Opposition Brief — Frank v. Gaos, 138 S. Ct. 1697 (2018) (No. 17-961)

Supreme Court brief2018

Ask Donna

What actually matters in this document.

Text

No. 17-961

In the Supreme Court of the Anited States

THEODORE H. FRANK and MELISSA ANN HOLYOAK,

Petitioners,

Vv.

PALOMA GAOS, on behalf of herself and all others

similarly situated, et al.,

Respondents.

On Petition for a Writ of Certiorari to the

United States Court of Appeals

for the Ninth Circuit

BRIEF IN OPPOSITION OF RESPONDENT

GOOGLE LLC

RANDALL W. EDWARDS DONALD M. FALK

O'Melveny & Myers LLP Counsel of Record

Two Embarcadero EDWARD D. JOHNSON

Center, 28th Floor Mayer Brown LLP

San Francisco, CA Two Palo Alto Square

94111 3000 El Camino Real

(415) 984-8716 Palo Alto, CA 94306

redwards@omm.com (650) 331-2000

dfalk@ma rown.com

DANIEL E. JONES

Mayer Brown LLP

1999 K Street, NW

Washington, DC 20006

(202) 263-3000

Counsel for Respondent Google LLC

QUESTION PRESENTED

Whether this Court should review the determi-

nation of the two courts below that it was infeasible

to distribute $5.3 million in settlement funds to 129

million class members who had been unable to plead

any concrete injury resulting from the challenged

feature of Internet searches, so that the district court

acted within its discretion in approving the settle-

ment’s cy pres awards to six organizations with es-

tablished Internet privacy programs.

‘i

RULE 29.6 STATEMENT

Respondent Google LLC is a wholly owned sub-

sidiary of XXVI Holdings Inc., which is a wholly

owned subsidiary of Alphabet Inc., a publicly traded

company. No publicly held corporation owns more

than 10% of Alphabet Inc.’s stock.

ill

TABLE OF CONTENTS

Page

QUESTION PRESENTED......................00-0eeceeeeeeeeeeeees i

RULE 29.6 STATEMENT .......................2cccceeeeeeeeeeees ii

, gE iv

ERE ab SS EE eer eae oe OTE 4

A. Referrer headers and web history. ............... 4

B. The Gaos and Priyev actions. ........................ 5

C. Settlement and approval proceedings. ......... 8

D. The Ninth Circuit affirms. .......................... 11

REASONS FOR DENYING THE PETITION......... 14

A. The Purported Conflict Does Not

RSENS RE eA eS A 15

B. The Purported Issue Lacks Recurrent

Importance Because Cy Pres-Only

Settlements Are Sharply Declining

And Reserved For Exceptional Cases. ........ 21

C. This Case Is An Exceedingly Poor

Vehicle To Consider Unprecedented

Limitations On Cy Pres Settlements. ......... 23

1. Class members were unharmed

and unlikely to receive anything

from further litigation........................... 23

2. This case implicates none of the cy

pres abuses that petitioners decry........ 24

D. The Decision Below Is Correct. ................... 28

SUEY cinncisinicunntennictsitognatstedanctinntecbininnemiionahiones 32

TABLE OF AUTHORITIES

Page(s)

CASES

In re Agent Orange Prod. Liab. Litig.,

GES BG BVO GG Ge, BID ccccccencccccccccccccecccovcesees 25

Amchem Prods., Inc. vy. Windsor,

ccna 27

Atl. Marine Constr. Co. v. Dist. Ct. for the

Western Dist. of Texas,

EE a 28

In re Baby Products Antitrust Litig.,

708 F.3d 163 (3d Cir. 2013) ............... 16, 17, 20, 25

In re BankAmerica Corp. Securities Litig.,

775 F.3d 1060 (8th Cir. 2015).............................. 19

Blackman v. Gascho,

No. 16-364, cert. denied 137 S. Ct.

EE SERENE aR 3

Briseno v. ConAgra Foods, Inc.,

844 F.3d 1121 (9th Cir. 2017).............................. 27

Edwards v. First American Corp.,

610 F.3d 514 (9th Cir. 2010)........................ 6, 7, 23

Fraley v. Batman,

638 F. App’x 594 (9th Cir. 2016).......................... 30

Fraley v. Facebook, Inc.,

966 F. Supp. 3d 939 (N.D. Cal. 2013).................. 30

Frank v. Poertner,

No. 15-765, cert. denied 136 S. C°.

ERG SESE SRR ee eee amr 3

TABLE OF AUTHORITIES—continued

Page(s)

Hanlon v. Chrysler Corp.,

150 F.3d 1011 (Oth Cir. 1988).............................- 26

K.D. y. Facebook, Inc.,

ee eee 30

Klier v. Elf Autochem N. Am., Inc.,

658 F.3d 468 (5th Cir. 2011).................... 16, 19, 20

Koby v. ARS Nat Servs., Inc.,

846 F.3d 1071 (9th Cir. 2017)..........................---. 25

Lane v. Facebook, Inc.,

696 F.3d 811 (9th Cir. 2012).............. 10, 11, 20, 22

Marek v. Lane,

I ae 14

Masters v. Wilhelmina Model Agency, Inc.,

473 F.3d 423 (2d Cir. 2007) ..................... 16, 20, 21

Mirfasihi v. Fleet Mortg. Corp.,

356 F.3d 781 (7th Cir. 2004).......................... 16, 19

Mirfasihi v. Fleet Mortg. Corp.,

551 F.3d 682 (7th Cir. 2008)................................ 19

Nachshin v. AOL, LLC,

663 F.3d 1034 (9th Cir. 2011).................. 10, 19, 25

In re Netflix Privacy Litig.,

2013 WL 1120801 (N.D. Cal. Mar.

IRE ORES Se ee 30

New York v. Reebok Int'l Ltd.,

re I HII sctiiscenccccnicsnnnnenscnmmssenntets 16

TABLE OF AUTHORITIES—continued

Page(s)

Pearson v. NBTY, Inc.,

Be re BN vcccecncecccescccsscerecovesses 18

Powell vy. Georgia-Pacific Corp.,

119 F.3d 703 (8th Cir. 1997).................... 16, 17, 25

Six Mexican Workers v. Ariz. Citrus

Growers,

904 F.2d 1301 (9th Cir. 1990)............ 17, 20, 21, 25

Spokeo, Inc. v. Robins,

136 S. Ct. 1540 (2016).......................... 2, 14, 22, 23

Texas Dep't of Housing & Community Affairs

v. Inclusive Communities Project, Inc.,

136 S. Ct. 3507 (BO16)................2.2..00000000002 1, 17, 29

In re Zynga Privacy Litig.,

750 F.3d 1098 (9th Cir. 2014).................... 5, 11, 23

STATUTES, RULES AND REGULATIONS

ETT 6, 19, 23

ER EL a ne Oe I 24

GE 8 oan 27, 28

§ ef 8 nee en 24

bE 12

4 8 8 15

OTHER AUTHORITIES

Natalie Rodriguez, Era of Mammoth Cases

Tests Remedy of Last Resort, Law 360

wii

TABLE OF AUTHORITIES—continued

Page(s)

Principles of the Law of Aggregate Litigation

§ 3.07 cmt. b (Am. Law Inst. 2010) ............... 13, 31

BRIEF IN OPPOSITION OF

RESPONDENT GOOGLE LLC

Petitioners (and professional objectors) Ted

Frank and Melissa Holyoak seek this Court's review

based on a phony circuit split along with broad policy

objections to aspects of cy pres settlements not pre-

sent here.

Petitioners describe a conflict that does not exist.

The courts of appeals agree on the legal standards

governing review of cy pres settlements. They agree

that, as the court below acknowledged, cy pres set-

tlements are “the exception, not the rule.” Pet. App.

8. And they agree on the narrow circumstance that

justifies the exception: when direct distribution to

class members is infeasible

At bottom, then, the petition presents a case-

specific disagreement with the findings below that

distribution was infeasible in this case, with its class

estimated at 129 million members. Although this

Court “dofes] not overturn a finding of fact accepted

by two lower courts,” Texas Dep't of Housing &

Community Affairs v. Inclusive Communities Project,

Inc., 135 S. Ct. 2507, 2544 (2015), petitioners’ assert-

ed conflict rests on their insistence that it is feasible

to distribute a little over $5 million—an amount that

petitioners concede is adequate to resolve the low-

value claims in this case—to this enormous class.

And while they ask this Court to depart from its es-

tablished practice in order to overturn both lower

courts’ finding of infeasibility in this case, petitioners

cannot offer even a speculative example of a case

where distribution of a settlement fund would be in-

feasible under the legal standard uniformly prevail-

ing in the courts of appeals.

Petitioners’ restrictive approach would unduly

constrain district courts and settling parties while

providing no corresponding benefit to absent class

members. The exacting standards applied by the

courts of appeals to cy pres relief already protect

against the abuses petitioners inveigh against, while

appropriately permitting district courts to evaluate

case by case whether generalized relief inuring to the

entire class serves class interests more effectively

than token payments to a tiny self-selected minority

of class members.

Here, the cy pres payments will benefit the class

as a whole by funding closely targeted projects that

are directly connected to the Internet privacy issues

raised by plaintiffs’ claims. And because the recipient

institutions are of the highest quality and specialize

in Internet consumer issues, petitioners’ quibbles

with the district court’s exercise of discretion in ap-

proving the recipients do not warrant this Court’s at-

tention. No more does petitioners’ preference to pro-

vide nominal sums to a vanishingly small percentage

of class members who were never harmed.

The circumstances of this case make it an espe-

cially inappropriate vehicle for changing the law to

impose petitioners’ vision. This case is unusually

well-suited to a cy pres remedy because no class

member appears to have been actually harmed by

the challenged practices—as two district judges con-

cluded in dismissing most of plaintiffs’ claims. The

case would have gone forward, if at all, based on the

remote (and now-obsolete) prospect of obtaining a

statutory damages award despite the absence of con-

crete injury. When this case settled, Ninth Circuit

law permitted uninjured plaintiffs to pursue statuto-

ry damages, but this Court’s decision in Spokeo, Inc.

v. Robins, 136 S. Ct. 1540 (2016), has since eliminat-

ed that risk. It is telling that neither petitioners, nor

the usual suspects joining them as amici, engage

with the obscure and hypertechnical claims, and lack

of injury, in this case.'

As the standards of the courts of appeals reflect,

parties should have the flexibility to compromise

cases like this one that limp past the pleadings stage

yet still pose a risk of huge liability based on uncer-

tain substantive law at the time of settlement. The

effectively black-and-white rule that petitioners ask

this Court to consider would make these cases im-

possible to settle when the relatively low settlement

value of the claims would yield a de minimis pay-

ment per class member. Instead, petitioners would

impose on the parties and the judicial system the

burden of continued litigation and continued risk. A

perfect and friction-free system of class certification,

liability determination, and award distribution

might obviate the need for cy pres settlements, but

the uncertainty of our own, entirely human legal sys-

tem does not permit that rigid result.

' For example, amicus Cato Institute submitted nearly identi-

cal arguments in Frank v. Poertner, No. 15-765, cert. denied 136

S. Ct. 1453 (2016). The same amicus joined a coalition of state

attorneys general led by the Arizona Attorney General in sup-

port of the unsuccessful petition in Blackman v. Gascho, No. 16-

364, cert. denied 137 S. Ct. 1065 (2017). Although they received

notice of the settlement as the Class Action Fairness Act re-

quires, none of the state attorneys’ general now supporting the

petition objected to the fairness of the settlement in either court

below. See Pet. App. 51.

STATEMENT

Petitioners and their amici have almost nothing

to say about this case—including the multiple dis-

missals of plaintiffs’ complaints. Petitioners’ two

paragraphs (Pet. 8-9) omit the context of the settle-

ment. Yet that context illuminates many of the defi-

ciencies in the petition.”

A. Referrer headers and web history.

Respondent Google LLC operates a free Internet

search engine, with over one billion user-generated

search requests every day. SER 747, 749-750. When

users submit search terms on Google Search, within

a fraction of a second Google returns to the user a

list of relevant websites in a new web page, “the

search results page.” SER 749-750, 761-762.

To generate each search results page, Google's

servers first generate a unique Uniform Resource

Locator, or URL, that includes information about the

user's search query used to generate that page. SER

761-762. Users can then click one of the links provid-

ed on the search results page, which will redirect the

user to the desired site. SER 749-750.

In the normal course of operation, the user’s web

browser (e.g., Internet Explorer, Firefox, Safari)

transmits to that website what is known as “referrer

header” information. See SER 760-762. The referrer

header communicates the URL of the web page that

2 In light of the posture of the case at settlement, these facts

are drawn largely from the allegations in plaintiffs’ pleadings.

Google does not admit the accuracy of these allegations.

’ “ER _” refers to petitioners’ Excerpts of Record in the court

of appeals, and “SER _” refers to respondents’ Supplemental

Excerpts of Record.

the user last visited and by doing so informs the re-

quested website how the user got to the page. SER

760-761; see also generally In re Zynga Privacy

Litig., 750 F.3d 1098, 1102 (9th Cir. 2014) (explain-

ing how referrer headers work). So if a user clicks on

a website from Google’s search results page, the re-

ferrer header generated by the user's web browser

will communicate the URL of the search results

page, which, according to plaintiffs, includes the us-

er’s search terms. SER 760-761.

Plaintiffs have also raised allegations about

Google’s “Web History” service, which Google pro-

vides for its authenticated account holders. SER 712,

717-718. The “Web History” service tracks and stores

account holders’ web browsing activity on Google

servers. Ibid. This service allows account holders to

search the full content of pages they have viewed in

the past, and also allows those who so desire to per-

sonalize their web browsing experience by creating

recommendations tailored to their individual user

preferences. /bid. Unlike referrer headers, which are

transmitted to third-party website operators as a

standard and default web browser function, only the

account holder has access to his or her Web History.

Ibid.

B. The Gaos and Priyev actions.

The plaintiffs alleged that Google violated their

privacy rights when referrer headers were disclosed

to the owners of third-party websites on which a user

clicked from a Google search results page. The pur-

ported privacy violation resulted from the alleged in-

clusion of search terms in the URL of the search re-

sults page.

1. In 2010, respondent Paloma Gaos filed a puta-

tive class action complaint against Google in the

Northern District of California. See 5:10-cv-04809-

EJD (N.D. Cal.), Dkt. No. 1. Gaos asserted violations

of the Stored Communications Act of 1986 (“SCA”),

18 U.S.C. §§ 2701 et seq.)), as well as several Califor-

nia state-law claims. /bid. Judge Ware granted

Google’s motion to dismiss the initial complaint be-

cause Gaos had “failed to plead facts sufficient to es-

tablish Article III standing” for any of her claims.

SER 799-800.

Gaos’s First Amended Complaint raised largely

the same claims. Google again moved to dismiss, and

Judge Davila (to whom the case had been reas-

signed) dismissed all but Gaos’s SCA claim, which al-

leged only that the content of search queries, not

linked back to Gaos in any way, was disclosed in re-

ferrer headers.

Judge Davila reasoned, much as Judge Ware

had, that “Gaos does not identify what injury result-

ed” from the alleged dissemination of search queries

in referrer headers, and thus had failed to “allege[]

injury sufficient for Article III standing with respect

to her non-statutory causes of action.” SER 791-792.

Judge Davila concluded that Gaos nonetheless

had standing to pursue her SCA claim based on

then-current Ninth Circuit law holding that assert-

ing the bare violation of a statute was sufficient to

establish Article III standing “without additional in-

jury.” SER 792-793 (citing Edwards v. First Ameri-

can Corp., 610 F.3d 514 (9th Cir. 2010), cert. grant-

ed, 131 S. Ct. 3022 (2011), cert. dismissed, 132 S. Ct.

2536 (2012)). Judge Davi'a further cautioned that

“although Gaos ha[d] a.i-ged sufficient injury for

standing based on a viola :on of the SCA” in light of

Edwards, “this finding does not mean Gaos has

properly stated a claim for relief under the SCA.”

SER 794.

Gaos, joined by additional plaintiff Anthony

Italiano, filed a second amended complaint (SER

746-788) in an attempt to cure the defects identified

by Judge Davila. That complaint speculated that a

website operator or other unidentified third party

could identify someone based on their search terms

because Google’s users, including plaintiffs, some-

times conduct “vanity searches”—i.e., searches for

their own name. SER 776. They further speculated

that, although referrer headers do not specify a us-

ers name, a hypothetical “adversary” could use the

“Science of Reidentification” to “combine anonymized

data” from Referrer Headers “with outside infor-

mation to pry out obscured identities.” SER 770-773.

Google again moved to dismiss, arguing that the

plaintiffs still had failed to allege facts supporting

Article III standing for any of their claims or for re-

lief under them, and that the SCA preempted plain-

tiffs’ state-law claims (which were in Italiano’s name

only). That motion had been fully briefed at the time

of settlement, but was never decided.

2. In February 2012, respondent Gabriel Priyev

filed a similar action in the Northern District of Illi-

nois. See 12-CV-01467 (N.D. IIl.), Dkt. No. 1 (Feb. 29,

2012). Google moved to dismiss that complaint as

well, contending, as in Gaos, that Priyev lacked Arti-

cle III standing and failed to state a claim on which

relief could be granted.

Rather than defend the original complaint,

Priyev twice amended it. In the second amended

complaint, Priyev, again like the plaintiffs in Gaos,

alleged that Google violated the SCA and asserted

additional state-law claims. SER 708-745. He simi-

larly speculated that unidentified third parties might

be able to identify him from referrer headers because

he “regularly engaged in vanity searches.” SER 723.

Priyev further alleged that, by including search

terms in referrer headers, Google had disclosed ac-

count holders’ Web Histories to third parties one

search at a time, allegedly in violation of its privacy

policies. SER 712, 717-718.

Before Google responded to the second amended

complaint, the Priyev action was transferred to the

Northern District of California and consolidated with

Gaos. All of the plaintiffs then filed a consolidated

class action complaint (SER 660-707), seeking to rep-

resent “[ajll persons in the United States who sub-

mitted a search query to Google at any time between

October 25, 2006 and the date of notice to the class of

certification.” SER 697 (emphasis omitted).

C. Settlement and approval proceedings.

After an all-day mediation session, the parties

entered into a settlement agree. -nt and release in

March 2013. Pet. App. 69-111. The settlement class

tracks the class definition in the consolidated com-

plaint and is estimated to include approximately 129

million members. Pet. App. 5. Google agreed to pay

$8.5 million, which was to be put into a settlement

fund. After covering notice and administration costs,

$5,000 incentive awards for each of the three named

plaintiffs, and whatever amount of reasonable attor-

neys’ fees and costs the district court awarded, the

remainder (approximately $5.3 million) would be dis-

tributed proportionally among six cy pres recipients

to fund projects specifically devoted to Internet pri-

vacy issues: Carnegie-Mellon University; World Pri-

vacy Forum; Chicago Kent College of Law Center for

Information, Society, and Policy; Stanford Law

School Center for Internet and Society; Berkman

Center for Internet & Society at Harvard University;

and AARP Foundation. Pet. App. 5.

Plaintiffs linked to each of the recipients’ de-

tailed proposals for the use of the cy pres distribution

on the website created for class notice. SER 3, 56-58;

see also SER 257-381. (A seventh proposed recipient,

the MacArthur Foundation, withdrew from consider-

ation during the settlement process. ER 142 n.3.)

Upon reviewing the proposals, the district court

compared them “to an application for a grant.” ER

4A.

Google also agreed to make additional disclo-

sures on its website to better inform its users how re-

ferrer headers operate and to direct them to Google’s

privacy policies for more information on how Google

handles search queries generally. Pet. App. 40, 82.

The district court granted preliminary approval

of the settlement. ER 135-148. The court observed

that “[t]his case is somewhat unique in that the size

and nature of the class renders it nearly impossible

to determine exactly who may qualify as a class

member.” ER 146. “In fact,” it noted, “due to the pop-

ularity of Defendant's search engine and its rather

ubiquitous position in American culture, this class

potentially covers all internet users in the United

States.” Ibid. Under this circumstance, the court

found, “the cost of sending out what would likely be

very small payments to millions of class members

would exceed the total monetary benefit obtained by

the class.” ER 145.

Recognizing that direct notice to the settlement

class was not feasible, the district court approved the

parties’ proposed notice campaign—-which petitioners

did not challenge below. The notice campaign con-

sisted of (1) Internet-based notice using banner ads

in both English and Spanish; (2) articles in the press;

(3) a website dedicated solely to the settlement, with

both English and Spanish versions; and (4) a toll-free

telephone number where class members can obtain

additional information and request a written class

notice. Pet. App. 6, 39; see also SER 146-161 (decla-

ration of the claims administrator explaining the de-

sign of the notice campaign and its implementation).

Despite this extensive campaign, only 13 of the

129 million putative class members—about one in

ten million, or 0.00001%—opted out of the class, and

only five class members (including the two petition-

ers) entered written objections. ER 17, 98-134.

After an extensive fairness hearing (ER 29-97),

the district court granted final approval of the set-

tlement (Pet. App. 31-61), holding that petitioners

and the other objectors had not met their burden of

demonstrating that the settlement is not fair, ade-

quate, and reasonable. In particular, the court re-

jected, as inconsistent with Ninth Circuit precedent,

petitioners’ argument that a cy pres-only settlement

class should never be certified. Pet. App. 37-38 (cit-

ing Lane v. Facebook, Inc., 696 F.3d 811, 826 (9th

Cir. 2012); Nachshin v. AOL, LLC, 663 F.3d 1034

(9th Cir. 2011)). The district court additionally noted

that petitioners and the other objectors failed to

overcome class counsel’s showing that the “cost of

distributing [the] settlement fund to the class mem-

bers would be prohibitive” in light of the size of the

class and the administrative expense of direct distri-

bution. Pet. App. 58.

The district court also recognized “the very real

risk of never obtaining or losing class status in the

absence of settlement.” Pet. App. 45. And the court

characterized the settlement amount as more than

adequate in light of the “significant and potentially

case-ending weakness in the SCA claim brought

about by the Ninth Circuit’s decision in Zynga Priva-

cy Litigation.” Pet. App. 58-59.

Finally, the court rejected the objectors’ conten-

tion that the cy pres recipients were inappropriate

due to alleged conflicts of interest. Having reviewed

the recipients’ proposals (Pet. App. 48; see also ER

44), the court observed that the recipients “have a

record of promoting privacy protection on the Inter-

net, reach and target interests of all demographics

across the country, were willing to provide detailed

proposals, and are capable of using the funds to edu-

cate the class about online privacy risks.” Pet. App.

47-48.

D. The Ninth Circuit affirms.

The Ninth Circuit affirmed. Pet. App. 1-23. The

court first rejected petitioners’ argument that cy

pres-only settlements are categorically improper. It

recognized that “cy pres-only settlements are consid-

ered the exception, not the rule.” Pet. App. 8. That

exception is appropriate “where the settlement fund

is ‘non-distributable’ because ‘the proof of individual

claims would be burdensome or distribution of dam-

ages costly.” /bid. (quoting Lane, 696 F.3d at 819).

The Ninth Circuit upheld the district court's

finding that the settlement fund was not distributa-

ble. The court observed that the fund amount to be

distributed ($5.3 million) to the class of 129 million

individuals left a “paltry” sum of four cents per class

member. Pet. App. 9. In addition, the court approved

the district court’s finding that trying to compensate

even a nontrivial percentage of class members would

consume the settlement fund: “sending out very

small payments to millions of class members would

exceed the total monetary benefit obtained by the

class.” Ibid. (quoting the district court).

The court of appeals further rejected petitioners’

argument that the only permissible form of settle-

ment is one that instead compensates a “miniscule

portion of the class”—either through a “random lot-

tery distribution” or by offering “$5 or $10 per

claimant’ on the assumption that few class members

will make claims.” Pet. App. 9. The court recognized

that this would amount to a categoricai ban on cy

pres-only settlements that would be inconsistent

with its prior precedents. Pet. App. 10. And it noted

that, although an alternative settlement of that kind

may be technically “possible,” it was not required so

long as the actual settlement is fair, reasonable, and

adequate under Rule 23(e) and the existing stand-

ards courts have adopted to scrutinize class-action

settlements. Pet. App. 9-10.

The court then turned to the selection of the six

cy pres recipients. It chastised petitioners for their

“unfair and untrue” assertion that the district court

“rubber-stamped the settlement.” Pet. App. 12. In-

stead, the court of appeals held, the district court

correctly found that the six cy pres recipients satisfy

the ““nexus’ requirement by being tethered to the ob-

jectives of the underlying statute and the interests of

the silent class members.” Pet. App. 12-13. The court

noted the rigorous “selection process employed to vet

the cy pres recipients in this litigation,” and the dis-

trict court’s “careful review” of the “detailed pro-

posals” submitted by the recipients. Pet. App. 16.

And because the recipients met the exacting stand-

ards mandated by the court’s precedents, Google’s

previous donations to some of the recipients and the

fact that some of the recipients “are organizations

housed at class counsel’s alma maters” were not the

“absolute disqualifier[s]” urged by petitioners. Pet.

App. 12-16.

In other words, as the court put it in terms of pe-

titioners’ own proposed rule, petitioners had failed to

“raise substantial questions about whether the se-

lection of the recipient[s] was made on the merits.”

Pet. App. 14 (quoting Principles of the Law of Aggre-

gate Litigation § 3.07 cmt. b (Am. Law Inst. 2010)).

The court observed, for instance, that “some of the

recipient organizations have challenged Google’s In-

ternet privacy policies in the past,” and that “[eJach

recipient's cy pres proposal identified the scope of

Google’s previous contributions to that organization,

and * * * explained how the cy pres funds were dis-

tinct from Google’s general donations.” Pet. App. 16-

17 & n.7. Likewise, the court concluded that because

“(t]he recipients are well-recognized centers focused

ou the Internet and data privacy,” petitioners’ argu-

ment that the settlement should be invalidated

based on the mere fact that class counsel have de-

grees from some of the institutions that house the re-

cipients “can’t be entertained with a straight face.”

Pet. App. 19.

Judge Wallace concurred in part and dissented in

part. Pet. App. 23-30. He “agree[d]” with the majority

“that a cy pres-only settlement was appropriate in

this case,” Pet. App. 23, and “express{[ed] no opinion

on the definitive fairness” of the settlement, Pet.

App. 30. Instead, he would have remanded to the dis-

trict court for further fact-finding about the selection

of the specific cy pres recipients. Pet. App. 24, 30.

The Ninth Circuit denied rehearing. Pet. App.

67.

REASONS FOR DENYING THE PETITION

The petition seeks review of a case-specific exer-

cise of discretion applying settled legal standards to

unique facts and circumstances. Further review is

unwarranted for several reasons.

First, there is no circuit conflict. All circuits

agree with the Ninth Circuit that a cy pres-only set-

tlement is appropriate in the rare circumstance

where direct distribution to class members is infea-

sible, and all insist on a close nexus between the cy

pres remedy and the interests of settling class mem-

bers. Both the district court and the Ninth Circuit

followed these established standards in carefully

scrutinizing the settlement here.

Second, petitioners and their amici are also flat

wrong that cy pres-only settlements are on the rise;

their own source confirms that, since Chief Justice

Roberts’ statement respecting the denial of certiorari

in Marek v. Lane, 134 S. Ct. 8 (2013), the number of

cy pres-only settlements has dropped dramatically.

That trend is all the more likely to continue in light

of this Court’s decision in Spokeo, which requires

courts to weed out at the pleadings stage the kind of

no-injury claims most likely to lead to proposed cy

pres-only settlements.

Third, this case is a poor vehicle for considera-

tion of the new and rigid limits on cy pres settle-

ments urged by petitioners. This case is especially

well-suited to a cy pres remedy because class mem-

bers have not alleged any actual harm from the chal-

lenged practice—making any direct payment, how-

ever modest, a windfall rather than compensation.

And none of petitioners’ concerns about cy pres abus-

es is actually presented here—which is not surpris-

ing because the circuits now uniformly demand that

district courts safeguard against abuse.

Finally, the decision below was correct. Petition-

ers do not dispute that the settlement amount was

more than adequate, and they have no real challenge

to the findings of both courts below that individual

distributions of the $5 million settlement amount to

a non trivial portion of the 129-million-member class

would be infeasible. Petitioners also cannot deny

that the district court thoroughly evaluated the de-

tailed proposals submitted by the cy pres recipients

and ensured that the cy pres distribution would be

spent on privacy and security issues closely tied to

the class claims. Given these circumstances, as well

as “the shakiness of the plaintiffs’ claims” (Pet. App.

9), the district court acted well within its discretion

in finding the settlement “fair, reasonable, and ade-

quate,” Fed. R. Civ. P. 23(e)(2), and the Ninth Circuit

was correct to affirm that exercise of discretion.

A. The Purported Conflict Does Not Exist.

1. While petitioners rail against cy pres remedies,

every circuit to address the issue has held that cy

pres remedies may be appropriate in class action set-

tlements. That’s why, in one of the cases petitioners

cite the most, the Third Circuit said that it “joinjed]

other courts of appeals in holding that a district

court does not abuse its discretion by approving a

class action settlement agreement that includes a cy

pres component.” In re Baby Products Antitrust

Litig., 708 F.3d 163, 172 (3d Cir. 2013).

In particular, the courts of appeals have allowed

cy pres remedies where a cash distribution of all or

part of a settlement fund would not be feasible—

either because class members cannot be identified, or

because the amounts at issue would produce a negli-

gible payment per class member (or be entirely con-

sumed by the administrative costs of distribution). In

those rare circumstances, the circuits uniformly have

held that the parties may direct the distribution to a

charity aligned with the interests of the settlement

class. See, e.g., Baby Products, 708 F.3d at 173 (not-

ing general agreement that “cy pres distributions are

most appropriate where further individual distribu-

tions are economically infeasible”); Klier v. Elf

Autochem N. Am., Inc., 658 F.3d 468, 475 & n.15 (5th

Cir. 2011) (cy pres awards are appropriate “when di-

rect distributions to class members are not feasible”);

Mirfasihi y. Fleet Mortg. Corp., 356 F.3d 781, 784

(7th Cir. 2004) (rejecting cy pres distribution where

“there would be no reason for thinking distribution to

the class members infeasible”); Masters v. Wilhelmi-

na Model Agency, Inc., 473 F.3d 423, 436 (2d Cir.

2007) (cy pres remedy available in “circumstances in

which direct distribution to individual class members

is not economically feasible”); Powell v. Georgia-

Pacific Corp., 119 F.3d 703, 706 (8th Cir. 1997) (a “cy

pres distribution is potentially appropriate” where “it

would be extremely difficult to distribute the funds

pro rata”); New York v. Reebok Int'l Lid., 96 F.3d 44,

49 (2d Cir. 1996) (approving cy pres distribution

where the “impracticality of attempting to distribute

the settlement proceeds among the multitude of uni-

dentifiea possible claimants is obvious” and “distri-

bution ‘would be consumed in the costs of its own

administration”).

This uniformity reflects a consistent rationale.

Rather than distribute negligible amounts to class

members (or, as petitioners suggest, leaving almost

all class members with nothing at all), the circuits

reason that settlement funds are better used for pro-

jects that advance the interests of the class as a

whole. See Baby Products, 708 F.3d at 172-73; Pow-

ell, 119 F.3d at 706; Six Mexican Workers v. Ariz.

Citrus Growers, 904 F.2d 1301, 1305, 1307 (9th Cir.

1990). No court has joined petitioners in asking that

federal class actions be turned into lottery where a

tiny percentage of class members, either self-selected

or selected by the court, receive payments that are

not tied to any injury.

Petitioners try to fracture this uniformity by at-

tacking a caricature of the decision below. Petition-

ers read the decision as adopting a new rule that a cy

pres-only settlement is permissible whenever “it

would be ‘infeasible’ to make payments to every sin-

gle class member.” Pet. 2; accord id. at 17. But the

Ninth Circuit said no such thing. It pointed out the

“4 cents ir recovery” per class member to highlight

the “dramatic” “gap” between the settlement fund

and the size of the class. Pet. App. 9. And it upheld

the district court’s finding “that the cost of verifying

and ‘sending out very small payments to millions of

class members would exceed the total monetary ben-

efit obtained by the class.” /bid. Petitioners have no

answer to that dispositive “finding of fact accepted by

two lower courts.” Texas Dep't of Housing, 135 S. Ct.

at 2544. That imbalance of costs and benefits re-

mains even if the “millions of class members” receiv-

ing payments amount to less (even far less) than the

full 129 million.

2. Sweeping away petitioners’ mischaracteriza-

tion of the decision below makes clear that the sup-

posed conflict reflects only the application of the

same legal standards to different facts.

To begin with, Judge Posner’s opinion in Pearson

v. NBTY, Inc., 772 F.3d 778 (7th Cir. 2014), adopted

the same legal standard used below, recognizing that

a “cy pres award is supposed to be limited to money

that can’t feasibly be awarded to the *** class

members.” Jd. at 784. The class in Pearson had only

4.72 million members (less than 4% of the class

here)—notified at a cost of about 30 cents per person.

Ibid. The much larger settlement amount in Pearson

allowed all members of the much smaller class to be

paid directly ($3 each by the court’s calculation, less

the incremental distribution expenses of processing

the payments). /bid. Thus, the parties “hald] not

* * * demonstrated” that it was “infeasible to provide

that compensation to the victims.” /bid.

Here, by contrast, both courts below found that it

would be infeasible to pay directly a “unique[ly]”

enormous class that “potentially covers all internet

users in the United States.” ER 146; see also Pet.

App. 8-9, 46-47. And, in furthur contrast with Pear-

son, where there was no dispute that class members

had purchased the product at issue at least poten-

tially in reliance on the allegedly misleading labels,

here no out-of-pocket injury is alleged. Indeed, as the

district court held, plaintiffs had not alleged any con-

crete injury at all. See SER 791-792.

There also is no inconsistency between the deci-

sion below and the Eighth Circuit’s decision in /n re

BankAmerica Corp. Securities Litigation, 775 F.3d

1060 (8th Cir. 2015). The Eighth Circuit expressly

“agree[d] with the Ninth Circuit” that cy pres awards

are permissible so long as they provide an “indirect

class benefit” by going to “uses consistent with the

nature of the underlying action.” /d. at 1066-67 (em-

phasis added; quotation marks omitted) (citing

Nachshin, 663 F.3d at 1040). The court of appeals re-

jected the cy pres award proposed in BankAmerica

because the single recipient—a Missouri legal-

services organization—had an insufficient nexus to

the securities fraud claims asserted by the nation-

wide class. See id. at 1067. No similar objections

have been or could be raised against the targeted

projects proposed by the cy pres recipients here,

whose Internet privacy activities have nationwide

scope and effects.

The Fifth Circuit’s decision in Klier has nothing

to do with the decision below. Indeed, the court in

‘ Petitioners also rely (Pet. 18 n.4) on the Seventh Circuit's

2004 decision in Mirfasihi v. Fleet Mortgage Corp., 356 F.3d 781

(7th Cir. 2004). But they fail to mention that the Seventh Cir-

cuit ultimately affirmed a cy pres distribution in that case on a

rationale similar to the lower courts’ here: that the underlying

claims were of dubious value and presented significant litiga-

tion risks for the plaintiffs. Mirfasihi v. Fleet Mortg. Corp., 551

F.3d 682, 686 (7th Cir. 2008). Just as the courts here ques-

tioned the applicability of the Stored Communications Act to

plaintiffs’ claims, see Pet. App. 58-59, the Seventh Circuit found

the inapplicability of the federal Fair Credit Reporting Act to

weigh in favor of the cy pres settlement. See 551 F.3d at 685-86.

that case explained that its decision did not “impli-

cate the line of authority giving careful scrutiny to

class settlement agreements in which the parties

agree to a cy pres distribution.” 658 F.3d at 478 n.29

(citing, inter alia, Six Mexican Workers, 904 F.2d at

1304, 1307). Rather, the district court in Klier sua

sponte ordered a cy pres distribution of unclaimed

settlement funds in direct violation of the settlement

agreement’s explicit direction that any leftover funds

in a subclass fund “shall be distributed pro rata to

all Claimants in that subclass.” Jd. at 476. That is,

Klier involved a district court’s injection of cy pres in

an attempt to override a settlement rather than ap-

prove or enforce it. Jd. at 476-77.

The decision below does not conflict with the

Third Circuit’s Baby Products decision; nor are the

two decisions “in tension,” as petitioners half-

heartedly assert in the alternative. Pet. 19. Baby

Products approvingly cited the Ninth Circuit's deci-

sion in Lane in “join[ing] other courts of appeals in

holding that a district court does not abuse its discre-

tion by approving a class action settlement agree-

ment that includes a cy pres component.” 708 F.3d at

172. Although the court noted that “cy pres distribu-

tions are most appropriate where further individual

distributions [to class members] are economically in-

feasible,” it declined to “hold that cy pres distribu-

tions are only appropriate in this context.” /d. at 173.

Instead, cy pres relief is permissible when it is “fair,

reasonable, and adequate” under “the same frame-

work developed for assessing other aspects of class

action settlements.” /d. at 174.

Finally, there is no “tension” between the deci-

sion below and the Second Circuit's decision in Mas-

ters. Pet. 20. Masters cited with approval the Ninth

Circuit’s infeasibility standard from Six Mexican

Workers, but faulted the district court for directing

funds to charities without even considering whether

a distribution to the class would be possible and

when “neither side contend[ed] that * * * individual

distribution [was] economically impracticable.” 473

F.3d at 435-36.

As these cases illustrate, there is no conflict

among the circuits over the standards for judicial re-

view of cy pres settlements. Courts applying these

standards properly reach different outcomes based

on different facts and circumstances. These outcomes

present no conflict on any important question of fed-

eral law that could warrant this Court’s review.

B. The Purported Issue Lacks Recurrent

Importance Because Cy Pres-Only Set-

tlements Are Sharply Declining And Re-

served For Exceptional Cases.

There is a complete mismatch between petition-

ers’ assertion that “cy pres settlements are increas-

ingly prevalent” and the question presented, which

asks this Court to review cy pres-only settlements—

i.e., ones that “provide[] no direct relief to class

members.” Pet. i, 31-32.

As explained in an article petitioners cite, the

number of cy pres-only settlements is in fact sharply

declining, with only 3 percent of all cy pres settle-

ments falling into that category from 2013 through

2016—down from 29 percent from 2000 through

2012. See Natalie Rodriguez, Era of Mammoth Cases

Tests Remedy of Last Resort, Law 360 (May 2, 2017),

https://www.law360.com/articles/9 18296/era-of-

mammoth-cases-tests-remedy-of-last-resort (cited at

Pet. 32; State AG Br. 7-8). As the article explains,

that drop is consistent with the view that Chief Jus-

tice Roberts’s statement respecting the denial of re-

view in Lane in November 2013 “mark[ed] a turning

point for cy pres.” /bid. Since that time, the article

summarizes, class action settlements have “mostly

use[d] cy pres to get rid of so-called drippings—

unclaimed funds and uncashed checks sent to class

members.” [bid.®

Moreover, while 2016 was the most recent year

covered in the analysis, the decline in the number of

cy pres-only settlements is likely to continue further

still in light of this Court’s 2016 decision in Spokeo.

Spokeo clarifies that Article II] requires lower feder-

al courts to weed out putative class actions seeking

statutory damages in the absence of actual harm.

These are the cases that are most likely to result in

proposed cy pres-only settlements if they make it

past the pleadings stage; settlements are naturally

smaller when the presence of any injury is unlikely.

In short, empirical data confirm the Ninth Cir-

cuit’s observation below that “cy pres-only settle-

ments are considered the exception, not the rule.”

Pet. App. 8. Petitioners offer no compelling reason

why this Court should review a perceived problem

that is neither real (see pages 24-28, infra) nor wide-

spread.

® The article also notes the positive development of parties and

courts using a “grant-like approach” to thoroughly vet potential

cy pres recipients—precisely the approach followed in this case.

See page 9, supra.

C. This Case Is An Exceedingly Poor Vehi-

cle To Consider Unprecedented Limita-

tions On Cy Pres Settlements.

1. Class members were unharmed and un-

likely to receive anything from further lit-

igation.

The petition again and again refers to class

members who have not been “compensated” for their

“injuries.” E.g., Pet. 2-3, 7, 19-20, 23, 25. The under-

lying premise is that a class is undercompensated

unless some class members receive a direct distribu-

tion—no matter how few class members would be

paid or how little they would receive.

But the circumstances in this case provide no op-

portunity to test that premise. That is because plain-

tiffs failed to allege any actual harm that they or

class members suffered as a result of the challenged

practices—which led both Judge Ware and Judge

Davila to dismiss most of plaintiffs’ claims for lack of

“injury sufficient for Article III standing.” SER 792;