United States Court of Appeals
For the First Circuit
No. 09-1196
IN RE PHARMACEUTICAL INDUSTRY AVERAGE
WHOLESALE PRICE LITIGATION
M. JOYCE HOWE,
Plaintiff, Appellant,
v.
LEROY TOWNSEND,
Plaintiff, Appellee.
APPEAL FROM THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF MASSACHUSETTS
[Hon. Patti B. Saris, U.S. District Judge]
Before
Lynch, Chief Judge,
Torruella and Howard, Circuit Judges.
Donald E. Haviland with whom Michael J. Lorusso and The
Haviland Law Firm, LLC were on brief for the appellant.
Steve Berman with whom Sean R. Matt, Thomas M. Sobol, Edward
Notargiacomo, Hagens Berman Sobol Shapiro LLP, Kenneth R. Wexler,
Jennifer Fountain Connolly, Wexler Wallace LLP, Jeffrey Kodroff,
John A. Marcoretta, Spector, Roseman, Kodroff & Willis P.C., Marc
H. Edelson, and Hoffman & Edelson were on brief for the appellee.
November 19, 2009
LYNCH, Chief Judge. One unhappy named plaintiff appeals
from an order approving a $24 million class action settlement in
one multidistrict litigation, in which class members allege that
AstraZeneca Pharmaceuticals LP (AstraZeneca) published artificially
inflated prescription drug prices. In re Pharm. Indus. Average
Wholesale Price Litig., MDL No. 1456, Civ. A. No. 01-12257 (D.
Mass. Dec. 15, 2008) (order approving final settlement) (In re
Pharm. Final Settlement Approval). The settlement is between
AstraZeneca and a class of consumer plaintiffs who claimed they
overpaid Medicare co-payments because AstraZeneca inflated the
price of Zoladex.
M. Joyce Howe, one of the class representatives for the
AstraZeneca consumer subclass, appeals. The other AstraZeneca
subclass representative, Leroy Townsend, asks us to uphold the
settlement. Howe argues that the settlement must be rejected on
the grounds that it creates a cy pres fund of up to $10 million
rather than distributing all recovery to class members; that the
settlement is not fair, reasonable, and adequate because its method
for calculating and distributing class members' damages is flawed;
and that the parties allegedly improperly negotiated fees
simultaneously with the settlement. Howe failed to raise the last
argument before final approval of the settlement.
Howe also argues to us two procedural objections she did
not make in the district court. Howe contends the district court's
-2-
order approving the settlement, which expanded the settlement
class, did not sufficiently define the new class and its claims as
required by Rule 23(c)(1)(B). Howe also argues that the district
court did not properly reappoint class counsel under Rule 23(g),
see Fed. R. Civ. P. 23(g), when approving the expanded settlement
class.
The cy pres and Rule 23(c)(1)(B) issues are ones of first
impression for this court. We find Howe's challenges meritless and
affirm the district court, which handled this matter with great
sensitivity and care.
I.
This appeal is in one case of a series of class actions
alleging pharmaceutical companies fraudulently inflated a figure
known as the "average wholesale price" (AWP) between 1991 and 2003
to boost sales.1 This court has already decided other appeals from
1
The AWP was supposed to reflect the average amount
wholesalers charged providers for pharmaceutical drugs.
Wholesalers bought drugs from manufacturers, marked up the price
about 20 to 25 percent to make a profit, and resold them to
physicians. Patients obtaining drugs from physicians either paid
out of pocket or had their insurers reimburse physicians, while
usually making a co-payment. Insurers and Medicare reimbursed
physicians based on published AWPs, which in turn determined
patients' co-payments.
Over time, the acquisition costs for physicians plummeted.
Wholesalers added much smaller markups, around 2.5 percent, due to
increasing competition. And pharmaceutical companies began using
rebates and similar devices to further drive down the cost of
drugs. Yet pharmaceutical companies did not adjust published AWPs
to reflect this market change. Instead, companies continued to
push costs down but reported AWPs based on the 20 to 25 percent
markup.
-3-
these multidistrict litigations and related cases. See In re
Pharm. Indus. Average Wholesale Price Litig., 582 F.3d 231 (1st
Cir. 2009); In re Pharm. Indus. Average Wholesale Price Litig., 582
F.3d 156 (1st Cir. 2009); Nat'l Ass'n of Chain Drug Stores v. New
England Carpenters Benefits Fund, 582 F.3d 30 (1st Cir. 2009); U.S.
ex rel. Duxbury v. Ortho Biotech Prods., L.P., 579 F.3d 13 (1st
Cir. 2009).
The healthcare industry treated the AWP, which
pharmaceutical companies self-reported, as the sticker price for
drugs. Insurers used AWPs to decide how much to reimburse
providers when patients obtained drugs, which in turn affected
patients' co-payments. Most of the drugs involved in these
litigations are expensive drugs for serious illnesses, including
cancer. The drug involved in this appeal, Zoladex, is commonly
used to treat prostate cancer. Consequently, Medicare, insurers,
and patients allegedly overpaid by many millions of dollars for
critical treatments based on manipulated AWPs.
A coalition of citizen, healthcare, senior citizen, and
consumer advocacy groups sued dozens of drug manufacturers and
Because Medicare and insurers continued to rely on published
AWPs, the "spread" between physicians' acquisition costs and their
reimbursements grew dramatically, generating increasing profits for
doctors. Pharmaceutical companies marketed this spread to
physicians to encourage doctors to prescribe the companies' drugs.
-4-
healthcare-product companies2 in the District Court for the
District of Massachusetts on December 19, 2001. Similar lawsuits
were filed across the country, and on April 30, 2002, the Judicial
Panel on Multidistrict Litigation consolidated the lawsuits and
assigned the case to the District Court for the District of
Massachusetts. In re Immunex Corp. Average Wholesale Price Litig.,
201 F. Supp. 2d 1378, 1379-81 (J.P.M.L. 2002).
The district court ultimately divided the multidistrict
litigation into two tracks, a fast track (Track One) and a normal
track (Track Two). Track One included three classes. One of the
three--Class 1, called the Medicare Part B Co-Payment Class--
consisted of patients covered by Medicare Part B who made co-
payments for these drugs. A subclass of Class 1, which sued
AstraZeneca for inflating the price of Zoladex, is involved in this
appeal.
The Medicare Part B Co-Payment Class alleged that the
defendants violated state consumer protection statutes by
artificially inflating AWPs. Between 1992 and 2005, Medicare Part
B reimbursements were based on published AWPs under federal law.3
2
The complaint also named numerous co-conspirators and Doe
defendants, possibly including other companies and doctors who
participated in fraudulent billing.
3
Medicare reimbursed physicians 100 percent of the AWP
from 1992 to 1998, up to 95 percent of the AWP until 2003, and 85
percent of the AWP until 2005, when the federal government finally
stopped relying on AWPs altogether.
-5-
Medicare covered 80 percent of this cost, requiring patients to
make a 20 percent co-payment.
Most patients had supplemental insurance that covered
some but not all of this co-payment. Those who had supplemental
coverage for the full co-payment were excluded from the class.
Most class members had Medigap insurance that covered 80 percent of
the 20 percent co-payment (leaving patients to pay 20 percent of
the 20 percent Medicare co-payment). Thus patients in the class
were differently affected depending on whether they had
supplemental insurance and how much of the co-payment that
insurance covered.
On August 16, 2005, the district court denied some
proposed classes and deferred certifying the Medicare Part B Co-
Payment class until the plaintiffs located individuals to act as
class representatives. In re Pharm. Indus. Average Wholesale Price
Litig., 230 F.R.D. 61, 96 (D. Mass. 2005) (order partially denying
and partially deferring class certification) (In re Pharm.
Preliminary Class Certification).4 The court issued the order
after carefully analyzing whether the Medicare Part B Co-Payment
Class, as well as all proposed Track One classes, met the
requirements for class certification under Rule 23 of the Federal
Rules of Civil Procedure. Id. at 76-96. The plaintiffs filed an
4
The district court issued many orders in this case, only
a handful of which are relevant to this appeal. We refer to orders
based on their content for the purposes of this appeal only.
-6-
amended complaint in October 2005 that reflected the court's
instructions.
The district court certified the Medicare Part B Co-
Payment Class under Rule 23(b)(3) of the Federal Rules of Civil
Procedure on January 30, 2006. In re Pharm. Indus. Average
Wholesale Price Litig., 233 F.R.D. 229, 230-31 (D. Mass. 2006)
(order granting and denying class certification) (In re Pharm.
Class Certification). It defined the class as "[a]ll natural
persons nationwide who made, or who incurred an obligation
enforceable at the time of judgment to make, a co-payment based on
AWP for a Medicare Part B covered Subject Drug that was
manufactured by" the defendants, including AstraZeneca, between
January 1, 1991, and January 1, 2005. Id. at 229-32 (footnote
omitted). The court excluded residents in nine states whose
consumer-protection statutes did not create a cause of action,
unlike the law in states whose residents were certified as class
members. Id. at 230. It certified class counsel under Rule 23(g).
Id. at 232.
The court divided the plaintiffs into four subclasses
according to which pharmaceutical-company defendant had distributed
the subject drugs to them. Id. at 230. The court certified Howe5
5
The court certified Howe individually and on behalf of
her husband's estate.
-7-
and Townsend as subclass representatives for the class members
suing AstraZeneca. Id.
After months of intense litigation and negotiation,
AstraZeneca and the Zoladex subclass reached a final settlement in
May 2007, on the eve of trial.6 That settlement is at issue in
this appeal. The plaintiffs' expert had calculated the class's
damages were $31,128,851. In the settlement, AstraZeneca agreed to
pay $24 million in total compensation to all class members or, for
deceased class members, to their spouses or the legal
representatives of their estates. In an effort to achieve a
comprehensive settlement, the parties also agreed AstraZeneca would
pay the consumers from nine states who had previously been excluded
from the class. If the amount of submitted claims exceeded $24
million, class members' recovery would be reduced proportionally.
AstraZeneca also promised to pay all costs of
distributing this money as well as $8.7 million in attorneys' fees
and costs, representing about one-third of the settlement amount.
The parties agreed AstraZeneca would pay fees and costs separately
and not deduct that from the class members' $24 million recovery.
The parties expected a large portion of the total sum
AstraZeneca was willing to pay would go unclaimed. Because Zoladex
6
In June 2007 some members of Classes 2 and 3 of the Track
One plaintiffs proceeded to a bench trial against four defendants,
including AstraZeneca, and the district court ruled in favor of the
plaintiffs. In re Pharm. Indus. Average Wholesale Price Litig.,
491 F. Supp. 2d 20, 29-32 (D. Mass. 2007) (In re Pharm. Trial).
-8-
is used to treat prostate cancer, many class members were elderly,
had died, or could die soon, and not all of them could be found.
The parties addressed this issue through two mechanisms.
First, rather than paying the money up front into a fund
that would distribute money to class members, AstraZeneca agreed to
pay class members only on a comparatively easy claims-made basis.
Class members only had to submit relatively minimal proof of when
they took Zoladex. Then, based on a formula created by the
plaintiffs' expert, class members would receive compensation for
double their actual losses. Because the formula tried to calculate
actual losses, it calculated the damages of class members who had
supplemental insurance were 20 percent of their 20 percent Medicare
co-payment, based on the average Medigap coverage that class
members had.
Second, the parties agreed to create a fund, called a cy
pres fund, that would pay any amount remaining (after payout to the
claimants) of the $24 million to "mutually acceptable charitable
organizations funding cancer research or patient care" that the
court would approve in the future. The parties capped the cy pres
amount at $10 million if not all of the $24 million was paid out.
The parties submitted this agreement to the district
court for preliminary approval on May 21, 2007. Howe quickly
objected to the proposed settlement. She then raised six concerns
but did not repeat all of them later in the settlement process.
-9-
Those concerns were (1) the settlement changed the composition of
the class, (2) the settlement compensated members on a claims-made
basis rather than sending class members a check for their share of
the class's damages, (3) the cy pres fund unfairly deprived class
members of recovery, (4) the loss formula should not have reduced
payments to those with supplemental insurance and also failed to
calculate different coverage levels among those with supplemental
insurance, (5) the settlement exposed class members to having other
classes subrogate their claims, and (6) the settlement was tainted
because the parties negotiated it and attorneys' fees
simultaneously.
The district court granted preliminary approval of the
settlement. In re Pharm. Indus. Average Wholesale Price Litig.,
MDL No. 1456, Civ. A. No. 01-12257 (D. Mass. Nov. 1, 2007) (order
preliminarily approving settlement) (In re Pharm. Preliminary
Settlement Approval). In the preliminary approval the court
certified an expanded class that included the residents in nine
states who were previously excluded, finding the class satisfied
all requirements in Rules 23(a) and (b)(3). Id. at 2-4. The court
again certified class counsel under Rule 23(g). Id. at 4.
In 2008 the parties sought final approval of the proposed
settlement. The morning of the fairness hearing, May 1, 2008, Howe
filed a reply brief objecting to the settlement. She again argued
that class members, not a cy pres fund, should receive the entire
-10-
settlement amount; the claims-made process and loss formula were
flawed; and the settlement risked subrogation of the plaintiffs'
claims. Howe proposed adopting the settlement distribution method
used in one Track Two settlement. She argued AstraZeneca should
use a Center for Medicare and Medicaid Services (CMS) database
containing information about Medicare reimbursements to mail all
class members a check for their equal share of the total amount
class members overpaid, regardless of whether they had supplemental
insurance. She also contended the attorneys' fees were excessive.
Although Howe's reply brief was tardy and its arguments
were arguably waived, the district court considered each objection
at the fairness hearing.7 The court found $24 million was a
reasonable settlement. Class counsel reported that they expected
$14 million of the $24 million settlement to go unclaimed. Under
the proposed settlement, up to $10 million of those leftover funds
would be distributed through cy pres. The district judge agreed
with Howe that more money should go to the plaintiffs rather than
to the cy pres fund. She instructed the parties that the
settlement should pay treble damages to those plaintiffs who had
7
Howe had earlier filed an objection to the settlement
that essentially cited, for three pages, Howe's earlier pleadings
before the court without stating any specific grounds for the
objection. During the fairness hearing, the district court
properly held that anything raised in this pleading that Howe did
not explain in the reply brief was waived. See TAG/ICIB Servs.
Inc. v. Sedecu Servicio de Descuento en Compras, 570 F.3d 60, 66
n.6 (1st Cir. 2009) (holding an argument waived in part because the
party failed to develop it in the district court).
-11-
the strongest claims because they obtained Zoladex in the
"heartland period"8 and should correspondingly reduce the cy pres
fund.
The court reserved deciding the attorneys' fees issue and
rejected Howe's other arguments. The district judge agreed with
class counsel that no subrogation issue existed. The court
additionally accepted class counsel's justification for calculating
damages for class members with supplemental insurance based on the
average Medigap coverage, even if some members' insurance covered
less than 80 percent of the Medicare co-payment. Class counsel
reported that Howe's attorney had identified no plaintiff other
than Howe who had different supplemental coverage, the plaintiffs'
expert had reaffirmed that 80 percent coverage fairly approximated
the class's supplemental insurance, and those few with different
coverage would still be compensated beyond their losses.
Class counsel also explained, and the district court
accepted, why Howe's proposal to mail equal checks to all class
members was inappropriate. First, the CMS database did not reveal
whether individuals had supplemental insurance. Because those with
full supplemental insurance were excluded from the class, Howe's
8
The "heartland period," which the district court set as
between December 1997 and 2003, was the time when plaintiffs who
have the strongest claim took the subject drugs. See In re Pharm.
Trial, 491 F. Supp. 2d at 31-32. The court found statutes of
limitations barred claims before 1997 and the AWP scheme after 2003
was not as egregious because Congress changed the Medicare
reimbursement policy. Id.
-12-
method would have the detriments of compensating non-class members,
overcompensating those with supplemental insurance, and
undercompensating those without it. By failing to consider
individual damages, Howe's approach actually would dramatically
reduce individual payouts.9 Second, unlike in this Track One case,
it was prohibitively difficult for the parties in the Track Two
settlement to calculate plaintiffs' individual damages. Those
cases involved many drugs that people took at a variety of dosages
over varying periods. Zoladex, by contrast, is taken at regular
dosages at regular intervals. With information about when a
plaintiff started Zoladex and whether the plaintiff had
supplemental insurance, it was easy to calculate a Zoladex
plaintiff's actual damages.
The parties amended the proposed settlement agreement to
address the court's instruction to increase the payout to class
members. The amended agreement continued to provide that all class
members would receive double their damages. But the new agreement
no longer simply gave leftover money to charity. After all
claimants were paid double damages, all heartland plaintiffs would
receive treble damages pro rata. Then leftover money would be paid
to charity through cy pres. Like the original cy pres payout, the
9
The plaintiffs estimated that, under Howe's proposed
method, class members would get just under $54 each, well short of
the hundreds or thousands of dollars some are entitled to under the
settlement the district court ultimately approved.
-13-
parties capped the additional treble damages and cy pres
distribution at $10 million.
On October 2, 2008, the district court approved the
amended settlement, subject to two further qualifications. In re
Pharm. Indus. Average Wholesale Price Litig., MDL No. 1456, Civ. A.
No. 01-12257 (D. Mass. Oct. 2, 2008) (order partially approving
settlement) (In re Pharm. Partial Settlement Approval). First, the
court instructed the parties to adjust the settlement to pay
heartland plaintiffs treble damages, then, if money remained, to
pay treble damages to all class members pro rata. If money was
still left over, the parties agreed the remaining money would be
distributed to charity through cy pres. Id. at 3. Second, it
reduced the attorneys' fees to 30 percent of the recovery. Id. at
3-4. The district court discussed and rejected Howe's other
concerns. Id. at 5-6.
The parties once more amended the settlement agreement to
reflect the court's instructions. They again limited the total
payout of treble damages to all class members and remaining money
to charity through cy pres to $10 million. The plaintiffs' expert
predicted in February 2009 that, even with class-wide treble
damages, class members would claim only about $17.2 million,
leaving about $6.8 million of the $24 million settlement left over
for charity.
-14-
It is this amended settlement that the district court
approved on December 15, 2008, as a fair, adequate, and reasonable
product of "extensive arm's-length negotiations." In re Pharm.
Final Settlement Approval, MDL No. 1456, Civ. A. No. 01-12257, at
4. Incorporating its prior orders dated August 16, 2005, and
January 16, 2006, the district court found the settlement class
satisfied the requirements of Rules 23(a) and (b)(3). Id. at 2-3.
The court again addressed Howe's objections and rejected
them. It found Howe's proposal that class members should all
receive a check rather than submit a claim was unreasonable because
that approach would ultimately reduce the amount all class members
received. Id. at 5. The court also rejected Howe's contention
that the damages formula unfairly reduced recovery for class
members with supplemental insurance. The reduction in the
settlement better reflected class members' real damages, and all
members would receive more than their losses regardless. Id. at 5-
6. The court similarly held Howe's argument that the formula
unfairly calculated all supplemental-insurance reductions based on
the 80 percent average supplemental coverage rather than based on
individuals' insurance policies was meritless because the number of
affected class members was small and because those with less
coverage would receive more than their damages. Id. at 6. The
district court rejected Howe's other objections. Id.
-15-
Howe now appeals the court order approving the settlement
agreement.
II.
We review the district court's approval of a settlement
for abuse of discretion. City P'ship Co. v. Atl. Acquisition Ltd.
P'ship, 100 F.3d 1041, 1043 (1st Cir. 1996). We review any legal
conclusions de novo. McKenna, 475 F.3d 418, 422 (1st Cir. 2007).
In turn, "[a] district court can approve a class action
settlement only if it is fair, adequate and reasonable," City
P'ship, 100 F.3d at 1043, "or (in shorthand) 'reasonable,'" Nat'l
Ass'n of Chain Drug Stores, 582 F.3d at 44. If the parties
negotiated at arm's length and conducted sufficient discovery, the
district court must presume the settlement is reasonable. City
P'ship, 100 F.3d at 1043. "[T]he district court enjoys
considerable range in approving or disapproving a class settlement,
given the generality of the standard and the need to balance [a
settlement's] benefits and costs." Nat'l Ass'n of Chain Drug
Stores, 582 F.3d at 45.
On appeal Howe argues the settlement is unreasonable
because it creates a cy pres fund, the agreement's claims process
and formula for calculating loss are inappropriate, and the parties
allegedly discussed settlement and attorneys' fees together. Howe
also claims that the court's final approval order did not satisfy
the requirements of Rule 23(c)(B)(1) to define the expanded class
-16-
and its claims and that the court did not properly certify class
counsel under Rule 23(g) when approving the settlement. The
district court did not abuse its discretion by rejecting these
arguments.
A. The District Court Did Not Abuse Its Discretion by
Finding the Settlement Provision Creating a Cy Pres Fund
Was Fair, Adequate, and Reasonable
Appellate courts review a district court's conclusion
that a settlement agreeing to a cy pres distribution is reasonable
for abuse of discretion. E.g., In re Airline Ticket Comm'n
Antitrust Litig., 307 F.3d 679, 682 (8th Cir. 2002); Wilson v. Sw.
Airlines, 880 F.2d 807, 811 (5th Cir. 1989). Howe raises two
challenges to the cy pres distribution in this settlement: that the
class members are entitled to any surplus settlement money and that
class counsel had a conflict of interest when negotiating the cy
pres distribution. We find no abuse of discretion.
1. The Cy Pres Distribution
This circuit has not had occasion to consider whether
settlement agreements in class actions may establish cy pres funds.
Other circuits have approved agreements that provide for cy pres
funds, particularly those negotiated at arm's-length by the
parties. See A. Conte & H.B. Newberg, 4 Newberg on Class Action
§ 11:20, at 28-31 (4th ed. 2002).
-17-
A cy pres10 distribution results from application of a
legal doctrine that courts have imported into class actions from
trusts and estates law. In trusts and estates law, cy pres, taken
from the Norman French expression cy pres comme possible ("as near
as possible"), "save[s] testamentary gifts that otherwise would
fail" because their intended use is no longer possible. In re
Airline Ticket Comm'n, 307 F.3d at 682. Courts permit the gift to
be used for another purpose as close as possible to the gift's
intended purpose. Id.
In class actions, courts have approved creating cy pres
funds, to be used for a charitable purpose related to the class
plaintiffs' injury, when it is difficult for all class members to
receive individual shares of the recovery and, as a result, some or
all of the recovery remains. Id.; 4 Conte & Newberg, supra, §
11:20. The use of a cy pres fund sometimes "prevent[s] the
defendant from walking away from the litigation" without paying a
full recovery because of practical obstacles to individual
distribution. Mirfasihi v. Fleet Mortgage Corp., 356 F.3d 781, 784
(7th Cir. 2004); see also 3 Conte & Newberg, supra, § 10:15, at 513
10
Some courts and commentators have used the terms "cy
pres" and "fluid recovery" interchangeably. E.g., Mirfasihi v.
Fleet Mortgage Corp., 356 F.3d 781, 784 (7th Cir. 2004); Molski v.
Gleich, 318 F.3d 937, 954 (9th Cir. 2003); Democratic Cent. Comm.
of D.C. v. Washington Metro. Area Transit Comm'n, 84 F.3d 451, 455
(D.C. Cir. 1996); 4 Conte & Newberg, supra, § 11:20, at 28. These
terms may be different. See 3 Conte & Newberg, supra, § 10:17, at
517 & n.7. Here we simply use "cy pres."
-18-
("[T]he cy pres . . . distributions serve the objectives of
compensation for the class (albeit in an indirect manner), access
to judicial relief for small claims, and deterrence of illegal
behavior.").
Court-mandated cy pres distributions, as opposed to
court-approved settlements using cy pres, are controversial. See
Democratic Cent. Comm. of D.C. v. Washington Metro. Area Transit
Comm'n, 84 F.3d 451, 455 n.2 (D.C. Cir. 1996); 3 Conte & Newberg,
supra, §§ 10:20-24, at 529-39; 4 id. § 11:20, at 28. But "courts
are not in disagreement that cy pres distributions are proper in
connection with a class settlement, subject to court approval of
the particular application of the funds." 4 Conte & Newberg,
supra, § 11:20, at 28. Because this case involves a settlement, we
need not reach the questions raised by court-created cy pres funds.
Courts have approved cy pres funds in settlements in at
least two circumstances. Id. First, cy pres settlement funds have
been approved when it is economically infeasible to distribute
money to class members. Id.; see, e.g., Mirfasihi, 356 F.3d at 784
(rejecting a settlement because it was feasible to compensate class
members individually); Molski v. Gleich, 318 F.3d 937, 954-55 (9th
Cir. 2003) (same). Distribution of all funds to the class can be
infeasible, for example, when class members cannot be identified,
when the class changes constantly, or when class members'
individual damages--although substantial in the aggregate--are too
-19-
small to justify the expense of sending recovery to individuals.
Mirfasihi, 356 F.3d at 784; Powell v. Ga.-Pac. Corp., 119 F.3d 703,
706 (8th Cir. 1997); 3 Conte & Newberg, supra, § 10:16, at 513 n.1.
In these cases, some courts have permitted the parties to create a
cy pres fund in lieu of a class payout. See Mirfasihi, 356 F.3d at
784. That is not our situation.
Second, courts have allowed parties to establish cy pres
funds when money remained from the defendant's payout after money
for damages had been distributed to class members. In re Airline
Ticket, 307 F.3d at 684. This situation often arises because some
class members never claimed their share. Six (6) Mexican Workers
v. Ariz. Citrus Growers, 904 F.2d 1301, 1307 (9th Cir. 1990).
Among other solutions, courts have approved giving money unclaimed
after payout to class members to charities related to the
plaintiffs' injuries. Id.; e.g., In re Simon II Litig., 407 F.3d
125, 131 (3d Cir. 2005); In re Mexico Money Transfer Litig., 267
F.3d 743, 746 (7th Cir. 2001); Powell, 119 F.3d at 706-07 (8th Cir.
1997); Six (6) Mexican Workers, 904 F.2d at 1307. This kind of cy
pres fund is at issue in this case.
The settlement provision creating a cy pres fund here is
somewhat unusual in its timing. Under the final settlement, all
class members will receive treble damages before any money goes to
cy pres. The parties already know that some original class members
will not file claims because they are elderly or have died and
-20-
their heirs will not stand in their shoes. They expect a
significant amount of the $24 million will go unclaimed, and that
expectation is supported by expert evidence predicting that class
members will claim only about $17.2 million, leaving approximately
$6.8 million left over. The parties have agreed that any remainder
will go, after distribution of treble damages to class members, to
cancer or patient-related charities.
Although unusually timed, the cy pres fund in this case,
contrary to Howe's argument, is not taking damages away from the
class members. The settlement permits all plaintiffs to claim and
be paid their damages--indeed treble their damages--before any
money is paid to charity through cy pres. This process is like
other, routinely approved cy pres distributions. See, e.g.,
Powell, 119 F.3d at 705-06 (refusing, after money in a settlement
fund remained, to distribute the rest to class members because
"neither party ha[d] a legal right" to the unclaimed funds); In re
Folding Carton Antitrust Litig., 744 F.2d 1252, 1253-54 (7th Cir.
1984) (holding a cy pres distribution was appropriate when $6
million remained in a fund created to pay costs and extra claims in
a settlement because "neither the plaintiff class nor the settling
defendants ha[d] any right" to the money). It would elevate form
over substance to require the parties to wait until after all
claims are paid before reaching an agreement as to how to
distribute any remaining money to charity.
-21-
Howe nevertheless insists that the class members are
entitled to receive any remaining money. She argues that, when
deciding how to allocate remaining money, the primary consideration
is whether it is economically infeasible to distribute the
remaining proceeds to all claimants. Howe relies on and, in our
view, Howe misunderstands the American Law Institute's ("ALI")
recent draft of the Principles of the Law of Aggregate Litigation
discussing the appropriateness of cy pres distributions in
settlements. See American Law Institute, Principles of the Law of
Aggregate Litigation § 3.07 (Apr. 1, 2009) (proposed final draft)
[hereinafter "ALI Aggregate Litigation Draft"].11 The question
before us is not whether the settlement complies with the ALI
draft, but whether the district court abused its discretion in
approving the cy pres part of the settlement. Cf. Bobby v. Van
Hook, ___ S. Ct. ___, 2009 WL 3712013, at *3 (Nov. 9, 2009)
(noting, in a criminal appeal arguing that the defendant's counsel
was ineffective, that "'American Bar Association standards and the
like' are 'only guides' to what reasonableness means" (quoting
Strickland v. Washington, 466 U.S. 668, 688 (1984))) (per curiam).
The district court's actions in this case were entirely
congruent with the proposed draft's purposes. The ALI's proposed
11
Howe cites one district court case relying on the ALI
proposed draft. But that court used the draft's reasoning to
approve a settlement, not reject a settlement. In re Tyco Int'l
Multidistrict Litig., 535 F. Supp. 2d 249, 262 (D.N.H. 2007).
-22-
draft expresses a policy preference, when settlement money remains,
for redistributing that money to class members to ensure they
recover their losses. The ALI was concerned that cy pres funds are
often inappropriate because "few settlements award 100 percent of
a class member's losses, and thus it is unlikely in most cases that
further distributions to class members would result in more than
100 percent recovery." ALI Aggregate Litigation Draft § 3.07 cmt.
b. The ALI also believed "that in most circumstances distributions
to class members better approximate the goals of the substantive
laws than distributions to third parties that were not directly
injured by the defendant's conduct." Id.
The district court shared these concerns. The court
ultimately insisted that the settlement pay class members treble
damages before any money is distributed through cy pres. This set
the benchmark well above the ALI's hope that class members might
receive 100 percent recovery.
And the court recognized that the cy pres fund serves the
goals of civil damages by ensuring AstraZeneca fairly pays for the
class's alleged losses. We asked at oral argument why AstraZeneca
would be willing to pay a total sum more than the treble damages
for each class member. Counsel for Plaintiff Townsend replied that
the plaintiffs had insisted on AstraZeneca paying a larger sum to
better represent the losses of the entire class, including those
class members who would never claim their recovery.
-23-
The district court's approval reflected another important
concern: facilitating a settlement in a hard-fought, complex class
action. See Durrett v. Housing Auth., 896 F.2d 600, 604 (1st Cir.
1990) (recognizing a policy encouraging class action settlements).
Achieving settlement in such cases is not easy. District judges
must realistically evaluate settlements based on the circumstances
of the case. "[T]he ultimate decision by the judge involves
balancing the advantages and disadvantages of the proposed
settlement as against the consequences of going to trial or other
possible but perhaps unattainable variations on the proffered
settlement." Nat'l Ass'n of Chain Drug Stores, 582 F.3d at 44.
The court's judgment here was not an abuse of discretion.
2. Conflict of Interest
As we read her argument, Howe also objects that the
district court abused its discretion by finding the settlement was
reasonable even though class counsel allegedly had a conflict of
interest. We have not before considered what effect potential
conflicts of interest by attorneys have on class settlements.
Although they have substantial discretion to approve settlements,
district judges do have responsibility to monitor class counsel's
performance.12 "Because class actions are rife with potential
12
When reviewing settlements, district courts also must
ensure class counsel have met their ongoing duty, imposed by Rule
23(g)(4), to fairly and adequately represent the class. See Key v.
Gillette Co., 782 F.2d 5, 7 (1st Cir. 1986). The duty of adequate
representation requires counsel to represent the class competently
-24-
conflicts of interests between class counsel and class members,
district judges are expected to give careful scrutiny to the terms
of the proposed settlements in order to make sure that class
counsel are behaving as honest fiduciaries for the class as a
whole." Mirfasihi, 356 F.3d at 785 (citations omitted) (collecting
cases).
Howe alleges class counsel had a conflict of interest
with the class because some attorneys had previously represented
Prescription Action Litigation (PAL), an organization dedicated to
reducing costs of pharmaceuticals in part by bringing class actions
challenging industry practices. Howe argues that loyalty to the
class obligated class counsel to demand the highest possible payout
to class members. Counsel's affiliation with PAL, she contends,
instead caused the attorneys to push for a cy pres fund because
PAL's express purpose is to create cy pres funds and because
counsel intended to distribute the fund's proceeds to PAL.
We reject Howe's description of counsel's obligations,
PAL's purposes, and the settlement's terms. As we have explained,
and vigorously and without conflicts of interest with the class.
See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 626 n.20 (1997);
Hanlon v. Chrysler Corp., 150 F.3d 1011, 1020 (9th Cir. 1998);
Rutherford v. City of Cleveland, 137 F.3d 905, 909 (6th Cir. 1998);
Andrews v. Bechtel Power Corp., 780 F.2d 124, 130 (1st Cir. 1985);
see also 1 Conte & Newberg, supra, § 3:22. The duty to avoid
conflicts is serious. Class counsel are fiduciaries to the class.
Rodriguez v. W. Publishing Co., 563 F.3d 948, 968 (9th Cir. 2009);
see also Sondel v. Nw. Airlines, Inc., 56 F.3d 934, 938 (8th Cir.
1995); 1 Conte & Newberg, supra, § 3:40, at 520 & n.5.
-25-
there was nothing inherently wrong with class counsel advocating
for a cy pres fund to benefit those class members whose recovery
would go unclaimed after counsel ensured all claimants would
receive more than their damages. Howe presents no evidence that
PAL's purpose is to promote cy pres distributions, and the record
is otherwise. PAL cannot receive any cy pres funds from this
settlement. The settlement provides the cy pres money will go to
"mutually acceptable charitable organizations funding cancer
research or patient care," a description PAL does not meet.
B. The District Court Did Not Abuse Its Discretion by
Approving the Settlement's Method for Calculating Class
Members' Damages and Distributing the Settlement Amount
Howe raises three objections on appeal to the district
court's approval of the method for calculating and distributing
class members' individual recovery. First, Howe argues in one
sentence that the district court should not have approved the
settlement because it reduced class members' payments if they had
supplemental insurance. This argument is waived because Howe has
failed to develop it on appeal. See United States v. DeCologero,
530 F.3d 36, 60 (1st Cir. 2008).
Even were it not waived, the argument has no merit. The
settlement formula tailors class members' recovery to their actual
losses, since those who had supplemental insurance paid less out of
-26-
pocket than those who did not.13 As the district court found, this
agreement therefore avoids undercompensating class members who
lacked supplemental insurance. See In re Pharm. Final Settlement
Approval, MDL No. 1456, Civ. A. No. 01-12257, at 5; In re Pharm.
Partial Settlement Approval, MDL No. 1456, Civ. A. No. 01-12257, at
5. And by limiting those with supplemental insurance to their
actual co-payment, the settlement makes more money available to
more class members. See In re Pharm. Final Settlement Approval,
MDL No. 1456, Civ. A. No. 01-12257, at 5.
Second, Howe asserts that the settlement should pay class
members with supplemental insurance according to their individual
coverage. Howe's insurance covered 50 percent of her Medicare co-
pay, not 80 percent as the loss formula assumes, and she wants her
share adjusted. The district court rejected this argument because
Howe and any similar class members, at double or treble damages,
would still receive well above their damages. Id. at 6.
That conclusion was not an abuse of discretion. Howe was
the only plaintiff known to counsel (and the court) who had
different coverage. The plaintiffs' expert had reaffirmed his
formula was appropriate despite Howe's objection. And all class
members, even if others existed who, like Howe, had less
13
As the plaintiffs explained to the district court, these
insurers are members of another class against these defendants.
-27-
supplemental coverage, would receive more than their actual
damages.14
Third, Howe again advocates a distribution system like
the one in a Track Two settlement. She argues all class members
should simply receive a check representing their share of the total
losses reported in the CMS database.
The district court could determine this plan was not more
equitable than the distribution method in the settlement. Because
the CMS database does not show who had supplemental insurance, this
system would ultimately reduce class members' payouts. It would
compensate nonmembers and overcompensate class members who had
supplemental insurance. The district court understood that, unlike
in Track Two, it was possible to avoid this problem in this case
because the parties could calculate Zoladex patients' losses.
C. Howe's Objection to Class Counsel Allegedly Negotiating
Attorneys' Fees with the Settlement Is Waived
Howe contends the settlement agreement is "tainted"
because the parties negotiated the settlement and attorneys' fees
together. Class counsel argue they did not. Howe raised this
objection when the district court was considering granting
preliminary but not final approval to the settlement. She
therefore waived it. See Iverson v. City of Boston, 452 F.3d 94,
102 (1st Cir. 2006).
14
The plaintiffs estimate Howe will receive about $200 more
than her actual losses.
-28-
D. The District Court Satisfied the Requirements of Rule
23(c)(1)(B)
The meaning of Rule 23(c) is a legal question we
interpret de novo. See NEPSK, Inc. v. Town of Houlton, 283 F.3d 1,
5 (1st Cir. 2002) (interpreting a federal rule of procedure de
novo).
Howe objects to the court's certification order under
Rule 23(c)(1)(B). She presents this argument for the first time on
appeal, and it is therefore waived. See In re New Motor Vehicles
Canadian Exp. Antitrust Litig., 533 F.3d 1, 6 (1st Cir. 2008).
Nevertheless, we address this issue in order to provide guidance to
courts interpreting and applying this rule.
Rule 23(c)(1)(B), which was added in 2003, requires
orders certifying classes to "define the class and the class
claims, issues, or defenses."15 Fed. R. Civ. P. 23(c)(1)(B). The
advisory notes do not clarify the rule further. This court has not
before addressed the meaning of 23(c)(1)(B). Only the Third
Circuit has interpreted it. Wachtel ex rel. Jesse v. Guardian Life
Ins. Co. of Am., 453 F.3d 179, 184-88 (3d Cir. 2006).
Howe's arguments assume that Rule 23(c)(1)(B) applies to
amended certification orders. As we read her brief, Howe solely
objects that the district court's December 15, 2008, order did not
15
Orders also must "appoint class counsel under Rule
23(g)." Fed. R. Civ. P. 23(c)(1)(B). We address Howe's objections
to class counsel appointment below.
-29-
meet her reading of Rule 23(c)(1)(B). That order certified the
final settlement class, which was expanded to include the residents
from nine states who were previously excluded, and finally approved
the settlement agreement. In re Pharm. Final Settlement Approval,
MDL No. 1456, Civ. A. No. 01-12257. Howe contends the district
court could not satisfy Rule 23(c)(1)(B) by incorporating by
reference, as it did, its prior orders issued on August 16, 2005,
and January 30, 2006, certifying the original class. Id. at 2.
She says the court's December 2008 certification order had to
exhaustively explain and justify the certification decision for the
expanded class.
Howe's brief presents three arguments why the December
2008 order erred under Rule 23(c)(1)(B). First, by incorporating
its prior orders and not giving a full restatement of those orders,
Howe argues, the district court did not define the class or its
claims in sufficient detail. Second, even if the district court
could incorporate prior orders, Howe contends that the August 2005
order, which discussed whether to certify the original class, did
not define the class claims, issues, or defenses with sufficient
clarity. Third, Howe argues, neither the December 2008 order nor
the prior orders explained why the court approved the expanded
settlement class.
We begin with the rule's plain text as well as its
structure and purpose. See Pavelic & LeFlore v. Marvel Entm't
-30-
Group, 493 U.S. 120, 123 (1989) (giving "the Federal Rules of Civil
Procedure their plain meaning"); cf. Simmons v. Galvin, 575 F.3d
24, 35 (1st Cir. 2009) (explaining that we interpret statutes
according to "the language itself, the specific context in which
that language is used, and the broader context of the statute as a
whole" (quoting Nken v. Holder, 129 S. Ct. 1749, 1756 (2009))
(internal quotation marks omitted)). To resolve any ambiguities,
we may also consider the rule's drafting history. Cf. Bercovitch
v. Baldwin Sch., Inc., 133 F.3d 141, 149 (1st Cir. 1998) ("[C]ourts
look first to the text of a statute, and then to the legislative
history if the meaning of the text is ambiguous.").
Rule 23(c)(1)(B) explains the contents of a certification
order: the order must clarify and detail the identity of a class
and the class claims, issues, or defenses in a class action. See
Merriam-Webster's Collegiate Dictionary 303 (10th ed. 1993)
(defining to "define" as "to make distinct, clear, or detailed");
see also Wachtel, 453 F.3d at 185. Rule 23(c)(1) contains two
other provisions governing class certification orders. First,
"[a]t an early practicable time after a person sues or is sued as
a class representative, the court must determine by order whether
to certify the action as a class action." Fed. R. Civ. P.
23(c)(1)(A). And "[a]n order that grants or denies class
certification may be altered or amended before final judgment."
Fed. R. Civ. P. 23(c)(1)(C). The federal rules contemplate
-31-
district courts issuing an order certifying a class and detailing
the class composition and the case's issues and claims, an order
the court can amend before final judgment.
Rule 23(c)(1) does not explain whether 23(c)(1)(B)'s
substantive requirements apply to amended orders issued under
23(c)(1)(C). But Rule 23(c)(1)(B)'s text appears to apply to any
order certifying a class, including orders certifying an amended
class. The rule governs "[a]n order that certifies a class
action." Fed. R. Civ. P. 23(c)(1)(B). The text uses "an" rather
than "the," which suggests the drafters did not mean to restrict
23(c)(1)(B) to a single certification order in a case. Courts have
interpreted similar language in Rule 23(f), which applies to "an
order granting or denying class-action certification," to permit
interlocutory review of amended orders. See Gutierrez v. Johnson
& Johnson, 523 F.3d 187, 199 n.12 (3d Cir. 2008); Carpenter v.
Boeing Co., 456 F.3d 1183, 1191 (10th Cir. 2006).
The text of Rule 23(c)(1) does not specify how courts can
accomplish the rule's aims, but the rule is meant to serve certain
functions. The depth of explanation courts should provide in
amended certification orders depends on the circumstances. Courts
can amend certification orders to reflect major changes or minor
adjustments to the class. See Fed. R. Civ. P. 23(c)(1)(C).
District courts should ensure that, after the new order, the
class's composition and claims remain well defined.
-32-
This interpretation makes sense in light of Rule
23(c)(1)(B)'s history and purpose. Rule 23 was substantially
revised in 2003 to "provide the district courts with the tools,
authority, and discretion to closely supervise class-action
litigation." See Comm. on Rules of Practice and Procedure,
Judicial Conference, Report of the Judicial Conference 8 (Sept.
2002). As part of this project, the rules committee made several
amendments to Rule 23(c) in 2003 that put greater emphasis on
understanding and articulating the "contours" of the class action
throughout the litigation. Wachtel, 453 F.3d at 186; see also L.
J. Hines, "Challenging the Issue Class Action End-Run," 52 Emory
L.J. 709, 763 (2003) (noting the 2003 amendments' emphasis on
defining a class's issues).
An amendment to 23(c)(1)(A) adjusted the timing of class
certification to give courts flexibility to wait to certify the
class until the court feels it understands the case and the issues
it raises.16 Wachtel, 453 F.3d at 186; see also Fed. R. Civ. P.
23(c)(1)(A) advisory comm. notes on 2003 amendments. Rule
23(c)(2)(B) was amended to make notice to class members more
16
The amendment was intended in part to address the
"critical need" to "determine how the case will be tried."
Wachtel, 453 F.3d at 186 (quoting Fed. R. Civ. P. 23(c)(1)(A)
advisory comm. notes on 2003 amendments). The drafters observed
that "[a]n increasing number of courts require a party requesting
class certification to present a 'trial plan' that describes the
issues likely to be presented at trial and tests whether they are
susceptible to class-wide proof." Id. (internal quotation marks
omitted).
-33-
informative. It now requires notice to state "the definition of
the class certified" and "the claims, issues, or defenses." Fed.
R. Civ. P. 23(c)(2)(B)(ii)-(iii).17 Rule 23(c)(1)(B) was added, at
least in part, to help appellate courts reviewing an order better
understand the district court's decision. See Comm. on Rules of
Practice and Procedure, supra, at 11.18
We see many possible benefits to the 2003 amendments to
Rule 23(c). Overall, better comprehension and explanation of the
class and the class claims helps district courts, appellate courts,
attorneys, and parties all proceed with more information and mutual
understanding. See Fed. R. Civ. P. 23 advisory comm. notes on 2003
amendments; see also Wachtel, 453 F.3d at 186-87 (discussing the
effect of the 2003 amendments); Comm. on Rules of Practice and
Procedure, supra, at 9-12 (explaining the purpose of the 2003
amendments to Rule 23(c)).
17
The drafters intentionally made this language parallel
with Rule 23(c)(1)(B). Fed. R. Civ. P. 23(c)(1)(C) advisory comm.
notes on 2003 amendments.
18
The drafting history says Rule 23(c)(1)(B) was added to
assist interlocutory review of certification orders under Rule
23(f). Comm. on Rules of Practice and Procedure, supra, at 11.
Because appellate courts have discretion whether to permit
interlocutory review under 23(f), however, appellate courts often
do not review certification orders until after final judgment. See
Waste Mgmt. Holdings, Inc. v. Mowbray, 208 F.3d 288, 293-94 (1st
Cir. 2000). Rule 23(c)(1)(B) does not limit itself to orders that
will be reviewed on interlocutory appeal. One of its purposes
seems to be to facilitate appellate review at any stage.
-34-
In this case the district court satisfied Rule
23(c)(1)(B)'s text and purpose. The court could incorporate its
prior orders by reference when certifying the expanded class.
Looking at the incorporated orders, the August 2005 order plainly
defined the class and the class claims, issues, and defenses in
sufficient detail. The district court devoted many pages to the
class's factual allegations against the defendant. See In re
Pharm. Preliminary Class Certification, 230 F.R.D. at 65-76. It
then carefully analyzed the proposed class's suitability for
certification, again explaining the issues common to the class.
See id. at 77-85, 96. The court also discussed the state consumer
protection statutes underlying the class's claim, noting
differences among them. Id. at 83-85.
It is also perfectly clear why the district court
expanded the settlement class in the December 2008 order.19 The
court originally excluded residents in these nine states because
the defendants had objected that including residents from these
states defeated the predominance requirement, see Fed. R. Civ. P.
23(b)(3), since their consumer-protection statutes differed. In re
Pharm. Preliminary Class Certification, 230 F.R.D. at 83-85; In re
Pharm. Class Certification, 233 F.R.D. at 230. In the settlement,
however, AstraZeneca bargained for "total peace" to resolve all
19
The rule says only that a court must define, not
necessarily justify, a class and its claims. See Fed. R. Civ. P.
23(c)(1)(B).
-35-
remaining claims against it. When expanding the class, the
December 2008 order references these prior orders and "the record
before the Court." In re Pharm. Final Settlement Approval, MDL No.
1456, Civ. A. No. 01-12257, at 2. That is enough.
E. Howe's Argument That the District Court Did Not Properly
Certify Class Counsel under Rule 23(g) Is Waived
Rule 23(g) requires district courts to appoint class
counsel and governs how courts should choose counsel. Fed. R. Civ.
P. 23(g)(1)-(2). Howe argues the district court did not appoint
class counsel under Rule 23(g) when approving the settlement in
this case and that the district court did not satisfy Rule 23(g)'s
requirements.
Howe is mistaken. The district court did find, while
approving the settlement, that class counsel met Rule 23(g)'s
requirements. In re Pharm. Preliminary Settlement Approval, MDL
No. 1456, Civ. A. No. 01-12257, at 3-4; In re Pharm. Final
Settlement Approval, MDL No. 1456, Civ. A. No. 01-12257, at 3.
Any argument the district court did not satisfy Rule
23(g) is waived because Howe raises this claim for the first time
on appeal. See Rocafort v. IBM Corp., 334 F.3d 115, 121 (1st Cir.
2003).
III.
The approval of the settlement is affirmed.
-36-