Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 1 of 21
`
`
`
`
`
`
`
`
`
`
`
`UNITED STATES DISTRICT COURT
`
`NORTHERN DISTRICT OF CALIFORNIA
`
`IN RE FACEBOOK BIOMETRIC
`INFORMATION PRIVACY LITIGATION
`
`Case No. 15-cv-03747-JD
`
`
`ORDER RE FINAL APPROVAL,
`ATTORNEYS’ FEES AND COSTS,
`AND INCENTIVE AWARDS
`
`Re: Dkt. Nos. 499, 517
`
`
`
`
`
`By any measure, the $650 million settlement in this biometric privacy class action is a
`
`landmark result. It is one the largest settlements ever for a privacy violation, and it will put at
`
`least $345 into the hands of every class member interested in being compensated. At the Court’s
`
`request, the parties jointly developed an innovative notice and claims procedure that generated an
`
`impressive claims rate. The settlement attracted widespread support from the class, and drew only
`
`three objections out of millions of class members.
`
`All of this was achieved in the context of a new and untested statute, the Illinois Biometric
`
`Information Privacy Act (BIPA), 740 Ill. Comp. Stat. 14/1 et seq. (2008). The case raised several
`
`complicated and intensely litigated issues, including the question of whether a statutory privacy
`
`injury was sufficiently “real” and concrete to establish an injury in fact for standing under
`
`Article III and the BIPA. The Court determined that it was, and two other courts -- the Illinois
`
`Supreme Court and the Ninth Circuit -- reached the same conclusion. The standing issue makes
`
`this settlement all the more valuable because Facebook and other big tech companies continue to
`
`fight the proposition that a statutory privacy violation is a genuine harm. In a pending Supreme
`
`Court case about the Fair Credit Reporting Act, which is unrelated to this action, Facebook,
`
`Google, and eBay filed an amicus brief that points to this settlement and asks the Court to reverse
`
`the “mistaken holding” that a statutory privacy violation is an actual injury. See Brief for Amici
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 2 of 21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Curiae eBay, Inc. et al. Supporting Petitioner at 4, Trans Union LLC v. Ramirez, No. 20-297 (U.S.
`
`Feb. 8, 2021). This underscores the considerable legal risks both sides would have faced had this
`
`case continued on.
`
`Overall, the settlement is a major win for consumers in the hotly contested area of digital
`
`privacy. Final approval of the class action settlement is granted. Attorneys’ fees and costs, and
`
`incentive awards to the named plaintiffs, are also granted, although in lesser amounts than
`
`requested.
`
`BACKGROUND
`
`This case has traveled a long and winding road since it was filed in 2015, and the Court
`
`focuses here on the key events most pertinent to final approval and the fees award. A heavily
`
`litigated class certification motion was resolved in favor of certification of a class of Facebook
`
`users located in Illinois for whom Facebook created and stored a face template after June 7, 2011.
`
`Dkt. No. 333 at 15. The certified class, which is also the class for the settlement, Dkt. No. 468,
`
`Ex. A at 5-6 (¶ 1.7), challenged Facebook’s “Tag Suggestions” program, which looks for and
`
`identifies people’s faces in photographs uploaded to Facebook to promote user tagging. The class
`
`members alleged that Facebook collected and stored their biometric data -- namely digital scans of
`
`their faces -- without prior notice or consent, thereby violating Sections 15(a) and 15(b) of the
`
`BIPA, 740 Ill. Comp. Stat. 14/15(a)-(b).
`
`After substantial briefing and argument, the Court concluded that a violation of the BIPA
`
`was enough to establish an injury in fact for purposes of Article III standing. See Dkt. No. 294.
`
`This was not the only serious disagreement between the parties in the early stage of the case.
`
`A significant battle was also fought over a choice-of-law provision in Facebook’s user agreements,
`
`and whether California law applied to the exclusion of a claim under the BIPA, as Facebook
`
`argued. The Court held an evidentiary hearing to help resolve that dispute, and concluded that
`
`Illinois law applied. See Dkt. No. 120. But the injury-in-fact issue stands out both because it was
`
`a threshold issue for the litigation as a whole, and because it became a central issue in Facebook’s
`
`discretionary appeal to the Ninth Circuit of the Court’s certification of the class. See Dkt.
`
`Nos. 406, 416. The circuit court affirmed the Court’s conclusions about standing and certification.
`
`2
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 3 of 21
`
`
`
`Dkt. No. 416. The Supreme Court denied Facebook’s petition for certiorari. Dkt. No. 426. While
`
`these events were unfolding in federal court, the Illinois Supreme Court published a decision in
`
`another BIPA case that cited and adopted the Court’s construction of the statute on the question of
`
`who counts as an “aggrieved” person with a right of action under the statute. See Rosenbach v. Six
`
`Flags Entm’t Corp., 129 N.E.3d 1197 (Ill. 2019).
`
`The settlement of the case was its own long chapter in this litigation. Three days before a
`
`trial setting conference, the parties advised that a settlement in principle had been reached. Dkt.
`
`No. 427. The Court denied preliminary approval of the initial settlement proposal over concerns
`
`about the discount on statutory damages under the BIPA, a conduct remedy that did not appear to
`
`require meaningful changes by Facebook, over-broad releases by the class, and the form of notice
`
`to class members. Dkt. No. 456.
`
`The parties responded substantively to the Court’s concerns. After another round of
`
`negotiations, they revised the proposed settlement in several significant ways. To start, Facebook
`
`agreed to increase its settlement payment by $100 million, which brought the total amount of the
`
`non-reversionary common fund to $650 million. The settlement administration expenses, taxes,
`
`and class representative incentive awards and attorney’s fees and costs will be paid from this fund,
`
`and the balance will be distributed on a pro rata basis to each class member who submits a valid
`
`claim. Dkt. No. 468, Ex. A at 10 (¶ 1.31). The parties also clarified the changes that Facebook
`
`will implement. For all users who have not affirmatively opted in or consented to biometric scans,
`
`Facebook will set its “Face Recognition” default user setting to “off,” and it will delete all existing
`
`and stored face templates for class members unless Facebook obtains a class member’s express
`
`consent after a separate disclosure about how Facebook will use the face templates. Id. at 13-14
`
`(¶ 2.9). It will also delete the face templates of any class members who have had no activity on
`
`Facebook for three years. Id. at 14 (¶ 2.9(c)). At the preliminary approval hearing, Facebook’s
`
`product manager for the Face Recognition program stated under oath that this change to the Face
`
`Recognition default setting to “opt in” would be not just in Illinois or the United States, but
`
`“global.” Dkt. No. 470 at 10:11-17.
`
`3
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 4 of 21
`
`
`
`For the notice and claims procedures, the parties maximized outreach to class members by
`
`leveraging Facebook’s direct access to users and using other internet channels. Facebook advised
`
`the Court that it would deploy “the most aggressive methods that we use to reach users in the ways
`
`that we best can at Facebook,” and the Court further encouraged the parties to optimize outreach
`
`by consulting a behavioral economist or psychologist to provide expert advice on techniques. Dkt.
`
`No. 470 at 19:19-21; 25:13-25. The claim form was easy to complete online or by mail, and
`
`required only basic information about class members’ Facebook accounts, and in the event their
`
`email address was not found in Facebook’s records, the class member’s “approximate dates and
`
`addresses where you lived in Illinois” for at least 6 months during the class period. See Dkt.
`
`No. 468, Ex. A (Ex. A). Class members had their choice of four payment options: paper check,
`
`Zelle, PayPal, or direct deposit. Id. These measures generated a claims rate of approximately
`
`22%, a result that vastly exceeds the rate of 4-9% that is typical for consumer class actions. See
`
`Fed. Trade Comm’n, Consumers and Class Actions: A Retrospective and Analysis of Settlement
`
`Campaigns 11 (2019) (“Across all cases in our sample requiring a claims process, the median
`
`calculated claims rate was 9%, and the weighted mean (i.e., cases weighted by the number of
`
`notice recipients) was 4%.”).
`
`The scope of the release was revised to limit it to the boundaries of this litigation. Class
`
`members who did not opt out will release Facebook, Inc., Face.com, and their associated entities,
`
`from known or unknown claims “arising from or related to plaintiffs’ allegations regarding the
`
`alleged collection, storage, or dissemination of biometric data related to facial recognition
`
`technology from Facebook users located in Illinois, including all claims and issues that were
`
`litigated in the action or that could have been brought in the Action and claims for any violation of
`
`the Illinois Biometric Information Privacy Act (‘BIPA’) or other Illinois statutory or common law
`
`related to facial recognition.” Dkt. No. 468, Ex. A at 8 (¶¶ 1.25, 1.26). Put more simply, the class
`
`is releasing only the claims that were or could have been litigated here. In addition, in response to
`
`a concern on the Court’s part, the release now excludes Facebook’s affiliated entities Instagram,
`
`Inc., WhatsApp Inc., and Oculus VR Inc., which were included originally even though they did
`
`4
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 5 of 21
`
`
`
`not use the Tag Suggestions feature. See id. at 8 (¶ 1.26). Claims may be pursued against those
`
`companies under the BIPA irrespective of this settlement.
`
`Based on these and other substantive revisions, the Court granted preliminary approval of
`
`the settlement. Dkt. No. 474. An unusual development ensued. A law firm that hitherto had
`
`played no role whatsoever in this case posted a misleading online solicitation about the settlement.
`
`See Dkt. No. 477. The post deceptively invited people to “sign up for a claim” when the law firm
`
`actually was trying to recruit opt-outs. The parties brought the problem to the Court’s attention,
`
`and the Court held a hearing on the matter at which the law firm appeared and participated. Dkt.
`
`No. 486. The law firm agreed to take down the posts and to send a curative notice to
`
`approximately 3,724 class members who had responded to the misleading solicitation. Dkt.
`
`No. 494. This appears to have fully corrected the situation, and there is no evidence of any
`
`residual issues.
`
`As the notice and claims procedures unfolded, the parties kept the Court apprised of
`
`progress with regular status reports. See Dkt. Nos. 475, 476, 492, 496, 501, 503, and 506. The
`
`claims period ended on November 23, 2020.
`
`This order resolves plaintiffs’ motion for final approval, and motion for fees, expenses, and
`
`costs. Dkt. Nos. 499, 517. It also addresses and overrules three objections filed by four objectors.
`
`See Dkt. Nos. 497 (objection of Kevin C. Williams), 504 (objection of Dawn Frankfother and
`
`Cathy Flanagan), 519 (objection of Kara Ross). Objectors Frankfother and Flanagan also filed an
`
`opposition to plaintiffs’ motion for final approval. Dkt. No. 518.
`
`The Court has carefully reviewed all of the voluminous filings with respect to final
`
`approval and fees, which include expert declarations, letters from class members, and other
`
`materials. On January 14, 2021, the Court held a lengthy final approval hearing. Dkt. No. 530.
`
`The hearing featured arguments by the parties, the objectors Frankfother, Flanagan, and Ross, who
`
`appeared through their counsel, as well as remarks by the class’s expert, Professor William B.
`
`Rubenstein of Harvard Law School. Id.
`
`5
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 6 of 21
`
`
`
`DISCUSSION
`
`I.
`
`FINAL APPROVAL OF CLASS ACTION SETTLEMENT
`
`A.
`
`Legal Standards and Choice of Law
`
`Pursuant to Rule 23(e) of the Federal Rules of Civil Procedure, the claims of a certified
`
`class may be settled only with the Court’s approval. Rule 23(e) outlines the procedures that apply
`
`to the proposed class settlement, including the requirement to direct notice in a reasonable manner
`
`to all class members who would be bound by the proposal. Fed. R. Civ. P. 23(e)(1).
`
`Under Rule 23(e)(2), the Court may approve a proposal that would bind class members
`
`“only after a hearing and only on finding that it is fair, reasonable, and adequate after considering
`
`whether:
`
`(A) the class representatives and class counsel have adequately represented the
`class;
`
`(B) the proposal was negotiated at arm’s length;
`
`(C) the relief provided for the class is adequate, taking into account:
`
`(i) the costs, risks, and delay of trial and appeal;
`
`(ii) the effectiveness of any proposed method of distributing relief to the class,
`including the method of processing class-member claims;
`
`(iii) the terms of any proposed award of attorney’s fees, including timing of
`payment; and
`
`(iv) any agreement required to be identified under Rule 23(e)(3); and
`
`(D) the proposal treats class members equitably relative to each other.”
`
`Fed. R. Civ. P. 23(e)(2).1
`
`In addition, our circuit has held that “[t]he factors in a court’s fairness assessment will
`
`naturally vary from case to case, but courts generally must weigh: (1) the strength of the plaintiff’s
`
`case; (2) the risk, expense, complexity, and likely duration of further litigation; (3) the risk of
`
`maintaining class action status throughout the trial; (4) the amount offered in settlement; (5) the
`
`extent of discovery completed and the stage of the proceedings; (6) the experience and views of
`
`
`1 This amended version of Rule 23(e)(2) went into effect on December 1, 2018. The Court finds it
`just and practicable to apply it to this case, which was filed in 2015, especially given that plaintiffs
`referenced this version of the Rule in their final approval motion, Dkt. No. 517 at 8, and no party
`or class member has expressed any concerns about it.
`6
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 7 of 21
`
`
`
`counsel; (7) the presence of a governmental participant; and (8) the reaction of the class members
`
`of the proposed settlement.” In re Bluetooth Headset Products Liability Litig., 654 F.3d 935, 946
`
`(9th Cir. 2011) (quoting Churchill Vill., L.L.C. v. Gen. Elec., 361 F.3d 566, 575 (9th Cir. 2004)).
`
`Although this case was transferred here from the Northern District of Illinois and involved
`
`an Illinois state statute, the Court follows the Ninth Circuit on final approval of the settlement, and
`
`not the Seventh Circuit, as objectors Frankfother and Flanagan suggest. See Dkt. No. 518 at 4-6.
`
`Final approval under Federal Rule of Civil Procedure Rule 23 is a federal issue governed by the
`
`precedent in the transferee court. Newton v. Thomason, 22 F.3d 1455, 1460 (9th Cir. 1994).
`
`B.
`
`The Rule 23(e) Factors
`
`1.
`
`Notice
`
`In the course of denying the original proposed settlement, the Court challenged the parties
`
`to come up with forms of notice that fit the reality of our online lives. The old methods of U.S.
`
`Mail and a print ad were not going to cut it in 2020. The parties rose to the occasion and proposed
`
`an array of innovative ways to reach class members. These included notice via direct email,
`
`Facebook’s “jewel” and newsfeed notifications, publication in the leading Illinois newspapers, a
`
`dedicated settlement website, and an internet ad campaign on non-Facebook platforms. See Dkt.
`
`No. 474 at 6; Dkt. No. 468, Ex. A at 15 (¶ 4.2).
`
`These were robust measures, and they paid off in spades. The settlement administrator,
`
`Gilardi & Co., sent initial notices by email to 15,372,960 addresses, which were associated with
`
`12,340,049 accounts. Dkt. No. 517-1 ¶ 7. In this initial batch, 10,295,502 notices were
`
`successfully delivered to at least one of the email addresses linked to an account. Id. ¶ 9. When
`
`counsel discovered around 5.7 million emails on Gmail had been routed to users’ spam folders,
`
`they worked directly with Google to fix the problem, which resulted in a 99.9% delivery success
`
`rate outside the spam folders. Dkt. No. 517 at 4; Dkt. No. 492; Dkt. No. 517-1 ¶¶ 10-11. A
`
`reminder email notice was also sent to 12,888,208 addresses in the class list, of which 9,956,299
`
`were successfully delivered. Dkt. No. 517-1 ¶¶ 12-13.
`
`A summary notice approved by the Court was published in the September 23, 2020
`
`editions of the Chicago Tribune and Chicago Sun-Times. Dkt. No. 517-1 ¶ 15. Almost 28 million
`
`7
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 8 of 21
`
`
`
`impressions advertising the settlement were delivered through the Google Display Network. Id.
`
`¶ 16. A website was established at www.facebookbipaclassaction.com, which, as of December 1,
`
`2020, had received 6,230,922 visits. Id. ¶ 17. A toll-free telephone number was also set up, which
`
`logged a total of 5,063 calls as of the same date. Id. ¶ 18.
`
`For the jewel and newsfeed notices on the Facebook platform, a Facebook software
`
`engineer submitted a declaration stating that the total number of impressions to unique users
`
`delivered as of December 3, 2020, was 9,124,351. Dkt. No. 512 ¶ 2. The combined click-through
`
`rate for the notices was 30.47%. Id. ¶ 3.
`
`Overall, this was a thorough, effective, and far-reaching program for giving notice to the
`
`class, which even one of the objectors has acknowledged. Dkt. No. 532 at 25:6-8 (“I think these
`
`guys did a significantly better job than in most cases of actually getting notice out.”). Objector
`
`Kara Ross’s different views about the adequacy of notice, see Dkt. No. 519 at ECF p. 2, have been
`
`withdrawn and would have been overruled for these reasons.2 The Court concludes that the notice
`
`was extremely thorough and diligently given in a reasonable manner to all class members who
`
`would be bound by the proposal. Fed. R. Civ. P. 23(e)(1).3
`
`The Court appreciates that counsel acted on its suggestion at the preliminary approval
`
`hearing and consulted with Professor Dan Ariely, professor of psychology and behavioral
`
`economics at Duke University, “so as to maximize the likelihood that class members would file
`
`claims.” Dkt. No. 499-1 ¶ 134. Based on Professor Ariely’s advice about the “no-action bias,”
`
`which is the “principle that people generally prefer to do nothing over something,” class counsel
`
`and Facebook “agreed to change the claim form flow so that class members who try to leave the
`
`website without submitting a claim (i.e., ‘do nothing’) can’t proceed without clicking a button
`
`indicating their understanding that their share of the settlement would be distributed pro rata to
`
`other class members (i.e., ‘do something’).” Id.
`
`
`2 Class counsel has represented that objector Ross withdrew her objection as to the sufficiency of
`notice. See Dkt. No. 517 at 6.
`
` 3
`
` Gilardi & Co. also delivered notices to state and federal officials in compliance with the Class
`Action Fairness Act, 28 U.S.C. § 1715. See Dkt. No. 517-1 ¶¶ 2-6.
`
`
`8
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 9 of 21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`2.
`
`Adequacy of Class Representatives and Class Counsel
`
`As stated in the class certification and preliminary approval orders, there is no serious
`
`question about the adequacy of the class representatives and class counsel. Dkt. No. 474 at 4. The
`
`Court’s independent review validated both points. Consequently, Nimesh Patel, Adam Pezen, and
`
`Carlo Licata were confirmed as class representatives, and Edelson PC, Robbins Geller Rudman &
`
`Dowd LLP, and Labaton Sucharow LLP were confirmed as class counsel.
`
`Nothing has changed on either score. If anything, more evidence has been adduced in
`
`favor of those appointments. The Court has taken into account the comments at the final approval
`
`hearing by Facebook’s lead attorney, who complimented the “willingness of very experienced
`
`counsel . . . to negotiate with me and my team in a way where we were able to look honestly
`
`collectively at the issues.” Dkt. No. 532 at 65:14-17. He also said that class counsel should be
`
`“applauded” and that they had a “skill level, just in the negotiation, itself.” Id. at 65:20-66:1.
`
`These and other facts in the record speak well of the professionalism both sides brought to this
`
`challenging litigation, particularly in its later stages. The class representative and class counsel
`
`appointments remain in place and weigh in favor of final approval.
`
`3.
`
`Negotiation of Proposal at Arm’s Length
`
`The parties engaged in several rounds of mediation with respected neutrals, including a
`
`retired judge, a magistrate judge, and a former United States Ambassador who is also an attorney.
`
`The settlement agreement does not contain a “clear sailing” provision, and all indicia point to a
`
`finding that the proposed settlement was the product of serious, informed, and noncollusive
`
`negotiations. This factor supports final approval.
`
`4.
`
`Adequacy of Relief Provided for the Class
`
`For obvious reasons, the adequacy of relief provided for the class typically is the make-or-
`
`break factor in the final approval of a class settlement. The Court devoted substantial attention to
`
`it in denying preliminary approval to the original settlement proposal, and it is the factor that the
`
`pending objections focus on.
`
`In evaluating the adequacy of relief under Rule 23(e)(2)(C), the Court considers the costs,
`
`risks, and delay of trial and appeal; the effectiveness of the proposed method of distributing relief
`
`9
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 10 of 21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`to the class; the terms of any proposed award of attorney’s fees, including timing of payment; and
`
`any agreement required to be identified under Rule 23(e)(3). Several of the factors discussed in
`
`the Churchill Village case overlap with the Rule 23(e)(2)(C) subfactors, and also go to evaluating
`
`the adequacy of relief for the class. These include the strength of plaintiff’s case; the risks,
`
`expense, complexity, and likely duration of further litigation; the risks of maintaining class action
`
`status throughout the trial; the amount offered in settlement; the extent of discovery completed and
`
`stage of proceedings; and the presence or nonpresence of a governmental participant. Churchill
`
`Vill., 361 F.3d at 575.
`
`Our circuit has acknowledged that judging the adequacy of the payout to the class under
`
`these factors is “an amalgam of delicate balancing, gross approximations and rough justice.”
`
`Rodriguez v. West Publishing Corp., 563 F.3d 948, 965 (9th Cir. 2009) (quotations omitted). The
`
`Court declines the objectors’ invitation to adopt the Seventh Circuit’s approach of basing the
`
`evaluation on probabilities expressed in specific percentages. See Dkt. 518 at 4-6. That is not the
`
`law of our circuit, which “put[s] a good deal of stock in the product of an arms-length, non-
`
`collusive, negotiated resolution, and ha[s] never prescribed a particular formula by which that
`
`outcome must be tested.” Rodriguez, 563 F.3d at 965 (citations omitted). It can also cloak the
`
`adequacy analysis in a robe of pseudo-science. All good trial lawyers and judges know that it is a
`
`fool’s errand to predict a jury verdict. Consequently, it makes little sense to say a payout to class
`
`members is adequate or not because there is a 5% or 65% chance of success at trial. Those
`
`percentages create a false sense of analytical rigor about an outcome that is unknowable until the
`
`jury foreperson signs the verdict form.
`
`What matters is a realistic assessment of the overall risks and rewards of a trial in the
`
`context of each specific case. The record here leaves no doubt that the class would face
`
`substantial hurdles to prevailing at trial, and if successful, preserving the verdict on appeal. These
`
`hurdles include genuine disputes of fact about: (1) the geographic location of Facebook’s servers
`
`vis-à-vis its Illinois users, and whether servers located outside the state were within the purview of
`
`the BIPA; (2) whether Facebook’s facial recognition technology collects a “scan” of “face
`
`geometry” as required under BIPA, 740 Ill. Comp. Stat. 14/10, or used photographs, which are
`
`10
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 11 of 21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`expressly excluded as a basis for liability in the BIPA, id.; (3) whether Facebook obtained user
`
`consent for collection and storage of the scans; and (4) whether Facebook had a good-faith belief
`
`that its actions were permissible under the BIPA such that a finding of a reckless or even negligent
`
`violation would not be possible. These would be major factual disputes for the jury to decide on
`
`the basis of highly contested evidence. The Court has denied summary judgment for that very
`
`reason. See Dkt. No. 372. There is also the ongoing legal dispute over standing, which Facebook
`
`and other tech companies continue to press in other cases, including the FCRA case pending
`
`before the Supreme Court.
`
`Taken as a whole, these risks amply justify a settlement payout of at least $345 per
`
`claimant. To be sure, this is less than the full statutory rate available under the BIPA. For a
`
`“prevailing party,” the BIPA provides for “liquidated damages” of (1) $1,000 or actual damages,
`
`whichever is greater, against an entity that “negligently” violated the statute; and (2) $5,000 or
`
`actual damages, whichever is greater, against an entity that “intentionally or recklessly” violated
`
`the statute. 740 Ill. Comp. Stat. 14/20(1)-(2). But simply pointing out, as objectors do, that the
`
`$650 million settlement amount is less than the theoretical possibility of billions of dollars were
`
`the class to hit a home run at trial is not illuminating. See Dkt. No. 518 at 5. The objectors did not
`
`meaningfully account for the factual disputes that the jury would be called upon to decide, or for
`
`the threat of a change in the law of standing. An adverse determination on any of these disputed
`
`issues would defeat any recovery under the BIPA. And because the objectors did not
`
`meaningfully engage with these challenges, they did not propose any rational estimates of the
`
`probabilities of success at trial that were grounded in the facts of this case.
`
`Rather than doing that work, the objectors randomly pick various percentages of success at
`
`trial as purported measures of adequacy. They “believe,” for example, that it’s reasonable to
`
`assume at least a 15% likelihood that the class would win at trial, which they say implies a “total
`
`litigation value” of approximately $1.88 billion. See id. at 5-6. This comment is striking mainly
`
`for its lack of any connection to the issues and record in this case. The objectors made no effort to
`
`tie it and similar remarks to the disputes that would actually go to trial. In effect, the 15% is just a
`
`number pulled out of thin air. The objectors try to dress it up a bit saying it is based on “Seventh
`
`11
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 12 of 21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Circuit standards,” id., but even assuming that might be accurate, it does not change their straw
`
`into gold, particularly under the standards of our circuit. The objectors also overlook the strong
`
`likelihood that a complicated class action case with a 15% chance of success would never see a
`
`jury. Few if any lawyers would spend their time and money on a long trial in the face of an 85%
`
`probability of losing.
`
`Consequently, the objectors’ complaint that the settlement leaves a pot of money on the
`
`table is entirely unpersuasive. While that is enough to put the objections to rest, it is also worth
`
`noting that a multi-billion dollar verdict could create its own major risk for the class. A number of
`
`courts have commented on the potential for a due process problem when statutory damages are
`
`pursued by a large class, such as the one here. The resulting damages could be so large as to
`
`constitute a severe penalty out of all proportion to the harm the legislature sought to redress, or
`
`that the class members suffered. See Parker v. Time Warner Ent. Co., L.P., 331 F.3d 13, 21-23
`
`(2d Cir. 2003); see also Kline v. Coldwell Banker & Co., 508 F.2d 226, 234 (9th Cir. 1974)
`
`(noting in dicta that allowing thousands of class members to recover full statutory damages under
`
`the Truth in Lending Act would result in “outrageous amounts” and be an “absurd and stultifying”
`
`result beyond the goal of the statute) (quotation omitted). Some courts have denied class
`
`certification for that reason. See, e.g., Ratner v. Chemical Bank New York Trust Co., 54 F.R.D.
`
`412 (S.D.N.Y. 1972). That is a dubious outcome for many reasons, not least because it perversely
`
`insulates wrongful conduct from liability just for being widespread, but other solutions are also
`
`fraught with doubts. A concurrence in Parker suggests that statutory damages clauses simply
`
`should “not . . . be applied according to their literal terms when doing so achieves a result
`
`manifestly not intended by the legislature.” Parker, 331 F.3d at 27 (Newman, J., concurring).
`
`The problem, as the concurrence acknowledges, is that “this approach cannot be reconciled with
`
`the terms of the statute,” which “for some . . . would be an insuperable obstacle.” Id.
`
`This sticky issue need not be resolved here because the class has been certified and the
`
`$650 million settlement amount is fair, reasonable, and consistent with the Illinois Legislature’s
`
`intent in the BIPA. The issue bears mention mainly because it further underscores the holes in the
`
`objectors’ comments. The objectors demand that the class risk everything at trial for the
`
`12
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:15-cv-03747-JD Document 537 Filed 02/26/21 Page 13 of 21
`
`
`
`1
`
`