaps more significantly, however, [the Act] embod[ies] a
legislative rejection of the concept that the insurance industry is
outside the scope of the antitrust laws—a concept that had
prevailed before the South-Eastern Underwriters decision.” Id.
at 220.

The Court also found that the legislative history sheds
some light on the scope of that exemption—that is, on which
particular activities within the insurance industry Congress
intended to exempt. This history, the Court concluded,
“strongly suggest[s] that Congress understood the business of
insurance to be the underwriting and spreading of risk.” Id. at
221. More specifically, the “primary concern” of Congress
“was that cooperative ratemaking efforts be exempt from the
antitrust laws” because such concerted efforts were understood
to be necessary to ensure the adequate capitalization of
insurance companies. Id. The Court quoted a report by the
National Association of Insurance Commissioners (NAIC):55
“[I]t would be a mistake to permit or require the unrestricted
competition contemplated by the antitrust laws to apply to the

55
“The views of the NAIC are particularly significant,
because the Act ultimately passed was based in large part on the
NAIC[’s proposed] bill.” Id.

128
insurance business. To prohibit combined efforts for statistical
and rate-making purposes would be a backward step in the
development of a progressive business.” Id. at 221–22 (quoting
90 Cong. Rec. A4405 (1944)) (emphasis omitted). During the
floor debates, Senator Ferguson explained the purpose of the bill
in a similar manner:

This bill would permit—and I think it is fair to
say that it is intended to permit—rating bureaus,
because in the last session we passed a bill for the
District of Columbia allowing rating. What we
saw as wrong was the fixing of rates without
statutory authority in the States; but we believe
that State rights should permit a State to say that
it believes in a rating bureau. I think the
insurance companies have convinced many
members of the legislature that we cannot have
open competition in fixing rates on insurance. If
we do, we shall have chaos. There will be
failures, and failures always follow losses.

Id. at 223 (quoting 91 Cong. Rec. 1481 (1945)); see also id.
(noting that “[t]he consistent theme of the remarks of other
Senators also indicated a primary concern that cooperative
ratemaking would be protected from the antitrust laws”). The
Court also found instructive President Roosevelt’s statement
when signing the bill: “Congress did not intend to permit private
rate fixing, which the Antitrust Act forbids, but was willing to
permit actual regulation of rates by affirmative action of the

129
States.” Id. at 224 (quoting [1944–45 Volume] The Public
Papers and Addresses of Franklin D. Roosevelt 587 (Samuel I.
Rosenman ed., 1950)).

On the basis of this history, one might narrowly construe
the “business of insurance” to encompass only public
ratemaking efforts, not purely private collaboration unauthorized
or unsupervised by state agencies. Dicta in Royal Drug suggest
otherwise, however. The Court observed that the Act’s
legislative history does not indicate exactly “which of the
various practices alleged in the South-Eastern Underwriters
indictment Congress intended to be covered by the phrase
‘business of insurance’”; nonetheless, it noted that the
indictment had charged “that the defendants had fixed their . . .
premium rates,” and it concluded that the legislative history did
make clear “that the fixing of rates is the ‘business of
insurance.’” Id. at 224 n.32. Since the South-Eastern
Underwriters defendants appear to have been charged with
private rate-fixing, see South-Eastern Underwriters, 322 U.S. at
535–36, the implication is that such activity falls within the
scope of the Act’s antitrust exemption.

Relying in part on this reasoning, the United States Court
of Appeals for the Eighth Circuit has explicitly rejected the
claim that private agreements among insurance companies to fix
rates do not fall within the “business of insurance.” In re
Workers’ Comp. Ins. Antitrust Litig., 867 F.2d 1552, 1555–57
(8th Cir. 1989). The court agreed with the defendants that if
joint rate setting is the business of insurance when authorized by

130
the state, “it makes little sense to say that cooperative rate
setting, without state involvement, is not within the business of
insurance. It is the setting of the rates which constitutes the
business of insurance. This characterization is not dependent
upon the identity of the rate setters.” Id. at 1556 n.7; cf. Proctor
v. State Farm Mut. Auto. Ins. Co., 675 F.2d 308, 318–25 (D.C.
Cir. 1982) (finding that an alleged horizontal conspiracy by five
automobile insurance companies to fix the price of automobile
body damage repair work was the “business of insurance” for
purposes of the McCarran-Ferguson Act’s antitrust
exemption).56

Our Court has also had occasion to interpret the scope of
the “business of insurance.” In Owens v. Aetna Life & Casualty
Co., decided after Royal Drug but before Pireno, we held that
alleged cooperation between two insurers “in the decision to file
in New Jersey only a single mass market rating-schedule, and
perhaps a very high individual policy rate . . . would fall within

56
Furthermore, as commentators have observed, a
construction of the “business of insurance” that limited the
concept to state-authorized collaboration would arguably “be so
narrow as not to go beyond the state action antitrust exemption”
set forth in Parker v. Brown, 317 U.S. 341, 350–52 (1943), thus
rendering the Act’s antitrust exemption superfluous. Jonathan
R. Macey & Geoffrey P. Miller, The McCarran-Ferguson Act
of 1945: Reconceiving the Federal Role in Insurance
Regulation, 68 N.Y.U. L. Rev. 13, 28 n.58 (1993).

131
even the narrowest reading” of the “business of insurance” for
purposes of the Act’s antitrust exemption. 654 F.2d 218, 232
(3d Cir. 1981). Analyzing Supreme Court precedent, we stated
that “[t]he earmark of insurance is the underwriting and
spreading of risks in exchange for a premium.” Id. at 224; see
Royal Drug, 440 U.S. at 211 n.7. At the same time, however,
we noted that the “business of insurance” “encompasses . . .
more than making contracts between an insurer and an insured.”
Owens, 654 F.3d at 224. Specifically, we found

it is clear that at least the following activities are
the business of insurance, either because they
pertain to risk-spreading or to the contract
between the insurer and the insured:
1. preparing and filing a rating-schedule, either on
behalf of an individual company or jointly
through a rating bureau;
2. deciding upon rating classification differences
between individual policies and group marketing
plans, either individually or jointly through a
rating bureau;
3. authorizing agents to solicit individual or group
policies;
4. accepting or rejecting coverages tendered by
brokers.

Id. at 225–26 (footnote omitted).

The dissenting opinion in Owens did not dispute the

132
majority’s conclusions about the scope of the “business of
insurance.” Instead, it argued that the majority had
mischaracterized the alleged activity before it. The dissent
believed the proper McCarran-Ferguson Act question concerned
not ratemaking, as the majority had concluded, but rather
“whether a conspiracy by insurance companies to divide markets
can be construed as a matter of law to constitute ‘the business of
insurance’ within the meaning of the McCarran-Ferguson Act.”
Id. at 236–37 (Sloviter, J., dissenting). In the dissent’s view, the
Act “was enacted to protect the arrangements necessary to
preserve the writing of insurance within and under regulation of
the respective states. . . . [T]he scope of the statute can be no
broader than protection of insurance company activities that can
rationally be claimed to need anticompetitive regulation.” Id. at
242 (internal quotation marks omitted). Accordingly, the dissent
believed it was “unlikely that Congress thought it was protecting
agreements whereby an insurance company would completely
withdraw from writing one type of insurance within the state.”
Id. Nonetheless, the dissent was “reluctant to suggest that no
agreement between insurance companies which may result in
withdrawal from a market can ever be the business of insurance,
because we do not know enough of the economic and business
stuff out of which these arrangements emerge to be certain.” Id.
at 244 (internal quotation marks omitted). What could be said
for certain, the dissent concluded, was that the District Court
had erred in finding “that the alleged division of markets
constitutes ‘the business of insurance’ as a matter of law.” Id.
at 245; see also Maryland v. Blue Cross & Blue Shield Ass’n.,

133
620 F. Supp. 907, 917 (D. Md. 1985) (“[I]n order to meet the
first Pireno requirement the defendants must show the
challenged territorial allocation is related positively to
underwriting and ratemaking; that is, that exclusive geographic
territories directly facilitate risk spreading and transfer through
the provision of insurance.” Because “[t]he parties have
submitted affidavits which raise material factual issues” as to
this question, summary judgment is inappropriate.).

With this precedent in mind, we turn to the case before
us. As the disagreement between the majority and dissent in
Owens illustrates, the precise characterization of the defendants’
conduct can be dispositive. Here, having dismissed several
antitrust claims for failure to satisfy Twombly’s pleading
standard, we are left with plaintiffs’ allegations that Marsh’s
insurer-partners agreed with one another not to compete for
incumbent business. Applying the Pireno criteria to this alleged
conduct, we agree with defendants (as did the District Court)
that the third criterion is met because the parties to this alleged
agreement are all entities within the insurance industry. See
Pireno, 458 U.S. at 129 (asking “whether the practice is limited
to entities within the insurance industry”). There is also a strong
argument that the agreement would be “an integral part of the
policy relationship between the insurer and the insured,” id.
(describing the second criterion), insofar as it would affect the
insurers from which a prospective purchaser could obtain
coverage.

On the basis of the complaint before us, however, we

134
cannot conclude that the alleged agreement “has the effect of
transferring or spreading a policyholder’s risk.” Id. (describing
the first criterion). Given the Supreme Court’s declaration that
“underwriting or spreading of risk [is] an indispensable
characteristic of insurance,” Royal Drug, 440 U.S. at 212, we
think the failure to satisfy this first criterion is decisive. See id.
at 220–21 (“References to the meaning of the ‘business of
insurance’ in the legislative history of the McCarran-Ferguson
Act strongly suggest that Congress understood the business of
insurance to be the underwriting and spreading of risk.”).

Our conclusion as to the first criterion rests on the fact
that plaintiffs do not allege that defendants’ agreement involved
who could receive insurance coverage, or the type of coverage
they could obtain. Cf., e.g., In re Ins. Antitrust Litig., 938 F.2d
919, 927 (9th Cir. 1991) (holding that an alleged conspiracy
among members of the insurance industry to restrict the terms of
coverage of commercial general liability insurance qualified as
the “business of insurance”), aff’d in part, rev’d in part on other
grounds sub nom. Hartford Fire Ins. Co. v. California, 509 U.S.
764 (1993). Plaintiffs allege only that defendants colluded in
order to influence with which of them a given policy could be
placed. In other words, the complaint asserts conduct affecting
not whether or to what extent a prospective insurance purchaser
would transfer its risk to an insurer, but merely to which insurer
that risk would be transferred. See Comm. SAC ¶ 100 (stating
that under the Marsh-centered commercial conspiracy, “the
incumbent [insurer] who hits a [premium] target and provides

135
the coverages requested is protected” (internal quotation marks
omitted) (emphasis added)); id. ¶ 106 (relating a statement by a
Marsh employee that “if the incumbent [insurers] meet their
target price and does [sic] the coverage we want, [Marsh Global
Broking] will protect them and make sure they get the business”
(internal quotation marks omitted) (emphasis added)). While
discovery may reveal facts warranting a reassessment, we cannot
say that defendants’ challenged agreement, as alleged in the
complaint, affected the spreading of risk within the meaning of
the “business of insurance.”

Royal Drug’s examination of the purpose of the
McCarran-Ferguson Act bolsters this conclusion. As the
Supreme Court explained, in carving out only the “business of
insurance” from federal antitrust regulation—and not the
“business of insurance companies”—Congress had in mind
“[t]he relationship between insurer and insured, the type of
policy which could be issued, its reliability, interpretation, and
enforcement,” as well as “other activities of insurance
companies [that] relate so closely to their status as reliable
insurers that they too must be placed in the same class.” Royal
Drug at 215–16 (quoting SEC v. Nat’l Sec., Inc., 393 U.S. 453,
460 (1969)). We think it noteworthy that in this passage the
Court twice referred to the reliability of insurers—the second
time in defining the open-ended class of unenumerated activities
that would fall within the “business of insurance.” Indeed, a
focus on reliability was at the heart of what the Court described
as “the primary concern of both representatives of the insurance

136
industry and the Congress,” namely “that cooperative
ratemaking efforts be exempt from the antitrust laws.” Id. at
221 (explaining “the widespread view” at the time of the
McCarran-Ferguson Act’s passage “that it is very difficult to
underwrite risks in an informed and responsible way without
intra-industry cooperation”).

Here, not only is defendants’ alleged agreement not to
compete for incumbent business different than the cooperative
ratemaking efforts described in Royal Drug, but it also appears
to have been unrelated to reliability; it does not involve any
restriction on the type of coverage offered or the risk profile of
insurable entities. Royal Drug emphasized that Congress
understood the “business of insurance” as bound up with
actuarial considerations intrinsic to the underwriting process, but
the collusion alleged by the complaint is not addressed to these
considerations. Defendants’ alleged agreement was designed to
ensure only that, once an insurer had won a client’s
business—by providing the client with an acceptable coverage
package at an acceptable premium—another insurer would not
attempt to poach that business by offering a more attractive price
when it came up for renewal.

Nor does this alleged conduct fall into any of the
categories of the “business of insurance” we set forth in Owens,
each of which also implicated reliability issues. See Owens, 654
F.2d at 225–26 (majority opinion). The alleged agreement does
not involve rating schedules or rating classification differences.
Nor does it involve authorizing agents to solicit particular types

137
of policies. The final category, “accepting or rejecting
coverages tendered by brokers,” id. at 226, may seem at first
blush to describe defendants’ alleged behavior, in which non-
incumbent insurers refused to compete for the incumbent’s
renewal business. But in the context of Owens, “accepting or
rejecting coverages tendered by brokers” is best understood as
referring to attempts by insurers to discriminate among brokers
based on the type of policy tendered. See id. at 233 (noting
plaintiff’s “claim that there was a conspiracy to drive him out of
business”). The majority did not address whether a bare
agreement to divide the market (either geographically or by
customer) would constitute the business of insurance. See id. at
234–37 (Sloviter, J., dissenting). Examining Royal Drug, we
believe a horizontal agreement not to compete for renewal
business, at least as alleged here, is not within the scope of
activity exempted by the McCarran-Ferguson Act. Such an
agreement is incidental to the processes of risk assessment and
underwriting deemed integral to the “business of insurance” by
the Supreme Court.

Defendants dispute this analysis. Noting that plaintiffs’
theory of injury rests on a claim that insurance customers paid
higher prices for insurance as a result of the alleged agreement,
defendants argue that the allegations go directly to the heart of
the insurance contract. In defendants’ view, to deny that the
alleged agreement among the insurers implicates the transfer of
risk is to artificially segregate the element of premium price
from the element of risk spreading; since premiums are the price

138
paid for transferring risk, conspiracies that have the direct effect
(if not necessarily the explicit purpose) of driving up premium
prices necessarily affect risk spreading.

We do not deny that premiums are an integral part of the
transfer of risk. But as the District Court accurately observed,
Royal Drug stands for the proposition that “more than a mere
impact on the price of premiums must be demonstrated” in order
“[t]o establish that a particular practice has a substantial
connection to the spreading and the underwriting of risk.” 2006
WL 2850607, at *9. Without question, if defendant insurers
agreed not to compete with one another for renewal business,
the premiums charged for renewal policies would presumably be
higher than in a competitive market. But we think there is an
important difference between an agreement not to compete to
sell a given package of insurance coverage and an agreement to
fix the rates for such coverage. As noted, exempting
cooperative ratemaking from federal antitrust regulation was one
of the key purposes of the McCarran-Ferguson Act because such
joint ratemaking was deemed necessary at times in order to
protect insurers from underestimating risk and the attendant
threat of insolvency.57 The kind of agreement alleged here,

57
Cf. Arroyo-Melecio v. Puerto Rican Am. Ins. Co., 398 F.3d
56 (1st Cir. 2005). The plaintiffs in that case alleged a
horizontal agreement among private insurers not to provide
compulsory automobile insurance, in violation of the Sherman
Act. According to the plaintiffs, the insurers had stakes in a

139
however, would not serve such a purpose. Rather than
guarantee that every insurer will offer a given coverage package
at the same prescribed rate, the alleged agreement here, taken

publicly created entity providing such insurance and believed
they would ultimately reap greater profits if they did not
compete with the price set by that entity (which was fixed by
Puerto Rican law, although private insurers were allowed to
charge a lower premium in order to compete to provide the
compulsory insurance). The plaintiffs also alleged that the
private insurers had colluded to coerce brokers to refrain from
selling compulsory insurance contracts through private
companies.
The Court of Appeals for the First Circuit decided that
the challenged conduct constituted the “business of insurance”
under the second clause of section 2(b) of the McCarran-
Ferguson Act. The court determined that “[h]orizontal
agreements among insurers to fix the price and to issue policies
only through the residual market are within the business of
insurance.” Id. at 67. Unlike defendants’ alleged agreement
here, the insurers’ conduct in Arroyo-Melecio was rate fixing in
the classic sense: because of the public scheme of compulsory
insurance, it had the effect of ensuring that no one could
purchase insurance except at the price prescribed by the public
entity. The compulsory insurance scheme and public entity in
Arroyo-Melecio also brought that case much closer to the
McCarran-Ferguson Act’s concern to exempt public ratemaking
from federal antitrust regulation.

140
alone, would leave it to each incumbent insurer to assess its risk
of loss for itself, and then to determine its own profit margin by
setting the price/premium without fear of competition. In short,
there is nothing about the alleged agreement that is particular to
the business of insurance; it is simply an agreement not to
compete to sell a particular product to a particular customer,
which would be expected—in any industry, see, e.g., Petruzzi’s
IGA Supermarkets, Inc. v. Darling-Delaware Co., 998 F.2d
1224 (3d Cir. 1993) (examining alleged agreement among fat
and bone rendering companies not to bid on one another’s
existing accounts)—to yield a higher price than would prevail in
a competitive market. The mere fact that the product here
happens to be insurance is not enough to trigger the McCarran-
Ferguson Act’s exemptions. Cf. Royal Drug, 440 U.S. at
213–214 n.9 (“[I]t does not follow that because an agreement is
necessary to provide insurance, it is also the ‘business of
insurance.’”).

Defendants do not dispute that Royal Drug held that the
fact that challenged behavior has an impact on premiums is not
enough to make it the “business of insurance” for purposes of
the McCarran-Ferguson Act’s antitrust exemption. Defendants
point out, however, that while the agreement found not to be the
“business of insurance” in Royal Drug was between insurers and
third-party benefit providers, here it is between only insurers and
directly involves the formation of the insurance contract
between insurer and insured (and not merely, as in Royal Drug,
the specific manner in which the insurer would perform

141
obligations assumed under a preexisting contract). In Sabo v.
Metropolitan Life Insurance Co., we stated that “whatever the
precise contours of the insurance business phrase may be, there
is nothing more basically ‘insurance’ than the sale of an
insurance contract.” 137 F.3d 185, 191 (3d Cir. 1998).
Applying this principle, we found that the McCarran-Ferguson
Act precluded a RICO claim alleging that the defendant
insurance company had, inter alia, engaged in “a ‘churning’
scheme, whereby [the defendant] encouraged and coerced
agents to fraudulently trade insurance policies in order to
accumulate commissions and decrease the value of outstanding
policies.” Id. at 187. According to defendants, “Sabo could not
more clearly dispose of the ‘business of insurance’ question in
this case,” as “[t]he ‘sale of an insurance contract’ is at the core
of Plaintiffs’ antitrust claims here.” Defendants’ EB Br. 67.

Defendants overlook, however, an important distinction
between Sabo and this case. Because Sabo involved a RICO
rather than an antitrust claim, it was governed by the first clause
of § 2(b) of the McCarran-Ferguson Act. That clause provides
that “[n]o Act of Congress shall be construed to invalidate,
impair, or supersede any law enacted by any State for the
purpose of regulating the business of insurance . . . unless such
Act specifically relates to the business of insurance.” 15 U.S.C.
§ 1012(b). This “first clause . . . impos[es] what is, in effect, a
clear-statement rule, a rule that state laws enacted ‘for the
purpose of regulating the business of insurance’ do not yield to
conflicting federal statutes unless a federal statute specifically

142
requires otherwise.” U.S. Dep’t of Treasury v. Fabe, 508 U.S.
491, 507 (1993). Both clauses incorporate the phrase “business
of insurance,” but as the Supreme Court has emphasized, the
respective protections afforded to state law under the two
clauses are of different scopes. “The first clause commits laws
‘enacted . . . for the purpose of regulating the business of
insurance’ to the States, while the second clause exempts only
‘the business of insurance’ itself from the antitrust laws.” Id. at
504. Because “[t]he broad category of laws enacted ‘for the
purpose of regulating the business of insurance’ . . . necessarily
encompasses more than just the business of insurance,” id. at
505, judicial determinations made when applying one clause
may not be dispositive when applying the other. As Sabo itself
explained, “Fabe makes clear [that] the Royal Drug test is only
a starting point in the analysis for non-antitrust cases.” 137 F.3d
at 191 n.3; see also Jonathan R. Macey & Geoffrey P. Miller,
The McCarran-Ferguson Act of 1945: Reconceiving the Federal
Role in Insurance Regulation, 68 N.Y.U. L. Rev. 13, 22 (1993)
(“[I]t appears that the meaning of [the ‘business of insurance’]
varies depending upon whether the case involves antitrust [i.e.,
clause two] or other regulatory [i.e., clause one] matters.”).
Accordingly, we cannot reflexively transplant Sabo’s holding
into our analysis under the second clause of § 2(b).

In fact, a close reading of the Supreme Court’s cases may
counsel against indiscriminately classifying all conduct
involving the “sale of an insurance contract” as the “business of
insurance” for purposes of the Act’s antitrust exemption. Sabo

143
relied heavily on SEC v. National Securities, Inc., in particular
its statement that “[t]he selling and advertising of policies” falls
“within the scope of the [Act’s preemption of federal law].” 393
U.S. at 460; see Sabo, 137 F.3d at 190–92. National Securities,
like Sabo, was a case applying the first clause of § 2(b). When
the Court later attempted in Royal Drug to define the “business
of insurance” for purposes of clause two, it quoted from the
same paragraph of National Securities on which Sabo drew. See
Royal Drug, 440 U.S. at 215–16. Notably, however, the Royal
Drug Court omitted the passage about “the selling and
advertising of policies” and included only the second part of the
paragraph, which, as noted, focused on the issue of “reliability.”
Id. In light of Fabe, we interpret this to mean that although any
state law that regulates “the selling and advertising of
insurance” will qualify as a “law enacted by [a] State for the
purpose of regulating the business of insurance” under clause
one of the McCarran-Ferguson Act, “the selling and advertising
of insurance” is not the “business of insurance” under clause
two unless it has some effect on “reliability” or underwriting
issues.58 As noted, based on the face of the complaint, we
cannot say that defendants’ alleged behavior satisfies this

58
Cf. Blue Cross, 620 F. Supp. at 917 (qualifying the
proposition “that the selling and advertising of policies is part of
the business of insurance” by noting that the Supreme Court
“decided the National Securities case many years before Royal
Drug and Pireno, cases purporting to narrow the ‘business of
insurance’ definition”).

144
standard.

Furthermore, even if Sabo’s holding were directly
applicable to this case, plaintiffs’ allegations here are
distinguishable. Whereas the Sabo plaintiff challenged a
scheme by which defendants allegedly churned insurance
policies—that is, bought and sold them with excessive
frequency—plaintiffs here complain that defendants agreed not
to sell them insurance. Cf. Owens, 654 F.2d at 242 (Sloviter, J.,
dissenting) (“It appears . . . unlikely that Congress thought it was
protecting [from antitrust regulation] agreements whereby an
insurance company would completely withdraw from writing
one type of insurance within the state. Aetna’s argument [to the
contrary] seems to turn protection of the ‘business of insurance’
into the ‘business of non-insurance.’”). And unlike the
agreement alleged here, the alleged churning of policies in Sabo
arguably did affect the coverage terms of the policies
themselves, as they “decrease[d] the value of outstanding [life
insurance] policies.” Sabo, 137 F.3d at 187.

In sum, although the scope of the agreement alleged by
plaintiffs has been refined since the District Court first passed
on the issue of the McCarran-Ferguson Act, we agree with the
court’s conclusion that defendants’ alleged conduct does not
constitute the “business of insurance” for purposes of the Act’s
antitrust exemption. Accordingly, the Act does not provide a
basis for dismissing plaintiffs’ Sherman Act claims.

3. Antitrust Conclusion

145
 Because the McCarran-Ferguson Act does not bar
plaintiffs’ claims (at least, not at this stage of the litigation), our
earlier Twombly analysis is dispositive. Given the long path our
discussion has taken, a brief synopsis of that analysis is in order.
The Supreme Court has made clear that courts confronted with
a motion to dismiss must assess whether the complaint contains
“enough factual matter (taken as true) to suggest that an
agreement was made.” Twombly, 550 U.S. at 556.
“Determining whether a complaint states a plausible claim to
relief will . . . be a context-specific task that requires the
reviewing court to draw on its judicial experience and common
sense.” Iqbal, 129 S. Ct. at 1950; see also Arar v. Ashcroft, 585
F.3d 559, 617 (2d Cir. 2009) (Parker, J., dissenting)
(“Plausibility . . . depends on a host of considerations: The full
factual picture presented by the complaint, the particular cause
of action and its elements, and the available alternative
explanations [for the facts alleged].”), cert. denied, 2010 WL
390379 (U.S. June 14, 2010). Some claims will demand
relatively more factual detail to satisfy this standard, while
others require less.

In the context of claims brought under § 1 of the Sherman
Act, plausibility is evaluated with reference to well-settled
antitrust jurisprudence that “limits the range of permissible
inferences from ambiguous evidence.” Matsushita, 475 U.S. at
588. In particular, “when allegations of parallel conduct are set
out in order to make a § 1 claim,” that conduct must be placed
in “some setting suggesting the agreement necessary to make out

146
a § 1 claim.” Twombly, 550 U.S. at 557. In other words, the
complaint must allege some “further circumstance,” “something
more than merely parallel behavior,” “pointing toward a meeting
of the minds.” Id. at 557, 560. If, in the circumstances alleged,
the asserted “parallel conduct . . . could just as well be
independent action,” then the complaint has failed to plead a §
1 claim. Id. at 557.

Here, the bid-rigging allegations supply the requisite
“further circumstance.” Because they plausibly suggest an
unlawful horizontal conspiracy not to compete for incumbent
business, plaintiffs have adequately met Rule 8(a)(2)’s
requirement for setting forth a § 1 claim against those
defendants in the asserted Marsh-centered commercial
conspiracy who are alleged to have participated in bid rigging.
This agreement to divide the market, if proven, would be a
naked restraint of trade subject to per se condemnation. See
Leegin, 551 U.S. at 886; In re Japanese Elec. Prods. Antitrust
Litig., 723 F.2d 238, 310–11 (3d Cir. 1983) (observing that “a
horizontal agreement to allocate customers” is “ordinarily . . . a
per se violation” (citing United States v. Topco Assocs., 405
U.S. 596, 606–12 (1972))), rev’d on other grounds sub nom.
Matsushita, 475 U.S. 574 (1986).59

59
Even if we concluded that a per se rule were inappropriate
and instead applied a quick look analysis (as plaintiffs urge in
the alternative), we would still condemn the alleged restraint.
As noted, defendants have not put forward a plausible

147
 With respect to the remaining antitrust claims, however,
plaintiffs have failed to plead facts plausibly supporting their
allegations of horizontal conspiracies to unreasonably restrain
trade, notwithstanding their conclusory assertions of agreement.
Given plaintiffs’ exclusive reliance on a per se or quick look
analysis, the absence of a horizontal agreement is fatal to their
§ 1 claims.60 Accordingly, these antitrust claims must be
dismissed, as the District Court concluded.

B. RICO Claims

Plaintiffs also claim that defendants’ alleged conduct
violated the Racketeer Influenced and Correct Organizations
(RICO) Act, 18 U.S.C. § 1962(c), (d). Section 1962(c) makes
it unlawful “for any person employed by or associated with any
enterprise engaged in, or the activities of which affect, interstate
or foreign commerce, to conduct or participate, directly or
indirectly, in the conduct of such enterprise’s affairs through a
pattern of racketeering activity.” Section 1962(d) makes it
unlawful “for any person to conspire to violate” § 1962(c).

justification for a naked horizontal agreement not to compete for
renewal customers.
60
As noted, see supra Section II.A.1.a., plaintiffs who seek to
condemn a vertical agreement must proceed under the traditional
rule of reason, which requires plaintiffs to demonstrate
anticompetitive effects in the relevant market.

148
 To plead a RICO claim under § 1962(c), “the plaintiff
must allege (1) conduct (2) of an enterprise (3) through a pattern
(4) of racketeering activity.” Lum v. Bank of Am., 361 F.3d 217,
223 (3d Cir. 2004) (citing Sedima, S.P.R.L. v. Imrex Co., 473
U.S. 479, 496 (1985)). According to the statute, an “enterprise”
includes “any individual, partnership, corporation, association,
or other legal entity, and any union or group of individuals
associated in fact although not a legal entity.” 18 U.S.C. §
1961(4). In the Commercial Case, plaintiffs allege the existence
of one legal-entity enterprise, the Council of Insurance Agents
and Brokers (CIAB), and six association-in-fact enterprises
corresponding to the six broker-centered antitrust conspiracies
discussed above. In the Employee Benefits Case, plaintiffs
plead five association-in-fact enterprises corresponding to the
broker-centered conspiracies alleged in their antitrust claims.

According to the RICO statute, a “pattern of racketeering
activity” requires at least two acts of racketeering activity within
a ten-year period. 18 U.S.C. § 1961(5). “These predicate acts
of racketeering may include, inter alia, federal mail fraud under
18 U.S.C. § 1341 or federal wire fraud under 18 U.S.C. § 1343.”
Lum, 361 F.3d at 223; see 18 U.S.C. § 1961(1) (defining
“racketeering activity”). Plaintiffs assert that the defendant
brokers in both the Commercial Case and Employee Benefits
Case committed “numerous acts of mail and wire fraud” in
furtherance of the enterprises to which they allegedly belonged.
Comm. SAC ¶ 510; accord EB SAC ¶ 520. More specifically,
plaintiffs allege that defendants “knowingly and intentionally

149
made misrepresentations” in materials disseminated by mail and
wire, “wherein they routinely represented that they would act in
the best interests of their clients in providing unbiased advice
and assistance in the selection of insurance products and
services relating thereto and that they would act as fiduciaries of
their clients in placing insurance on the best terms possible and
at the best price available.” Comm. SAC ¶ 535; EB SAC ¶ 539.
At the same time, plaintiffs charge, defendants “knowingly and
intentionally . . . concealed material facts,” such as “the steering
of insurance placements from the Broker Defendants to the
Insurer Defendants,” and the fact “that the Broker Defendants
were not acting in the best interest of their clients but were
instead acting on behalf of themselves and the Insurer
Defendants who were associated with the Broker’s enterprise to
further their financial interests at the expense of their clients.”
Comm. SAC ¶ 535; EB SAC ¶¶ 539–40. In the Employee
Benefits case, plaintiffs also allege that defendants violated 18
U.S.C. § 1954;61 see 18 U.S.C. § 1961(1) (including violations

61
Section 1954 provides:
Whoever being—
(1) an administrator, officer, trustee,
custodian, counsel, agent, or employee of
any employee welfare benefit plan or
employee pension benefit plan; or
(2) an officer, counsel, agent, or employee
of an employer or an employer any of
whose employees are covered by such

150
 plan; or
(3) an officer, counsel, agent, or employee
of an employee organization any of whose
members are covered by such plan; or
(4) a person who, or an officer, counsel,
agent, or employee of an organization
which, provides benefit plan services to
such plan
receives or agrees to receive or solicits any fee,
kickback, commission, gift, loan, money, or thing
of value because of or with intent to be influenced
with respect to, any of the actions, decisions, or
other duties relating to any question or matter
concerning such plan or any person who directly
or indirectly gives or offers, or promises to give or
offer, any fee, kickback, commission, gift, loan,
money, or thing of value prohibited by this
section, shall be fined under this title or
imprisoned not more than three years, or both:
Provided, That this section shall not prohibit the
payment to or acceptance by any person of bona
fide salary, compensation, or other payments
made for goods or facilities actually furnished or
for services actually performed in the regular
course of his duties as such person, administrator,
officer, trustee, custodian, counsel, agent, or

151
of 18 U.S.C. § 1954 among the enumerated racketeering
activities). Plaintiffs assert that “[e]ach payment of a
Contingent Commission . . . is a violation of Section 1954 and
. . . is intended to influence the advice that the Defendant
Brokers give to the plan sponsors, plan administrators and/or
plan participants.” EB SAC ¶ 535.

In their motion to dismiss the RICO claims, defendants
argued that plaintiffs had failed adequately to plead the
enterprise and conduct elements of their § 1962(c) claims, and
that they had failed adequately to plead predicate acts of
racketeering. The District Court granted the motion, finding that
plaintiffs had insufficiently pled both the enterprise and conduct
elements of the § 1962(c) claims based on the alleged broker-
centered enterprises, and had insufficiently pled that defendants
conducted CIAB “through a pattern of racketeering activity.”
Having determined that plaintiffs had failed to allege adequately
that any defendant had violated § 1962(c), the court also
dismissed the claims of conspiracy under § 1962(d). 2007 WL
2892700, at *33; see id. at *13 (holding that “in the event all
substantive RICO claims in the action are dismissed, a plaintiff
cannot bring a § 1962(d) claim”).

1. Legal Standards

employee of such plan, employer, employee
organization, or organization providing benefit
plan services to such plan.

152
 a. Section 1962(c)

i. The Enterprise Element

The RICO statute “describes two categories of
associations that come within the purview of the ‘enterprise’
definition. The first encompasses organizations such as
corporations and partnerships, and other ‘legal entities.’ The
second covers ‘any union or group of individuals associated in
fact although not a legal entity.’” United States v. Turkette, 452
U.S. 576, 581–82 (1981) (quoting 18 U.S.C. § 1961(4)). As the
District Court here explained, when the enterprise asserted is a
legal identity, such as “a legitimate business or organization . . .,
the need to allege and prove the existence of enterprise structure
can be met without great difficulty, since all aspects of the
enterprise element . . . are satisfied by the mere proof that the
entity does in fact have a legal existence.” 2007 WL 2892700,
at *9; see, e.g., Webster v. Omnitrition Int’l, Inc., 79 F.3d 776,
786 (9th Cir. 1996) (“[C]orporate entities ha[ve] a legal
existence . . ., and the very existence of a corporation meets the
requirement for a separate [enterprise] structure.” (internal
quotation marks omitted) (alteration in original)); see also Boyle
v. United States, 129 S. Ct. 2237, 2249 (2009) (Stevens, J.,
dissenting) (“In cases involving a legal entity, the matter of
proving the enterprise element is straightforward . . . .”).

The statutory language does not, however, specify the
essential features of an association-in-fact enterprise. The
Supreme Court attempted to explicate this concept in Turkette,

153
where it reviewed a First Circuit decision limiting the definition
of “enterprise” to legitimate organizations. The Supreme Court
reversed, stating that “[t]here is no restriction upon the
associations embraced by [§ 1961(4)’s] definition [of
enterprise]: an enterprise includes any union or group of
individuals associated in fact. On its face, the definition appears
to include both legitimate and illegitimate enterprises within its
scope . . . .” Turkette, 452 U.S. at 580.

The First Circuit had expressed concern that including
criminal organizations within the definition of “enterprise”
would effectively collapse the distinction between the
“enterprise” and “pattern of racketeering” elements of a §
1962(c) violation. Id. at 582. The Supreme Court agreed that
the “enterprise” and the “pattern of racketeering activity” were
distinct elements of a § 1962(c) claim. But the Court rejected
the argument that preserving this distinction required the
exclusion of illegitimate organizations from the definition of
“enterprise.” Id. at 583. In setting forth its own understanding
of this distinction, the Court suggested several features defining
an association-in-fact enterprise:

The enterprise is an entity, for present purposes a
group of persons associated together for a
common purpose of engaging in a course of
conduct. The pattern of racketeering activity is,
on the other hand, a series of criminal acts as
defined by the statute. The former is proved by
evidence of an ongoing organization, formal or

154
 informal, and by evidence that the various
associates function as a continuing unit. The
latter is proved by evidence of the requisite
number of acts of racketeering committed by the
participants in the enterprise. While the proof
used to establish these separate elements may in
particular cases coalesce, proof of one does not
necessarily establish the other. The “enterprise”
is not the “pattern of racketeering activity”; it is
an entity separate and apart from the pattern of
activity in which it engages.

Id. (internal citation omitted).

Interpreting this language from Turkette, we identified
three elements essential to an association-in-fact enterprise.
United States v. Riccobene, 709 F.2d 214, 221–24 (3d Cir.
1983). We stated that, first, such an enterprise must have “some
sort of structure . . . within the group for the making of
decisions, whether it be hierarchical or consensual. There must
be some mechanism for controlling and directing the affairs of
the group on an on-going, rather than ad hoc, basis.” Id. at 222.
Second, “the various associates [must] function as a continuing
unit. This does not mean that individuals cannot leave the group
or that new members cannot join at a later time. It does require,
however, that each person perform a role in the group consistent
with the organizational structure established by the first element
and which furthers the activities of the organization.” Id. at 223
(internal quotation marks and citation omitted). Finally, we

155
reiterated Turkette’s requirement that the association-in-fact be
“an entity separate and apart from the pattern of activity in
which it engages.” Id. (quoting Turkette, 452 U.S. at 583). As
we understood this last requirement,

it is not necessary to show that the enterprise has
some function wholly unrelated to the
racketeering activity, but rather that it has an
existence beyond that which is necessary merely
to commit each of the acts charged as predicate
racketeering offenses. The function of overseeing
and coordinating the commission of several
different predicate offenses and other activities on
an on-going basis is adequate to satisfy the
separate existence requirement.

Id. at 223–24.

In evaluating the sufficiency of plaintiffs’ pleadings here,
the District Court understandably relied heavily on Riccobene.
See 2007 WL 2892700, at *9–11. After the District Court had
dismissed plaintiffs’ claims, and after we had heard argument in
this appeal, the Supreme Court decided Boyle v. United States,
129 S. Ct. 2237 (2009). Boyle sought to clarify the required
attributes of an association-in-fact enterprise in order to resolve
conflicts that had developed among the courts of appeals over
the proper interpretation of the Turkette factors. Id. at 2243.
Rejecting several proposed ways of cabining the definition of an
“enterprise,” the Boyle Court highlighted several elements of the

156
RICO statute that pointed toward a capacious construction of the
term. Most significant was the statute’s specific description of
possible enterprises. See 18 U.S.C. § 1961(4) (stating that an
“‘enterprise’ includes any individual, partnership, corporation,
association, or other legal entity, and any union or group of
individuals associated in fact although not a legal entity”).
“This enumeration of included enterprises is obviously broad,
encompassing ‘any . . . group of individuals associated in fact.’
The term ‘any’ ensures that the definition has a wide reach, and
the very concept of an association in fact is expansive.” Boyle,
129 S. Ct. at 2243 (quoting 18 U.S.C. § 1961(4)) (emphasis in
Boyle) (internal citation omitted).62 In addition, “the RICO
statute provides that its terms are to be ‘liberally construed to
effectuate its remedial purposes.’” Id. (quoting Organized Crime
Control Act of 1970, Pub. L. No. 91-452, § 904(a), 84 Stat. 922,
947).

Informed by these background principles, the Court
expounded the necessary elements of an association-in-fact
enterprise. Such an enterprise must have a structure.

62
The Court noted that § 1961(4), which lists entities
“include[d]” in the term “enterprise,” “does not purport to set
out an exhaustive definition of the term,” and that,
“[a]ccordingly, this provision does not foreclose the possibility
that the term might include, in addition to the specifically
enumerated entities, others that fall within the ordinary meaning
of the term ‘enterprise.’” Boyle, 129 S. Ct. at 2243 n.2.

157
Specifically, it “must have at least three structural features: a
purpose, relationships among those associated with the
enterprise, and longevity sufficient to permit these associates to
pursue the enterprise’s purpose.” Id. at 2244.63 But the Court
saw “no basis in the language of RICO” for requiring a
particular type of organizational structure. Id. at 2245. An
association-in-fact enterprise, it explained,

need not have a hierarchical structure or a “chain
of command”; decisions may be made on an ad
hoc basis and by any number of methods—by
majority vote, consensus, a show of strength, etc.
Members of the group need not have fixed roles;

63
Boyle thus sees what we described in Riccobene as the
second element, i.e., continuity, as an inherent component of the
structure requirement. See Boyle, 129 S. Ct. at 2244 (“Section
1962(c) . . . shows that an ‘enterprise’ must have some
longevity, since the offense proscribed by that provision
demands proof that the enterprise had ‘affairs’ of sufficient
duration to permit an associate to ‘participate’ in those affairs
through ‘a pattern of racketeering activity.’”). In other words,
while Riccobene used the term “structure” to describe one of
several necessary features of an “enterprise,” Boyle appears to
use “structure” as an overarching term encompassing all of the
requisite elements: common purpose, relationships among those
associated with the enterprise, and the continuity necessary to
allow the associates to pursue the enterprise’s purpose.

158
 different members may perform different roles at
different times. The group need not have a name,
regular meetings, dues, established rules and
regulations, disciplinary procedures, or induction
or initiation ceremonies. While the group must
function as a continuing unit and remain in
existence long enough to pursue a course of
conduct, nothing in RICO exempts an enterprise
whose associates engage in spurts of activity
punctuated by periods of quiescence. Nor is the
statute limited to groups whose crimes are
sophisticated, diverse, complex, or unique; for
example, a group that does nothing but engage in
extortion through old-fashioned, unsophisticated,
and brutal means may fall squarely within the
statute’s reach.

Id. at 2245–46; see also id. at 2243 & n.3 (rejecting, as
“extratextual,” the dissent’s argument “that the definition of a
RICO enterprise is limited to ‘business-like entities’” (citing id.
at 2247–50 (Stevens, J., dissenting)). Boyle makes clear, in
other words, that although the structure requirement demands
that “the parts” of the association in fact must be “arranged or
put together to form a whole,” the statute does not prescribe any
particular arrangement, as long as it is “sufficient to permit [the
enterprise’s] associates to pursue the enterprise’s purpose.” Id.
at 2244 (internal quotation marks omitted).

Boyle also clarified the relationship between the

159
“enterprise” and “pattern of racketeering activity” elements of
a § 1962(c) claim. The petitioner in Boyle had objected to the
trial judge’s jury instructions, which had stated that “the
existence of an association-in-fact [enterprise] is sometimes
more readily proven by what it does, rather than by abstract
analysis of its structure.” Id. at 2247. In the petitioner’s view,
the judge should have specified that, to qualify as a RICO
enterprise, the association’s structure must go “beyond that
inherent in the pattern of racketeering activity.” Id. at 2244.
The Supreme Court found the petitioner’s proffered language
unnecessary. If the language “is interpreted to mean that the
existence of an enterprise is a separate element that must be
proved,” the Court explained, “it is of course correct. . . . [T]he
existence of an enterprise is an element distinct from the pattern
of racketeering activity and ‘proof of one does not necessarily
establish the other.’” Id. at 2245 (quoting Turkette, 452 U.S. at
583). Indeed, the Court thought it was “easy to envision
situations in which proof that individuals engaged in a pattern
of racketeering activity would not establish the existence of an
enterprise.” Id. at 2245 n.4. If, for example, “several
individuals, independently and without coordination, engaged
in a pattern of crimes listed as RICO predicates[,] . . . [p]roof of
these patterns would not be enough to show that the individuals
were members of an enterprise.” Id. (emphasis added). Nor
would proof of a conspiracy to commit a RICO predicate
offense “necessarily establish that the defendant[s] participated
in the affairs of an . . . enterprise through a pattern of . . .
crimes.” Id. at 2246. While “a conspiracy is an inchoate crime

160
that may be completed in the brief period needed for the
formation of the agreement and the commission of a single overt
act in furtherance of the conspiracy,” § 1962(c) “demands much
more: the creation of an ‘enterprise’—a group with a common
purpose and course of conduct—and the actual commission of
a pattern of predicate offenses.” Id. The Court did not believe,
however, that the petitioner’s suggested language was necessary
in order to avoid a merger of the crime proscribed by § 1962(c)
and simple conspiracy to commit a predicate offense. Instead,
the Court found that the statute’s requirement of a pattern of
predicate acts, plus the requirement of an enterprise
structure—liberally defined as “a group with a common purpose
and course of conduct”—was sufficient to preserve the
distinction.

Not only was it unnecessary to require proof of a
structure “beyond that inherent in the pattern of racketeering
activity,” but the phrase was also potentially misleading. For “if
the phrase is used to mean that the existence of an enterprise
may never be inferred from the evidence showing that persons
associated with the enterprise engaged in a pattern of
racketeering activity, it is incorrect.” Id. at 2245. As the Court
had observed in Turkette, “the evidence used to prove the
pattern of racketeering activity and the evidence establishing an
enterprise ‘may in particular cases coalesce.’” Id. (quoting
Turkette, 452 U.S. at 583); see also id. at 2246 n.5 (“Even if the
same evidence may prove two separate elements, this does not
mean that the two elements collapse into one.”). In other words,

161
“proof of a pattern of racketeering activity may be sufficient in
a particular case to permit a jury to infer the existence of an
association-in-fact enterprise.” Id. at 2247.64 For this reason,

64
Writing in dissent, Justice Stevens agreed with the majority
that “[t]here may be cases in which a jury can infer [the
existence of an enterprise] from the evidence used to establish
the pattern of racketeering activity.” Id. at 2249 (Stevens, J.,
dissenting). But he believed that should “be true only when the
pattern of activity is so complex that it could not be performed
in the absence of structures or processes for planning or
concealing the illegal conduct beyond those inherent in
performing the predicate acts.” Id. By that standard, Justice
Stevens found the jury instructions approved by the majority to
be “plainly deficient.” Id. at 2251. In allowing the jury to “‘find
an enterprise where an association of individuals, without
structural hierarchy, forms solely for the purpose of carrying out
a pattern of racketeering acts,’” the instructions failed, he
argued, to “require the Government to prove that the alleged
enterprise had an existence apart from the pattern of predicate
acts.” Id.; see also id. at 2250 (“By permitting the Government
to prove [the ‘enterprise’ and ‘pattern of racketeering activity’]
elements with the same evidence, the [majority] renders the
enterprise requirement essentially meaningless in association-in-
fact cases.”). By contrast, the majority held that an association-
in-fact enterprise need not do anything other than engage in the
pattern of racketeering activity, so long as it has the requisite
structural features (common purpose, interrelationships among

162
the trial judge’s instructions to the jury, which explained that a
RICO enterprise could be “form[ed] solely for the purpose of
carrying out a pattern of racketeering acts,” and that “the
existence of an association-in-fact is oftentimes more readily
proven by what it does, rather than by abstract analysis of its
structure,” were “correct and adequate.” Id. at 2242, 2247
(internal quotation marks omitted).

In short, Boyle holds that the RICO statute defines an
“enterprise” broadly, such that the “enterprise” element of a §
1962(c) claim can be satisfied by showing a “structure,” that is,
a common “purpose, relationships among those associated with
the enterprise, and longevity sufficient to permit these associates
to pursue the enterprise’s purpose.” Id. at 2244; see id. at 2245
(“[A]n association-in-fact enterprise is simply a continuing unity
that functions with a common purpose.”). “[A]fter Boyle, an
association-in-fact enterprise need have no formal hierarchy or

its associates, and longevity). Cf. Odom v. Microsoft Corp., 486
F.3d 541, 551 (9th Cir. 2007) (en banc) (concluding that “the
Supreme Court’s statement in Turkette that an ‘enterprise’ is ‘an
entity separate and apart from the pattern of activity in which it
engages’” is not a requirement “that an associated-in-fact
enterprise have a structure beyond that necessary to carry out its
pattern of illegal racketeering activities,” but rather “merely a
statement of the obvious: The enterprise and its activity are two
separate things. One is the enterprise. The other is its
activity.”).

163
means for decision-making, and no purpose or economic
significance beyond or independent of the group’s pattern of
racketeering activity.” United States v. Hutchinson, 573 F.3d
1011, 1021 (10th Cir.), cert. denied, 130 S. Ct. 656 (2009). To
the extent our cases have interpolated additional requirements
into the statute, they are abrogated by Boyle.

Neither Turkette nor Boyle (nor Riccobene, for that
matter) specifically addressed requirements for pleading civil
RICO claims.65 Plaintiffs contend these cases speak only to
which “attributes of an enterprise must ultimately be proven”;
they do not define what must be plead in the complaint.
Plaintiffs’ EB Br. 70.66 Citing our opinion in Seville Industrial

65
Turkette evaluated a challenge to the validity of a criminal
indictment; Boyle reviewed jury instructions in a criminal case;
and Riccobene evaluated a challenge to the sufficiency of the
evidence supporting a criminal RICO conviction.
66
Boyle was decided after the parties filed their briefs in this
appeal. Although plaintiffs’ argument distinguishing burdens of
proof from pleading burdens thus does not explicitly refer to
Boyle, its logic would seem to call for interpreting that decision,
like Turkette and Riccobene, as addressed only to burdens of
proof.
Plaintiffs also argue, in the alternative, that their
allegations adequately plead the enterprise features set forth in
the Turkette/Boyle line of cases. See infra.

164
Machinery Corp. v. Southmost Machinery Corp., 742 F.2d 786
(3d Cir. 1984), plaintiffs argue that it is sufficient at the pleading
stage simply to identify the entities of which the alleged
association-in-fact enterprise is composed. If that were the
standard, plaintiffs would have satisfied it, as they have named
the defendant broker and defendant insurers within each alleged
broker-centered enterprise. See Comm. SAC ¶ 502; EB SAC ¶
523.

Plaintiffs, however, misconceive the pleading standard.
To begin with, Seville was decided before the Supreme Court
held in Bell Atlantic Corp. v. Twombly that Federal Rule of
Procedure 8 requires plaintiffs to plead “enough factual matter”
to state “a claim to relief that is plausible on its face.” 550 U.S.
at 556, 570. But putting Twombly aside for the moment, we are
skeptical that Seville, read purely on its own terms, deems it
sufficient merely to identify the members of an alleged
association-in-fact enterprise. It is true that Seville found neither
Turkette nor Riccobene “speaks to what must be pleaded in
order to state a cause of action.” Seville, 742 F.2d at 790. It is
also true that Seville concluded the complaint before it was
adequately pled because the plaintiff had “identified the four
entities it believed were the enterprises that had been marshalled
against it.” Id. (stating that “[t]he rules of pleading require
nothing more at this early juncture than that bare allegation”).
We find it significant, however, that Seville involved legal-entity
enterprises, not associations in fact. See id. at 789–90. Each of
the four enterprises alleged there consisted solely of a named

165
defendant—either a corporation or a natural person. Simply
identifying these legal entities was sufficient to “put the
defendant[s] on notice,” id. at 790, of the nature of the
enterprises alleged (and, for that matter, to indicate the existence
of the features set forth in Turkette).

When the asserted enterprise, however, is not itself a
legal entity, but rather an association of legal entities, simply
identifying the allegedly associated components does not serve
to put defendants on notice of the RICO claim alleged against
them—just as merely listing the names of alleged conspirators
would not give defendants adequate notice of an alleged
conspiracy. For that reason, even before Twombly, some courts
required plaintiffs to provide more detail in pleading the
existence of an association-in-fact enterprise. See, e.g.,
Richmond v. Nationwide Cassel L.P., 52 F.3d 640, 646 (7th Cir.
1995) (agreeing “with the district court that plaintiff’s naming
of a string of entities does not allege adequately an enterprise”).

In any case, it is clear after Twombly that a RICO claim
must plead facts plausibly implying the existence of an
enterprise with the structural attributes identified in Boyle: a
shared “purpose, relationships among those associated with the
enterprise, and longevity sufficient to permit these associates to
pursue the enterprise’s purpose.” Boyle, 129 S. Ct. at 2244; see
Rao v. BP Prods. N. Am., Inc., 589 F.3d 389, 400 (7th Cir. 2009)
(upholding dismissal of RICO claims because, inter alia, the
plaintiff’s allegations of an association-in-fact enterprise “do not
indicate how the different actors are associated and do not

166
suggest a group of persons acting together for a common
purpose or course of conduct”); Elsevier Inc. v. W.H.P.R., Inc.,
692 F. Supp. 2d 297, 307 (S.D.N.Y. 2010) (finding that the
complaint “fail[s] to plead the existence of . . . the so-called
association in fact enterprise” because it does not plausibly “tie[]
together the various defendants allegedly comprising the
association in fact into a single entity that was formed for the
purpose of working together,” that is, “acting in concert”); see
also Phillips v. County of Allegheny, 515 F.3d 224, 234 (3d Cir.
2008) (stating generally that Twombly “‘requires a complaint
with enough factual matter (taken as true) to suggest’ [each]
required element” of the claim alleged). To require less would
ignore Twombly’s interest in “insist[ing] upon some specificity
in pleading before allowing a potentially massive factual
controversy to proceed” to an “inevitably costly and protracted
discovery phase.” Twombly, 550 U.S. at 558 (internal quotation
marks omitted); see id. at 557–58 (quoting Dura
Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336, 347 (2005))
(noting that the potential cost of such cases may have an “in
terrorem” effect on defendants, coercing settlement of even
groundless claims). As other courts have recognized, the
concern expressed in Twombly is just “as applicable to a RICO
case, which resembles an antitrust case in point of complexity
and the availability of punitive damages and of attorneys’ fees
to the successful plaintiff. RICO cases, like antitrust cases, are
‘big’ cases and the defendant should not be put to the expense
of big-case discovery on the basis of a threadbare claim.”
Limestone Dev. Corp. v. Vill. of Lemont, 520 F.3d 797, 803 (7th

167
Cir. 2008).67

In Twombly, this concern focused on the agreement
element of § 1 of the Sherman Act. Because § 1 allows
plaintiffs to bring suit against conspiracies, it has the potential
to impose liability on a large number of defendants. To prevail,
however, a § 1 plaintiff must show not simply that the
defendants all engaged in similar wrongdoing, but that they
agreed to undertake concerted action in restraint of trade.

The enterprise element of RICO claims is a close
analogue of § 1’s agreement element. Unless a plaintiff is
required at the pleading stage to suggest plausibly the existence
of an enterprise structure—unless a plaintiff must “allege
something more than the fact that individuals were all engaged
in the same type of illicit conduct during the same time period,”
Elsevier, 692 F. Supp. 2d at 307—the RICO statute’s allowance
for association-in-fact enterprises becomes an open gateway to
the imposition of potentially massive costs on numerous
defendants, regardless of whether there is even a hint of the
collaboration necessary to trigger liability.

ii. The “Conduct” Element

Mere association with an enterprise does not violate §

67
We do not imply that Twombly’s pleading standard is
applicable only to “big” cases, but we note that the “practical”
reasons for this standard, see Twombly, 550 U.S. at 557–58, are
particularly evident in such cases.

168
1962(c). To be liable under this provision, a defendant must
“conduct or participate, directly or indirectly, in the conduct of
such enterprise’s affairs through a pattern of racketeering
activity.” 18 U.S.C. § 1962(c). The Supreme Court has held
that the “conduct or participate” element requires a defendant to
“have some part in directing those affairs.” Reves v. Ernst &
Young, 507 U.S. 170, 179 (1993). More precisely, “one is not
liable under [§ 1962(c)] unless one has participated in the
operation or management of the enterprise itself.” Id. at 183.
“An enterprise is ‘operated’ not just by upper management but
also by lower rung participants in the enterprise who are under
the direction of upper management.” Id. at 184. “Outsiders”
may also meet the statutory requirement if they “exert control
over” the enterprise, but such outsider defendants must have
“conducted or participated in the conduct of the ‘enterprise’s
affairs,’ not just their own affairs.” Id. at 184–85. Accordingly,
the Supreme Court found that an outside accounting firm did not
“conduct or participate, directly or indirectly, in the conduct” of
a farmer cooperative’s affairs when it prepared and presented
audit reports to the cooperative’s board, reports that allegedly
failed to disclose all of the information necessary to assess the
solvency of the cooperative. Id. at 185–86; see also Univ. of
Md. at Baltimore v. Peat, Marwick, Main & Co., 996 F.2d 1534,
1539–40 (3d Cir. 1993) (“It cannot be said that by merely
performing what are generic financial and related services to an
insurance company, even if they are later found to be deficient,
an accounting firm has opened itself to liability under the federal
racketeering statute.”).

169
 iii. The Requisite Nexus

Simply pleading that a defendant “participated in the
operation or management” of an enterprise, however, is not
enough to make out a violation of § 1962(c). The defendant
must have done so “through a pattern of racketeering activity.”
In other words, there must be not only a “nexus between the
[defendant] and the conduct [of] the affairs of an enterprise,”
Univ. of Md., 996 F.2d at 1539, but also a nexus between the
conduct of those affairs and the pattern of racketeering activity,
see Banks v. Wolk, 918 F.2d 418, 424 (3d Cir. 1990). The plain
language of the statute requires that the “pattern of racketeering
activity” be a means by which the defendant “participate[s],
directly or indirectly, in the conduct of [the] enterprise’s
affairs.” In United States v. Provenzano, we adopted the Second
Circuit’s test for determining when a defendant “conducts the
activities of an enterprise through a pattern of racketeering.”
688 F.2d 194, 200 (3d Cir. 1982) (citing United States v. Scotto,
641 F.2d 47 (2d Cir. 1980)). The test is satisfied if (1) the
defendant “is enabled to commit the predicate offenses solely by
virtue of his position in the enterprise or involvement in or
control over the affairs of the enterprise; or (2) the predicate
offenses are related to the activities of that enterprise.” Id.
(quoting Scotto, 641 F.2d at 54).

The Second Circuit has repudiated this standard,
however, finding it inconsistent with the definition of “conduct”
subsequently set forth by the Supreme Court in Reves. United
States v. Wong, 40 F.3d 1347, 1372 (2d Cir. 1994) (“Although

170
Reves does not mention Scotto, ‘we have recently recognized
that the Supreme Court’s holding in Reves . . . is irreconcilable
with the relevant portion of our decision in Scotto.’” (quoting
United States v. Viola, 35 F.3d 37, 40 (2d Cir. 1994)); see also
Bldg. Indus. Fund v. Local Union No. 3, 992 F. Supp. 162, 179
n.10 (E.D.N.Y. 1996) (“To the extent that Scotto held that a
person may be liable under RICO simply because the requisite
nexus exists between his activity and the affairs of the
enterprise, it was overruled by the Supreme Court’s decision in
Reves [], holding that a person must be involved in the
‘operation or management’ of an enterprise to be liable under
RICO.”).

We agree with the Second Circuit that the
Provenzano/Scotto standard is no longer good law. 68 Given

68
We recited the Provenzano standard post-Reves in United
States v. Irizarry, 341 F.3d 273, 304 (3d Cir. 2003), without
addressing the potential tension with Reves or acknowledging
that the Second Circuit had recognized the standard’s
abrogation. Irizarry’s invocation of the standard appears to be
dicta, as our opinion noted that the argument to which it was
addressed had been waived. Id. In any case, the issue in
Irizarry was whether the judge should have instructed the jury
that it could not convict under § 1962(c) unless the government
showed “that the defendant’s motive in committing the predicate
act was to further the affairs of the enterprise.” Id. In finding,
correctly, that “a defendant can commit a predicate act that is

171
Reves, the inquiry must be whether the defendant participated in
the “operation or management” of an enterprise’s affairs, and if
so, whether he did so “through a pattern of racketeering
activity.” As the plain language of the statute indicates, the
nexus element requires a plaintiff to show that the defendant
participated in the conduct of the enterprise’s affairs (per Reves)
through—that is, “by means of, by consequence of, by reason of,
by the agency of, or by the instrumentality of,” United States v.
Brandao, 539 F.3d 44, 53 (1st Cir. 2008) (internal quotation
marks omitted)—a pattern of racketeering activity.69

detrimental to the enterprise so long as the evidence establishes
the requisite nexus between the predicate act and the enterprise,”
id., we had no occasion to address any other aspects of the
Provenzano standard.
69
Although we adopt Brandao’s definition of the statutory
term “through,” we believe the decision inverts the relationship
specified by § 1962(c). Brandao states that “[a] sufficient nexus
. . . exists . . . if the defendant was able to commit the predicate
acts by means of, by consequence of, by reason of, by the
agency of, or by the instrumentality of [i.e., ‘through’] his
association with the enterprise.” 539 F.3d at 53 (internal
quotation marks omitted). But the statute provides that a
defendant must participate in the conduct of an enterprise’s
affairs through the racketeering activity. See 18 U.S.C. §
1962(c). This language dictates that it is the predicate acts of
racketeering that must be the “means.”

172
 b. Section 1962(d)

Under § 1962(d), it is unlawful to conspire to violate any
of the substantive provisions of RICO. 18 U.S.C. § 1962(d); see
18 U.S.C. § 1962(a)–(c) (substantive provisions). 70 In certain
circumstances, a defendant may be held liable under § 1962(d)
even where its own actions would not amount to a substantive
RICO violation. Salinas v. United States, 522 U.S. 52, 65
(1997) (stating that a conspirator can violate § 1962(d) “in any
number of ways short of agreeing to undertake all of the acts
necessary for the crime’s completion”)71 ; see Smith v. Berg, 247

70
Here, § 1962(c) is the only substantive provision invoked
by plaintiffs’ complaints.
71
In Salinas, the Supreme Court explained § 1962(d)’s
relationship to general conspiracy doctrine:
The relevant statutory phrase in § 1962(d) is “to
conspire.” We presume Congress intended to use
the term in its conventional sense, and certain
well-established principles follow. . . . When
Congress passed RICO in 1970, the American
Law Institute’s Model Penal Code permitted a
person to be convicted of conspiracy so long as he
“agrees with such other person or persons that
they or one or more of them will engage in
conduct that constitutes such crime.” As the
drafters emphasized, “so long as the purpose of

173
F.3d 532, 537 (3d Cir. 2001) (stating that, under Salinas, a
defendant “who opts into or participates in a conspiracy” to
violate § 1962(c) may be liable “even if the defendant did not
personally agree to do . . . any particular element” of the §
1962(c) violation); see, e.g., Salinas, 522 U.S. at 65 (holding
that the RICO statute “does not permit us to excuse from the
reach of [§ 1962(d)’s] conspiracy provision an actor who does
not himself commit or agree to commit the two or more
predicate acts requisite to the underlying [§ 1962(c)] offense”).
But a § 1962(d) claim must be dismissed if the complaint does
not adequately allege “an endeavor which, if completed, would
satisfy all of the elements of a substantive [RICO] offense,”
Salinas, 522 U.S. at 65. See Efron v. Embassy Suites (Puerto
Rico), Inc., 223 F.3d 12, 21 (1st Cir. 2000); Lightning Lube, Inc.

the agreement is to facilitate commission of a
crime, the actor need not agree ‘to commit’ the
crime.”. . . A conspirator must intend to further
an endeavor which, if completed, would satisfy all
of the elements of a substantive criminal offense,
but it suffices that he adopt the goal of furthering
or facilitating the criminal endeavor. . . . It is
elementary that a conspiracy may exist and be
punished whether or not the substantive crime
ensues, for the conspiracy is a distinct evil,
dangerous to the public, and so punishable in
itself.
522 U.S. at 63–65 (internal citations omitted).

174
v. Witco Corp., 4 F.3d 1153, 1191 (3d Cir. 1993); see, e.g.,
Edwards v. First Nat. Bank, 872 F.2d 347, 352 (10th Cir. 1989)
(finding that “the conspiracy claim falls when the substantive
claim based on 1962(c) is deficient” because the alleged
racketeering acts do not compose a “pattern”); cf. Berg, 247
F.3d at 538 (holding a defendant may be held liable for
conspiring to violate § 1962(c), even if he did not operate or
manage a RICO enterprise, as long as “he knowingly agree[d]
to facilitate a scheme which includes the operation or
management of a RICO enterprise”).72

72
Salinas appears to hold that a violation of § 1962(d) does
not require a consummated violation of a substantive RICO
provision; it is sufficient that the conspiracy have as its object
acts which, if completed, would constitute a substantive
violation. Whether a plaintiff who had not been injured by a
substantive violation would have standing to bring a civil action
for violation of § 1962(d), however, is a different question. See
18 U.S.C. § 1964(c) (creating a civil cause of action for “[a]ny
person injured in his business or property by reason of a
violation of section 1962” and providing for triple damages);
Beck v. Prupis, 529 U.S. 494, 501 n.6 (2000) (distinguishing
“the question of what constitutes a violation of § 1962(d)” from
“the meaning of a civil cause of action for private injury by
reason of such a violation”). In Beck, the Supreme Court
explained that “a civil conspiracy plaintiff cannot bring suit
under RICO based on injury caused by any act in furtherance of
a conspiracy that might have caused the plaintiff injury. Rather

175
 2. Application to This Case

a. Section 1962(c) Claims

i. The Broker-Centered Enterprises

Plaintiffs contend they have pled facts plausibly

. . . a RICO conspiracy plaintiff [must] allege injury from . . . an
act that is independently wrongful under RICO.” Id. at 505–06
(abrogating Shearin v. E.F. Hutton Group, Inc., 885 F.2d 1162,
1168–69 (3d Cir. 1989)). Accordingly, the Court held that the
petitioner, who had been terminated by his employer for
blowing the whistle on RICO activities, did not have standing to
allege a § 1962(d) violation, since his injury was not “caused by
an overt act that is . . . an act of racketeering or otherwise
wrongful under RICO.” Id. at 505. At the very least, then, Beck
stands for the proposition that a plaintiff bringing a § 1962(d)
claim for conspiracy to violate § 1962(c) must allege injury from
a racketeering act enumerated in § 1961(1). But Beck did not
make clear whether that requisite racketeering act must be part
of a consummated § 1962(c) violation. Indeed, the Court
explicitly reserved the question “whether a plaintiff suing under
§ 1964(c) for a RICO conspiracy must allege an actionable
violation under §§ 1962(a)–(c), or whether it is sufficient for the
plaintiff to allege an agreement to complete a substantive
violation and the commission of at least one act of racketeering
that caused him injury.” Id. at 506 n.10.

176
suggesting that each defendant broker and its insurer-partners
composed an association-in-fact enterprise. The District Court
disagreed. Central to its conclusion was its finding that
although plaintiffs had adequately alleged bilateral agreements
(regarding the steering of business and the payment of
contingent commissions) between each broker and its insurer-
partners, plaintiffs had failed to plead facts plausibly suggesting
collaboration among the insurers. The asserted hub-and-spoke
structures therefore lacked a “unifying ‘rim.’” 2007 WL
2892700, at *20. In the absence of a plausible “rim” or “wheel”
connecting the alleged insurer “spokes,” the District Court
determined that while plaintiffs may have alleged parallel,
bilateral structures connecting a broker to each of its insurer-
partners, they had failed to plead “broker-centered enterprises”
encompassing each broker “hub” and all of its strategic partners.

With respect to all but the Marsh-centered enterprise
alleged in the Commercial complaint, we agree with the District
Court that plaintiffs’ allegations of broker-centered enterprises
are fatally defective. In our analysis of the antitrust claims, we
determined that, with the exception of the alleged Marsh-
centered commercial conspiracy, the facts alleged in the
complaints do not plausibly imply a horizontal agreement among
the insurer-partners. In seeking to establish a “rim” enclosing
the insurer-partners in the alleged RICO enterprises, plaintiffs
rely on the same factual allegations we found deficient in the
antitrust context: that each insurer entered into a similar
contingent-commission agreement in order to become a

177
“strategic partner”; that each insurer knew the identity of the
broker’s other insurer-partners and the details of their
contingent-commission agreements; that each insurer entered
into an agreement with the broker not to disclose the details of
its contingent-commission agreements; that the brokers utilized
certain devices, such as affording “first” and “last looks,” to
steer business to the designated insurer; and that, in the
Employee Benefits Case, insurers adopted similar reporting
strategies with regard to Form 5500. As noted, these allegations
do not plausibly imply concerted action—as opposed to merely
parallel conduct—by the insurers, and therefore cannot provide
a “rim” enclosing the “spokes” of these alleged “hub-and-
spoke” enterprises.

Even under the relatively undemanding standard of
Boyle, these allegations do not adequately plead an association-
in-fact enterprise. They fail the basic requirement that the
components function as a unit, that they be “put together to form
a whole.” Boyle, 129 S. Ct. at 2244 (internal quotation marks
omitted). Because plaintiffs’ factual allegations do not plausibly
imply anything more than parallel conduct by the insurers, they
cannot support the inference that the insurers “associated
together for a common purpose of engaging in a course of
conduct.” Id. (quoting Turkette, 452 U.S. at 583); see id. at
2245 n.4 (stating that “several individuals” who “engaged in a
pattern of crimes listed as RICO predicates” “independently and
without coordination” “would not establish the existence of an
enterprise”); Elsevier, 692 F. Supp. 2d at 307 (stating that, as

178
with a § 1 Sherman Act claim, a RICO claim pleading “nothing
more than parallel conduct by separate actors” is insufficient:
“there has to be something that ties together the various
defendants allegedly comprising the association in fact into a
single entity that was formed for the purpose of working
together—acting in concert—by means of” racketeering acts);
Gregory P. Joseph, Civil RICO: A Definitive Guide 106 (3d ed.
2010) (stating that a “rimless hub-and-spoke configuration
would not satisfy the ‘relationships’ prong of Boyle’s structure
requirement”); see also Rao, 589 F.3d at 400 (finding the
plaintiff had failed to plead an association-in-fact enterprise
because the “allegations do not indicate how the different actors
are associated and do not suggest a group of persons acting
together for a common purpose or course of conduct”). Were
the rule otherwise, competitors who independently engaged in
similar types of transactions with the same firm could be
considered associates in a common enterprise. Such a result
would contravene Boyle’s definition of “enterprise.” 73

As the District Court acknowledged, although the
complaints do not adequately plead these asserted broker-
centered enterprises, it is possible that plaintiffs’ factual
allegations would provide a plausible basis for the assertion of
a number of bilateral enterprises, each encompassing a broker

73
We do not address the other grounds relied on by the
District Court to support its conclusion that these “broker-
centered enterprises” were inadequately pled.

179
and one of its insurer-partners, or even the assertion that
individual brokers or insurers each constituted an enterprise.
But the District Court determined that “Plaintiffs’ three previous
rounds of pleadings unambiguously indicate that Plaintiffs have
no interest in asserting” such smaller-scale enterprises. 2007
WL 2892700, at *34 n.32. We agree. While plaintiffs
strenuously insist they have adequately pled the existence of
“broker-centered enterprises,” they have conspicuously
refrained, throughout the district-court proceedings and on
appeal, from asserting alternative bilateral or single-entity
enterprises. Accordingly, the District Court properly dismissed
the claims based on these broker-centered enterprises with
prejudice.

As with the antitrust claims, we reach a different
conclusion with respect to the claims alleging bid rigging—the
bid-rigging allegations in the Commercial complaint suffice to
plead a “Marsh-centered enterprise.” 74 As Boyle clarified, a
RICO “enterprise” must have a structure, but it need not have
any particular structural features beyond “a purpose,
relationships among those associated with the enterprise, and

74
Plaintiffs also allege a “Marsh-centered enterprise” in the
Employee Benefits complaint, but they do not allege that the
associates of this “enterprise” engaged in bid rigging. For the
reasons given above, we agree with the District Court that
plaintiffs have not adequately pled the existence of such an
enterprise.

180
longevity sufficient to permit these associates to pursue the
enterprise’s purpose.” Boyle, 129 S. Ct. at 2244. We think the
allegations of bid rigging provide the “rim” to the Marsh-
centered enterprise’s hub-and-spoke configuration, satisfying
Boyle’s requirements. The Commercial complaint alleges that
Marsh prepared “broking plans” governing the placement of
insurance contracts that came up for renewal. According to
plaintiffs, “[t]he broking plans assigned the business to a
specific insurer at a target price and outlined the coverage. The
broking plans also included instructions as to which preferred
Insurers would be asked to provide alternative [i.e., intentionally
uncompetitive, or sham] quotes. If the incumbent Insurer hit the
‘target,’ it would get the business and then [Marsh employees]
would solicit ‘alternative’ . . . quotes from other members of the
conspiracy.” Comm. SAC ¶ 117. The complaint also alleges
the reasons why the insurers agreed to provide sham bids. For
example, it relates a statement by a former employee of a
defendant insurer that his employer had agreed to “provide[]
losing quotes” to Marsh in exchange for, among other things,
Marsh’s “getting ‘quotes from other [insurance] carriers that
would support the [employer, at least when it was the incumbent
carrier] as being the best price.’” Comm. SAC ¶ 109. This
statement plausibly evinces an expectation of reciprocity and
cooperation among the insurers.

In at least one sense, plaintiffs’ allegations regarding the
“Marsh-centered enterprise” exceed Boyle’s requirements.
Boyle explicitly disavowed the need for any particular

181
organizational structure. Boyle, 129 S. Ct. at 2245–46. It
upheld the conviction of a “loosely and informally organized”
group of bank robbers that neither “had a leader [n]or
hierarchy,” nor “ever formulated any long-term master plan or
agreement”; the group would meet before each robbery to
“assign the roles that each participant would play (such as
lookout and driver).” Id. at 2241. Here, by contrast, plaintiffs
allege a hierarchical structure according to which Marsh, in
accordance with its “broking plan,” decided from which insurer
each sham bid would be requested. Plaintiffs adequately allege
a “common interest” or “purpose,” id. at 2244, namely to
increase profits by deceiving insurance purchasers about the
circumstances surrounding their purchase. See Comm. ARCS
66. The alleged reciprocal bid rigging also adequately suggests
“relationships among” the insurers “associated with the
enterprise[s]”; if proved, it would plausibly demonstrate the
insurers “joined together” in pursuit of the aforementioned
common purpose. Boyle, 129 S. Ct. at 2244 (internal quotation
marks omitted).75 Finally, the complaint alleges that the bid
rigging occurred over a period of several years, plausibly
alleging “that the enterprise had ‘affairs’ of sufficient duration
to permit an associate to ‘participate’ in those affairs through ‘a
pattern of racketeering activity.’” Id. (quoting § 1962(c)).
Accordingly, plaintiffs have adequately pled the enterprise

75
Defendants do not dispute that the complaint sufficiently
alleges the requisite relationships between each insurer and
Marsh.

182
element of the RICO claims based on the alleged Marsh-
centered commercial enterprise.76

76
The number of defendants alleged to have engaged in bid
rigging appears to be slightly smaller than the number of
defendants alleged to be associates of the Marsh-centered
enterprise. See supra note 44. Compare Comm. SAC ¶ 502
(naming alleged members of an association-in-fact Marsh-
centered commercial enterprise), with Plaintiffs’ Comm. Br. 78
n.17 (claiming that the defendant insurers that engaged in bid
rigging are “AIG, ACE, Axis, Chubb, XL, Munich/AmRe,
Liberty Mutual, St. Paul Travelers, Fireman’s Fund and
Zurich”), and Comm. RPS ¶¶ 27–56 (detailing bid-rigging
allegations). While, as noted, plaintiffs have declined to plead
alternative enterprises consisting of only individual defendants
or bilateral relationships, they urge us to permit them “to
proceed with their RICO claims as to Marsh and the insurers
explicitly alleged to have been involved in bid-rigging.”
Plaintiffs’ Comm. Br. 78 n.17. Such a “downscaled” Marsh-
centered enterprise has been adequately pled.
It may also be worth reiterating that membership in an
enterprise is not the touchstone of § 1962(c) liability. Rather, it
is the operation of that enterprise’s affairs through a pattern of
racketeering that constitutes a violation. Accordingly, although
the Commercial complaint here appears to allege that each
defendant was a member of the enterprise it “conducted,” as a
general matter defendants who are outside an enterprise may in

183
 The District Court, proceeding without the benefit of
Boyle, found that the bid-rigging allegations were not sufficient
to plead a Marsh-centered enterprise. The court believed that
the defendant insurers’ participation in each bid-rigging
transaction was “ad hoc.” See 2007 WL 2892700, at *22
(stating that “in the picture painted by Plaintiffs[,] a refusal by
Insurer-Defendant ‘X’ to provide ‘B’ quotes to the requesting
Broker-Defendant was of little substantive consequence to
consummation of [each] transaction because alternative ‘B’
quotes were readily available to the Broker-Defendant from a
range of other Insurer-Defendants”). Given the alleged
existence of Marsh’s “broking plan,” which designated, for each
renewal contract, the insurer from which a sham bid would be
solicited, we cannot say that insurer participation in bid rigging
was completely “ad hoc.” 77 In any case, Boyle makes clear that
there is no need for a systematic plan ordaining in advance who

certain circumstances be held liable under § 1962(c), as the
Supreme Court has acknowledged. See Reves, 507 U.S. at
184–85.
77
The District Court suggested that defendant insurers may
have sometimes refused Marsh’s request for a false bid. But
even if there were occasional refusals, we do not believe they
would prove fatal to plaintiffs’ enterprise allegations. Cf. Boyle,
129 S. Ct. at 2245 (“[N]othing in RICO exempts an enterprise
whose associates engage in spurts of activity punctuated by
periods of quiescence.”).

184
is to provide a sham bid for a particular transaction; “decisions
may be made on an ad hoc basis and by any number of
methods.” Boyle, 129 S. Ct. at 2245. Nor do “[m]embers of the
group need [to] have fixed roles; different members may
perform different roles at different times.” Id.; cf. id. at 2241
(upholding RICO conviction predicated on an enterprise in
which the participants never “formulated any long-term master
plan or agreement” but instead assigned roles to each associate
on a robbery-by-robbery basis). Furthermore, we see no
indication that a RICO enterprise requires participants to have
non-interchangeable, non-substitutable functions.

The District Court also appeared to believe that the bid-
rigging allegations did not adequately plead interrelationships
among the insurers, as opposed to simply bilateral relationships
between Marsh and each insurer. As the District Court read the
complaint, “an Insurer-Defendant ‘X’ cared little which other
Insurer-Defendant would be ‘accommodated’ by its ‘B’ quotes
and whether these ‘B’ quotes would actually be used at all by
the requesting Broker-Defendant; the sole point of ‘X’s’ interest
was to ensure that ‘X’ would be rewarded” by the Broker, in
some form, in exchange for its willingness to generate sham
bids. 2007 WL 2892700, at *21. In other words, the insurers
provided sham bids to the broker, at the request of the broker, in
exchange for benefits provided by the broker. We agree that the
complaint does not allege that the bid rigging took the form of
quid pro quo transactions between insurers; Insurer “X” did not
promise to provide a sham bid facilitating the renewal of Insurer

185
“Y’s” account in exchange for Insurer “Y’s” providing a sham
bid facilitating the renewal of Insurer “X’s” account. But, as
noted, the complaint does allege that one reason the insurers
were willing to furnish sham bids was so that they would be the
beneficiaries of sham bids in the future. To be sure, an insurer
might not need a sham bid to win the renewal of any particular
account. It stands to reason, however, that sometimes a sham
bid would be necessary—for example, when Marsh’s client
insisted on seeing multiple bids. In our view, the alleged
agreement by insurers to provide sham bids plausibly suggests
an interrelationship among the insurers—mediated through
Marsh—in pursuit of achieving greater business and profits by
means of deceiving insurance purchasers. Through this
interrelationship, the insurers were allegedly able to advance this
common interest to a greater extent than would have been
possible on the strength of the bilateral relationships between
Marsh and each broker alone. Under Boyle, this (plus the
requisite longevity) is enough to plead an enterprise.

The District Court believed plaintiffs had also failed to
plead the “conduct” element of their claims regarding the
Marsh-centered enterprise, that is, failed to plead that defendants
“conduct[ed], or participate[d], directly or indirectly, in the
conduct of such enterprise’s affairs.” 18 U.S.C. § 1962(c). To
some extent, the District Court’s conclusion appears to derive
from its determination regarding the “enterprise” element of

186
these claims.78 At least one of its findings, however, appears
particular to the “conduct” element. In an implicit reference to
the Supreme Court’s Reves decision, the District Court stated
that it was “not convinced” that “Defendants operated” the
alleged Marsh-centered enterprise’s affairs “rather than
Defendants’ own affairs.” 2007 WL 2892700, at *31. Reves
explained that liability under § 1962(c) “depends on showing
that the defendants conducted or participated in the conduct of
the ‘enterprise’s affairs,’ not just their own affairs.” 507 U.S.
at 185. In Reves, the Court found that an outside accounting
firm hired to audit an enterprise’s books did not participate in
the conduct of the enterprise’s affairs simply by doing its job
(even if its job performance was deficient). Id. at 185–86; see
United States v. Oreto, 37 F.3d 739, 750 (1st Cir. 1994) (stating
that “the reason the accountants were not liable in Reves is that,
while they were undeniably involved in the enterprise’s
decisions, they neither made those decisions nor carried them
out”).

Here, however, the defendants are alleged to be members
of the enterprise. It will often be the case that the interests of
the enterprise are congruent with those of its members; such
congruence presumably provides the incentive for members to
participate in the enterprise. We think, therefore, that “if

78
The overlap between the District Court’s analyses of these
two elements is understandable. A defendant cannot participate
in conducting the affairs of a non-existent enterprise.

187
defendants band together to commit [violations] they cannot
accomplish alone . . . then they cumulatively are conducting the
association-in-fact enterprise’s affairs, and not [simply] their
own affairs.” Joseph, supra, at 74; see Brandao, 539 F.3d at 54.
Here, defendants’ alleged collaboration in the Marsh-centered
enterprise, most notably the bid rigging, allowed them to
deceive insurance purchasers in a way not likely without such
collusion.

Moreover, we believe that, based on the complaint’s
allegations, plaintiffs have adequately pled that defendants
engaged in activities constituting participation in the conduct of
the enterprise. The allegations that defendant broker Marsh
directed the placement of insurance contracts and solicited
rigged bids from insurers plausibly imply that Marsh
“participated in the operation or management of the enterprise
itself.” Reves, 507 U.S. at 183. And by allegedly supplying the
sham bids, Marsh’s insurer-partners are also adequately alleged
to have “operated” the enterprise within the meaning of Reves.
Cf. id. at 184 (stating that “[a]n enterprise is ‘operated’ not just
by upper management but also by lower rung participants in the
enterprise who are under the direction of upper management”);
e.g., MCM Partners, Inc. v. Andrews-Bartlett & Assocs., 62 F.3d
967, 978–79 (7th Cir. 1995) (finding that two businesses
participated in the conduct of an association-in-fact enterprise
“by knowingly implementing decisions” by the enterprise’s
managers to commit crimes). We thus find plaintiffs have
sufficiently pled that defendants in the Marsh-centered

188
enterprise satisfied the “conduct” requirement set forth in
Reves.79

In summary, we find that plaintiffs have adequately pled
both the “enterprise” and “conduct” elements of the § 1962(c)
claims based on the alleged Marsh-centered commercial
enterprise. Accordingly, we will vacate the District Court’s
judgments insofar as they dismissed those claims, and we will
remand to allow the District Court an opportunity to evaluate the
remaining elements of those claims.80

ii. The CIAB Enterprise

79
Whether plaintiffs have adequately alleged that defendants
participated in the conduct of the Marsh-centered enterprise’s
affairs “though a pattern of racketeering activity” is, of course,
another matter. This question necessarily turns in part on
whether plaintiffs have sufficiently pled a “pattern of
racketeering activity.” See infra note 80.
80
In their motion to dismiss, defendants also contended that
plaintiffs had failed to plead adequately a pattern of predicate
offenses. In a footnote, the District Court suggested this
challenge might have merit but did not conclusively adjudicate
it, instead granting the motion on the ground that “Plaintiffs’
pleading of the ‘enterprise’ and ‘conduct’ elements warrant
dismissal without granting Plaintiffs leave to amend.” 2007 WL
2892700, at *33 n.29.

189
 In the Commercial Case, plaintiffs also bring § 1962(c)
claims naming the Council of Insurance Agents and Brokers
(CIAB), a trade organization, as a RICO enterprise. According
to the complaint, the CIAB “represents the largest, most
profitable of all commercial insurance agencies and brokerage
firms. The Council’s primary audience is CEOs and their
management teams. The Council partners with its members and
provides not only vital intelligence on current market conditions
and trends, but also solutions to the next challenge before the
need arises.” Comm. SAC ¶ 513 (internal quotation marks
omitted). Plaintiffs aver that defendant insurers’ and brokers’
association with the CIAB violated the RICO statute insofar as
defendants “used vital intelligence gained through
communications and meetings facilitated by CIAB and
otherwise, including information about decreased profits and
demand to devise [their customer allocation] scheme[s] . . . to
replace competition. The [defendants] operating through the
CIAB Enterprise reached consensus on how they would change
the market and regarding non-disclosure” of the details of their
allegedly unlawful contingent-commissions-for-guaranteed-
premium-volume schemes. Id. ¶ 523.

The District Court acknowledged the complaint had
adequately pled that the CIAB, a legal entity, was an enterprise,
but it concluded the complaint failed to allege adequately that
defendants had participated in the conduct of the CIAB
“through” the acts of fraud pled as the requisite pattern of

190
racketeering activity.81 In the District Court’s view, the
complaint’s allegations “merely indicate that CIAB provided an
internal and external communication tool for the members of the
insurance industry. No fact stated in Plaintiffs’ submissions
indicates that Defendants’ alleged predicate acts of mail and
wire fraud (aimed at Defendants’ clients) were related in any
way to the activities of CIAB, or that Defendants committed the
alleged predicate offenses through the means of CIAB, or that
CIAB was somehow indispensable to Defendants in their
alleged goal to commit the underlying predicate offenses.” 2007
WL 1062980, at *17. Although the court believed that plaintiffs
had adequately alleged that the CIAB afforded defendants an
opportunity to meet and communicate about their challenged
schemes, it found that simply making use of the forum provided
by the Council did not plausibly imply the requisite nexus:
“[W]ere Plaintiffs to allege that Defendants habitually met with
each other . . . in a chain of coffee houses or through an Internet
portal[,] such allegations would not render the corporation
owning such chain of coffee houses or the online portal a RICO

81
Plaintiffs conclusorily allege that “Defendants have
conducted or participated in the conduct of the affairs of the
CIAB Enterprise through a pattern of racketeering activity,”
Comm. SAC ¶ 527, but as Twombly observed, “a plaintiff’s
obligation to provide the grounds of his entitlement to relief
requires more than labels and conclusions, and a formulaic
recitation of the elements of a cause of action will not do,”
Twombly, 550 U.S. at 555 (internal quotation marks omitted).

191
enterprise for the purposes of the inquiry at bar. Neither the
CIAB-based [allegations], nor the coffee house- or Internet
portal-based alternative provides this Court with a fact drawing
‘a nexus between [Defendants’ alleged offenses] and the
conduct in the affairs’ of [the alleged enterprise].” Id. (quoting
United States v. Parise, 159 F.3d 790, 796 (3d Cir. 1998)) (first
alteration in District Court opinion).

As the District Court observed, plaintiffs’ nexus theory
appears to rest primarily on the alleged fact that the CIAB
provided an “opportunity” or “forum” for defendants to discuss
and advance their schemes.82 We agree with the District Court

82
See, e.g., Comm. SAC ¶ 513 (“CIAB provides Defendants
with numerous opportunities to communicate, meet, use vital
intelligence on market conditions that is shared with its partner
members, and reach agreement on how they will address
challenges in the marketplace . . . .”); id. ¶ 518 (“CIAB provides
the Broker Defendants a forum to discuss and reach agreement
with each other and with the Insurer Defendants regarding,
among other things, compensation arrangements and other
aspects of their relationships, what each wants and needs from
the relationship, the market and market conditions, consolidation
and disclosure.”); id. ¶ 519 (“CIAB hosts ‘Executive Forums’
. . . [which] allow members the opportunity to ‘discuss common
problems and solutions.’”); id. ¶ 521 (“CIAB has also provided
Defendants the opportunity to discuss and reach agreement on
joint action in response to the regulatory investigations and

192
that the allegation that defendants took advantage of an
opportunity to meet provided by a legitimate enterprise in the
normal course of its business does not mean—or plausibly
imply—that defendants were participating in the conduct of the
enterprise. By extension, the allegation that defendants utilized
such an opportunity to plot or discuss or otherwise facilitate a
pattern of racketeering activity does not, without more, plausibly
imply that defendants conducted the enterprise’s affairs through
a pattern of racketeering activity. If it did, any coffee house or
hotel with conference facilities could, as the District Court
rightly recognized, be made into a RICO enterprise merely by
dint of the fact that racketeers used the facility as a meeting
place. In our view, such an expansive interpretation of §
1962(c) liability would not comport with the text or purpose of
the RICO statute, nor with the cases interpreting it. As we have
observed in interpreting the Supreme Court’s decision in Reves,
“[s]imply because one provides goods or services that ultimately
benefit the enterprise does not mean that one becomes liable
under RICO as a result.” Univ. of Md., 996 F.2d at 1539.
Similarly, we do not think a RICO violation occurs simply
because an enterprise provides goods or services that ultimately
benefit a defendant’s racketeering efforts; in order to constitute

regarding disclosure issues.”); id. ¶ 522 (“CIAB provides
Defendants with numerous opportunities to communicate, meet,
use vital intelligence on market conditions that is shared with its
partner members, and reach agreement on how they will address
challenges in the marketplace . . . .”).

193
a § 1962(c) violation, it is necessary that the defendant, in
utilizing these goods or services, “participate . . . in the conduct
of [the] enterprise’s affairs.” § 1962(c). Availing oneself of a
forum provided by an enterprise does not, without more,
plausibly imply that one has participated in the conduct of that
enterprise’s affairs.

In an apparent attempt to supply this “something more,”
plaintiffs allege that “[t]he purpose of the CIAB Enterprise is to
further the interests of larger brokers generally and to further the
Defendants’ scheme specifically.” Comm. SAC ¶ 524. But to
claim that the purpose of the CIAB was to further the
defendants’ scheme is simply to assert what must be shown
through well-pled factual allegations. See Twombly, 550 U.S.
at 555; cf. supra note 81. And the alleged fact that the CIAB
worked to further the interests of larger brokers generally does
not plausibly imply the necessary nexus between defendants’
racketeering activity and the conduct of the CIAB’s affairs. The
Chamber of Commerce exists to promote the interests of its
member businesses, but this does not mean that if some of its
members plot their racketeering activities at a Chamber meeting,
they have participated in the conduct of the Chamber’s affairs.
In short, plaintiffs cannot distinguish trade unions like the CIAB
from the District Court’s hypothetical coffee house simply by
invoking trade unions’ “general” mission to advance their
members’ interests.

Our analysis thus confirms the District Court’s
conclusion that the allegations it recited were inadequate to

194
plead the nexus element of plaintiffs’ CIAB-based RICO claims.
We believe, however, that the complaint might reasonably be
construed as asserting additional facts that present a closer
question. Specifically, the complaint appears to allege that
defendant brokers did not merely make use of the CIAB as a
forum, but actually utilized its institutional machinery to
formulate strategy and issue public statements in aid of their
fraudulent acts. The complaint avers that the defendant brokers
effectively operated and controlled the CIAB; the brokers sat on
its Board of Directors. CIAB allegedly formulated a “position
statement that was intended to stave off any meaningful
regulatory disclosure requirements and reassure insurance
purchasers that their brokers were acting in their best interests.”
Comm. SAC ¶ 445. “The CIAB position statement was made
available for review and editing by its members before being
finally approved by CIAB’s Executive Committee, which
included executives from Marsh, Aon, and HRH. . . . Further,
one or more representatives of all of the Broker Defendants, or
their predecessors, were on CIAB’s Board of Directors at the
time the CIAB position statement was issued.” Id. ¶ 446. The
Brokers then allegedly incorporated language from this position
statement into their allegedly fraudulent disclosures. See id.
¶ 458 (“The [Broker] disclosures were modeled after the CIAB
position statement in order to create the impression of
transparency, by stating that the brokers ‘may’ have contingent
commission agreements that might result in some additional
revenue, while failing to disclose any information regarding the
strategic partnerships that the Broker Defendants had entered

195
into with the Insurer Defendants or the significance of these
partnerships and the contingent payments arrangements [for] the
insurance placement process and the premiums charged.”)

Assuming the truth of these allegations, one might
arguably make a plausible inference that defendant brokers
participated in the conduct of the CIAB enterprise, and
furthermore, that one way they operated the enterprise was to
have it craft the allegedly misleading disclosure statements
incorporated by defendants into their allegedly fraudulent
communications with clients (and perhaps others).83 This
inference, however, would raise additional questions that would
have to be resolved before we could conclude that plaintiffs
have adequately pled their CIAB-based RICO claims. First,
under § 1962(c), which is the only substantive RICO provision
invoked by plaintiffs, the defendants must be distinct from the
enterprise they conduct. See Gasoline Sales, Inc. v. Aero Oil
Co., 39 F.3d 70, 72 (3d Cir. 1994); see also Cedric Kushner
Promotions, Ltd. v. King, 533 U.S. 158, 161 (2001) (“We do not
quarrel with the basic principle that to establish liability under
§ 1962(c) one must allege and prove the existence of two
distinct entities: (1) a ‘person’; and (2) an ‘enterprise’ that is not
simply the same ‘person’ referred to by a different name. The

83
This possible inference pertains only to the defendant
brokers. The complaint does not contain well-pled factual
allegations plausibly implicating the defendant insurers in the
creation of the disclosure statements.

196
statute’s language, read as ordinary English, suggests that
principle.”). Here, plaintiffs appear to allege that the CIAB
crafted fraudulent statements released directly to the public and
incorporated into defendants’ own disclosures. Defendants are
alleged to have sat on the board of directors of CIAB. Although
the Supreme Court has held that a corporate owner may
successfully be pled as a defendant distinct from the corporate
enterprise it owns, Cedric Kushner, 533 U.S. at 163–66, the
Court’s decision examined a complaint clearly alleging the
defendant owner’s role in the acts supposedly constituting both
a “pattern of racketeering activity” and participation in the
conduct of the enterprise. See Cedric Kushner, No. 98 Civ.
6859, 1999 WL 771366, at *1–2 (S.D.N.Y. Sept. 28, 1999),
aff’d, 219 F.3d 115 (2d Cir. 2000), rev’d, 533 U.S. 158. The
complaint here, by contrast, appears to say little about defendant
brokers’ involvement in CIAB’s crafting of the allegedly
fraudulent statements other than that defendants sat on CIAB’s
board at the time. There may therefore be a question about
whether plaintiffs have adequately pled not only a connection
between the CIAB and the alleged pattern of racketeering
activity, but also the requisite connection between that pattern
and defendants’ conduct of the CIAB.84

84
As noted, the requisite nexus element would not be satisfied
by the mere allegation that the defendants’ alleged racketeering
activity benefitted from goods or services provided by the
CIAB.

197
 Second, there is a question about whether plaintiffs have
adequately pled that defendant brokers’ allegedly misleading
disclosure statements constitute a “pattern of racketeering
activity.” In other words, even assuming plaintiffs have
adequately alleged a connection between defendants’ conduct of
the CIAB and the statements alleged to be fraudulent, it remains
to be determined whether the fraudulent nature of these
statements has itself been adequately alleged. Cf. Am. Dental
Ass’n v. Cigna Corp., 605 F.3d 1283, 1290–93 (11th Cir. 2010)
(holding that the district court properly dismissed plaintiffs’ §
1962(c) claims because plaintiffs had not sufficiently pled the
acts of mail and wire fraud alleged to form a pattern of
racketeering activity). Although the District Court expressed
significant skepticism about plaintiffs’ allegations of fraud, its
dismissal of the CIAB-based RICO claims does not appear to be
based on this issue.85

In light of these significant questions, as well as our
discussion of the nexus standard, see supra Section II.B.1.a.iii.,
we will vacate the dismissal of the CIAB-based claims. None
of these questions was squarely addressed by the parties on
appeal, and we believe the District Court is best positioned to
decide them in the first instance. On remand, the District Court

85
See supra note 80 and accompanying text.

198
may proceed in any way consistent with our opinion.86

b. Section 1962(d) Claims

The District Court dismissed the § 1962(d) RICO
conspiracy claims on the ground that plaintiffs had failed to
plead any viable § 1962(c) claims. Because we will vacate the
dismissal of the § 1962(c) claims relating to the alleged Marsh-
centered commercial enterprise and the CIAB enterprise, we
will also vacate the § 1962(d) claims based on those two
enterprises. We will affirm the District Court’s dismissal of the
remaining § 1962(d) claims because with respect to these
claims, plaintiffs have failed to allege adequately “an endeavor
which, if completed, would satisfy all of the elements of a
substantive [RICO] offense.” Salinas, 522 U.S. at 65.

C. State-Law Claims

A district court “may decline to exercise supplemental
jurisdiction” over state-law claims if it “has dismissed all claims
over which it has original jurisdiction.” 28 U.S.C. § 1367(c)(3).
Having dismissed all federal-law claims, the District Court here
exercised this statutory grant of discretion and dismissed
plaintiffs’ state-law claims. Because we will vacate in part the
judgment dismissing the federal claims, we will also vacate the

86
The court may choose in its discretion, for instance, to
address the adequacy of plaintiffs’ fraud allegations before
returning to the question of the nexus between that alleged fraud
and defendants’ alleged conduct of the CIAB.

199
dismissal of the state-law claims. See Shaev v. Saper, 320 F.3d
373, 384 (3d Cir. 2003).

III. Conclusion

For the forgoing reasons, we will vacate the dismissal of
the Sherman Act claims with respect to defendants alleged to
have engaged in bid rigging in the Marsh-centered commercial
conspiracy; the dismissal of the RICO claims based on the
alleged Marsh-centered commercial enterprise, with respect to
those same defendants; the dismissal of the RICO claims based
on the alleged CIAB enterprise, with respect to the defendant
brokers; and the dismissal of the state-law claims. We will
affirm the District Court’s judgment in all other respects and
remand for further proceedings consistent with this opinion.

200