Opinion
Case No. 22-mc-80026-AGT
2022-07-11
ZURU, INC., Plaintiff, v. GLASSDOOR, INC., Defendant.
Michele D. Floyd, Kilpatrick Townsend & Stockton LLP, San Francisco, CA, Eric Lorenz Olavson, Sterling, VA, Joseph Ronald Oliveri, Clare Locke LLP, Alexandria, VA, for Plaintiff. William J. Frimel, Seubert French Frimel & Warner LLP, Menlo Park, CA, for Defendant.
Michele D. Floyd, Kilpatrick Townsend & Stockton LLP, San Francisco, CA, Eric Lorenz Olavson, Sterling, VA, Joseph Ronald Oliveri, Clare Locke LLP, Alexandria, VA, for Plaintiff.
William J. Frimel, Seubert French Frimel & Warner LLP, Menlo Park, CA, for Defendant.
ORDER ON MOTION TO QUASH
Re: Dkt. No. 8
Alex G. Tse, United States Magistrate Judge
A former employee of Zuru, Inc., or perhaps more than one, anonymously left six scathing reviews of the company on Glassdoor.com. The reviews refer to Zuru as a "[b]urn out factory" with a "toxic culture," where an "incompetent" management team "consistently talk[s] down" to employees and treats them like "dirt." Zuru says these and similar statements in the reviews are false, and the company wants to sue the reviewers for defamation in New Zealand, the country where Zuru was founded and where the reviewers worked. First, Zuru must determine who wrote the reviews.
To do so, Zuru previously applied to this Court for permission to subpoena Glassdoor, Inc. Zuru wants Glassdoor to reveal who wrote the reviews, to identify how many people have seen the reviews, and to provide some related information. Glassdoor is headquartered in this district, and under 28 U.S.C. § 1782(a), a district court can order persons within the district to produce discovery "for use in a proceeding in a foreign or international tribunal."
The Court previously allowed Zuru to subpoena Glassdoor but made clear that Glassdoor could move to quash the subpoena. Glassdoor so moved, and its motion is now before the Court. At the heart of its motion, Glassdoor contends that it shouldn't be required to unmask the anonymous reviewers because Zuru can't plead or prove a defamation claim against them. As a backstop, Glassdoor argues that even if it must reveal who wrote the reviews, it shouldn't be required to identify how many people have seen the reviews because that information isn't relevant, is commercially sensitive, and would be impracticable to produce. These arguments are addressed in sections that follow. I.
Before jumping into the analysis, the Court makes several comments here as an aside: First, the Court allowed Zuru to serve two subpoenas on Glassdoor, a document subpoena and a deposition subpoena. See Dkts. 1-2, 1-3, 7. Glassdoor moved to quash both, but the parties later resolved their dispute over the deposition subpoena, see Dkt. 11 at 9 n.3, leaving only the document subpoena to be addressed in this order.
Second, Glassdoor notified the reviewers that Zuru had asked for their names and explained that the reviewers could timely move to quash Zuru's subpoenas. See Dkt. 16-2, Mahood Decl. ¶ 9; see also Dkt. 7 at 2 ¶¶ 2–3. The reviewers didn't join Glassdoor's motion or file their own.
Third, Zuru and Glassdoor consented to have Glassdoor's motion to quash resolved by a magistrate judge. See Dkts. 20–21; 28 U.S.C. § 636(c).
Given Glassdoor's focus on the merits of Zuru's defamation claim, first this must be noted: § 1782(a), the governing statute, doesn't require the Court to evaluate the merits of Zuru's claim before ordering Glassdoor to produce discovery in furtherance of it.
Congress enacted § 1782(a) "to provide federal-court assistance in gathering evidence for use in foreign tribunals." Intel Corp. v. Advanced Micro Devices, Inc. , 542 U.S. 241, 247, 124 S.Ct. 2466, 159 L.Ed.2d 355 (2004). The statute is generous and reflects a hope that if federal courts assist with foreign litigation, foreign courts will do the same when the tables are turned. See ZF Auto. US, Inc. v. Luxshare, Ltd. , ––– U.S. ––––, 142 S. Ct. 2078, 2088, 213 L.Ed.2d 163 (2022) ("[T]he animating purpose of § 1782 is comity: Permitting federal courts to assist foreign and international governmental bodies promotes respect for foreign governments and encourages reciprocal assistance.").
The statute conditions U.S. judicial assistance on only three prerequisites, each of which is satisfied here: (i) "the person from whom the discovery is sought" must "reside[ ] or [be] found in the district of the district court where the application is made;" (ii) the discovery must be "for use in a proceeding in a foreign or international tribunal;" and (iii) the application must be made "by a foreign or international tribunal or any interested person." Khrapunov v. Prosyankin , 931 F.3d 922, 925 (9th Cir. 2019) (simplified). The merits of the foreign claim aren't essential. And imagine if they were: federal district courts would be called upon to regularly interpret and apply laws from around the globe.
This isn't to say that a district court, in evaluating a § 1782(a) application, can't consider the merits of a foreign claim. Once § 1782(a) ’s three prerequisites are met, the statute gives district courts wide latitude to determine "what, if any, [judicial] assistance is appropriate." Intel , 542 U.S. at 266, 124 S.Ct. 2466. It is beyond the scope of this order to catalogue all scenarios in which a merits review may make sense. But one such scenario is present here.
There are competing interests at play. Glassdoor wants to safeguard anonymous speech on its website. Zuru wants to protect its reputation. Both interests can't simultaneously be accommodated. In deciding which interest should prevail, the merits matter. If Zuru's defamation claim is baseless, Glassdoor may have a legitimate interest in shielding the reviewers’ identities. But if the reviewers made false statements, their "right to remain anonymous may give way to [Zuru's] need to discover [their] identity in order to pursue its claim." Art of Living Found. v. Does 1-10 , No. 10-CV-05022-LHK, 2011 WL 5444622, at *4 (N.D. Cal. Nov. 9, 2011). A peek at the merits will ensure that Zuru has legitimate reasons for outing the anonymous reviewers. Even though a merits review isn't required, the Court will conduct one.
II.
Zuru intends to sue the reviewers in New Zealand. In evaluating the merits, the Court will thus look to New Zealand law. All New Zealand legal authorities cited in this order were submitted by the parties and can be found in the record. See Dkts. 1-5, 8-3, 12, 16-2.
In New Zealand, when a corporation sues for defamation, it must prove four elements: (1) that a statement was made, (2) that the statement was defamatory, (3) that the statement was published by the defendant, and (4) that the statement caused or is likely to cause the defendant to suffer a monetary loss. See Solomon v. Prater [2021] NZHC 481 at [12]; New Zealand Defamation Act of 1992 § 6. New Zealand doesn't require the plaintiff to prove falsity; "[i]t is for the defendant to make out the truth of the statements in defending the action." CPA Australia Ltd. v New Zealand Inst. of Chartered Accountants [2015] NZHC 1854 at [44].
In California, by contrast, at least when statements involve "matters of public concern," which online reviews sometimes do, e.g., Wong v. Jing , 189 Cal. App. 4th 1354, 1366, 117 Cal.Rptr.3d 747 (2010) (review of dental practice on Yelp! held to be a matter of public concern), the plaintiff "bears the burden of showing the statements the defendant made were false," Sonoma Media Invs., LLC v. Superior Ct. , 34 Cal. App. 5th 24, 37, 247 Cal.Rptr.3d 5 (2019) (simplified).
Putting aside whether Zuru will be able to prove these four elements at trial, it is clear that the company can plausibly plead them.
(1) Statements were made: they were written in reviews on Glassdoor.com.
(2) A trier of fact could reasonably conclude that the statements were defamatory. That is, the statements would cause a reasonable person reading them "to think worse of the person concerned in a more than minor way." Low Volume Vehicle Tech. Ass'n v. Brett [2018] 2 NZLR 587 (HC) at [48]. How so? The reviews make Zuru, the company concerned, sound like a horrible place to work.
(3) The statements were authored and disseminated, and thus "published," Tairawhiti Dist. Health Bd. v Perks [2002] NZAR 533 (HC) at [16, 19], by the proposed defendants, the unknown Zuru employees who posted the reviews on Glassdoor.com.
(4) The statements caused monetary loss. Zuru's co-founders state that Zuru "has had to expend money, time, and resources in combatting the negative publicity, negative perception, and harm to [Zuru's] reputation that the [r]eviews have caused." Dkt. 14, Mat Mowbray Decl. ¶ 14; accord Dkt. 15, Nick Mowbray Decl. ¶ 15. As an example, the co-founders explain that after the reviews were published, they extended a job offer to a candidate for a management position, but the candidate declined the offer and cited the reviews as a reason for doing so. See M. Mowbray Decl. ¶ 13; N. Mowbray Decl. ¶ 14. Zuru then had to spend "money and resources to recruit additional candidates for the position." M. Mowbray Decl. ¶ 13; accord N. Mowbray Decl. ¶ 14.
Spending more money to recruit a job candidate is a monetary loss; it is not simply a business disruption or a loss of management time. Cf. Tairawhiti [2002] NZAR 533 at [22–23, 27] (concluding that the latter two types of harm were not "financial loss[es]" and thus weren't actionable under New Zealand defamation law).
The four elements are satisfied, but Glassdoor argues that Zuru's proposed defamation claim lacks merit, for two principal reasons. (Glassdoor also put forward a third reason, but the Court just rejected it in footnote three.) First, Glassdoor contends that the reviews in question don't contain defamatory statements because the reviews "constitute opinion," which Glassdoor argues "is not subject to defamation liability." Dkt. 8 at 9. Second, Glassdoor asserts that Zuru has failed to prove that it suffered a monetary loss, which Zuru must do, Glassdoor says, before the Court allows Zuru to learn who wrote the anonymous reviews. These two arguments are considered next. As will be seen, neither is persuasive.
A.
Glassdoor's first argument conflates U.S. and New Zealand defamation law. In the United States, statements of "pure opinion[ ]" are protected by the First Amendment. Partington v. Bugliosi , 56 F.3d 1147, 1153 (9th Cir. 1995) (simplified). Even if a statement of pure opinion is harmful to a person's reputation, it cannot give rise to a claim for defamation. See, e.g., Gardner v. Martino , 563 F.3d 981, 988 (9th Cir. 2009) (affirming dismissal of a defamation claim because the claim was based on statements of "nonactionable opinion"); Summit Bank v. Rogers , 206 Cal. App. 4th 669, 695–96, 142 Cal.Rptr.3d 40 (2012) ("Although statements of fact may be actionable as libel, statements of opinion are constitutionally protected.") (simplified).
In New Zealand, statements of opinion are not categorically protected. New Zealand makes available a more limited protection, allowing the defendant in a defamation action to assert the defense of "honest opinion." Low Volume Vehicle [2018] 2 NZLR at [59]. As the name suggests, this defense turns on whether the defendant "honestly held" the opinions he expressed. Id. (simplified). If the defendant's opinions were not honestly held, the defense isn't available and the defendant can be held liable for defamation, even for statements of opinion. See Low Volume Vehicle [2018] 2 NZLR at [59] ("Any individual has the right to express an opinion on something. Provided this opinion is honestly held , and the speaker has got his or her facts right ....") (emphasis added) (quoting another source).
The honest-opinion defense is an affirmative defense: the defendant bears the burden of proving that he honestly held the opinions he expressed. See Mitchell v Sprott [2002] 1 NZLR 766 at [16]; Low Volume Vehicle [2018] 2 NZLR at [60]; CPA Australia [2015] NZHC at [142]. The defense is also fact intensive; it is based on "what was going on in the mind" of the defendant when he was speaking. Mitchell [2002] 1 NZLR at [47]. Courts "hesitate to determine the question of the honesty of the opinion in advance of a trial." Id. at [48].
Here, then, the question under New Zealand law isn't whether the reviews on Glassdoor.com were opinions, but whether, if they were opinions, they reflected the reviewers’ honest opinions. That question can't be answered now. The reviewers haven't even been identified, much less appeared in court and offered evidence on the genuineness of their opinions. Zuru, at this point, just wants discovery—to learn who wrote the reviews.
Glassdoor relies on Coastal Abstract Service, Inc. v. First American Title Insurance Co. , 173 F.3d 725 (9th Cir. 1999), and on Mitchell v Sprott [2002] 1 NZLR 766, to support its argument that a statement of opinion, full stop, cannot give rise to defamation in New Zealand. Neither decision is on point. Coastal Abstract isn't relevant: the decision interpreted California's defamation law, not New Zealand's. And Glassdoor misreads Mitchell. The Mitchell court didn't hold that a statement of opinion cannot be defamatory. The court held, in affirming the denial of the defendant's motion for summary judgment, that the defendant's honest-opinion defense couldn't be determined "in advance of a trial." 1 NZLR at [48]; see also id. at [49] ("There is a logical difficulty in making a final determination on the issue of honesty of opinion at the present time ...."). Mitchell supports this Court's interpretation of New Zealand law, not Glassdoor's.
Glassdoor's first merits argument—that the reviews aren't defamatory because they express opinions—is based on a misreading of New Zealand law. Even if the reviews are classified as statements of opinion, they may be actionable.
B.
As for Glassdoor's second merits argument—that Zuru, at this stage in the case, must prove that it suffered a monetary loss—Glassdoor hasn't identified any New Zealand authority to support it. As a corporation, Zuru will eventually need to prove that it suffered, or is likely to suffer, a monetary loss. See New Zealand Defamation Act of 1992 § 6. But monetary loss is "a matter for proof at trial." CPA Australia [2015] NZHC at [66]. Its factual existence "cannot affect the outcome of ... whether or not there has been pleaded ... harm to the plaintiff." Id. (quoting another source).
The cases Glassdoor relies on don't suggest otherwise. See Chinese Herald Ltd. v New Times Madia Ltd. [2004] 2 NZLR 749 (HC) at [18], [52–57] (dismissing corporate plaintiff's defamation claim after an evidentiary hearing , because the plaintiff didn't offer evidence of monetary loss); Tairawhiti [2002] NZAR at [19–27] (same).
In support of its position, Glassdoor turns away from New Zealand law and invokes the First Amendment. When it applies, the First Amendment's free-speech clause protects "an author's decision to remain anonymous." McIntyre v. Ohio Elections Comm'n , 514 U.S. 334, 342, 115 S.Ct. 1511, 131 L.Ed.2d 426 (1995). That protection "is not unlimited," and competing interests, including a person's interest in protecting his reputation against falsehoods, can outweigh an anonymous speaker's free-speech rights. In re Anonymous Online Speakers , 661 F.3d 1168, 1173 (9th Cir. 2011) ; see also In re Rule 45 Subpoenas Issued to Google LLC & LinkedIn Corp. , 337 F.R.D. 639, 646 (N.D. Cal. 2020) (noting that the First Amendment "does not protect tortious, defamatory, or libelous speech"). But to ensure that a proper balance is struck and that the First Amendment isn't unjustifiably cast aside, judges in this district do often require plaintiffs who want to unmask anonymous speakers to do more than "plead and pray" that they have a viable defamation claim. Highfields Cap. Mgmt., L.P. v. Doe , 385 F. Supp. 2d 969, 975 (N.D. Cal. 2005). They must persuade the court that there is "a real evidentiary basis" for believing that the defendant acted unlawfully and caused "real harm." Id.
See also Wirt v. Twitter, Inc. , No. 21-MC-80166-JSC, 2021 WL 5919846, at *1–2 (N.D. Cal. Dec. 15, 2021) (following Highfields ); Smythe v. Does , No. 15-MC-80292-LB, 2016 WL 54125, at *2–4 (N.D. Cal. Jan. 5, 2016) (same); Kechara House Buddhist Ass'n Malaysia v. Does , No. 15-CV-00332-DMR, 2015 WL 5538999, at *4 (N.D. Cal. Sept. 18, 2015) (same); Art of Living Found. , 2011 WL 5444622, at *5–10 (same).
Glassdoor says Zuru hasn't provided "a real evidentiary basis" for its defamation claim, and in particular for its alleged monetary loss. Arguably, Glassdoor is wrong on the facts: Zuru hasn't simply alleged that it suffered a loss; Zuru's cofounders declared, under penalty of perjury, that because of the negative reviews, Zuru had to spend more money to recruit job candidates for a particular position. See M. Mowbray Decl. ¶ 13; N. Mowbray Decl. ¶ 14. These statements aren't allegations; they are evidence, and they could be used to prove a monetary loss. But putting this issue aside, the bigger problem for Glassdoor is that the company relies on a standard that doesn't apply.
Courts use the "real evidentiary basis" standard to protect First Amendment rights. But the speakers here, the anonymous reviewers, don't have those rights. They worked for Zuru in New Zealand, and there's no reason to believe they were U.S. citizens. "[F]oreign citizens outside U.S. territory do not possess rights under the U.S. Constitution," including under the First Amendment. Agency for Int'l Dev. v. All. for Open Soc'y Int'l, Inc. , ––– U.S. ––––, 140 S. Ct. 2082, 2086, 207 L.Ed.2d 654 (2020). In reply, Glassdoor doesn't argue that the reviewers do have First Amendment rights. But the company suggests that the Court, in its discretion, should still consider First Amendment free-speech principles. As Glassdoor sees it, those principles reflect a pro-free-speech U.S. policy, and the Court should reject § 1782(a) discovery requests that don't further that policy.
Some courts have applied Highfields ’s "real evidentiary basis" standard even when a § 1782(a) applicant has sought to identify an anonymous speaker who was a foreign citizen speaking outside U.S. territory. See, e.g., Tokyo Univ. of Soc. Welfare v. Twitter, Inc. , No. 21-MC-80102-DMR, 2021 WL 4124216, at *4–5 (N.D. Cal. Sept. 9, 2021) ; In re PGS Home Co. Ltd. , No. 19-MC-80139-JCS, 2019 WL 6311407, at *4–6 (N.D. Cal. Nov. 25, 2019). These courts overlooked that the basis for applying Highfields , the First Amendment, wasn't implicated.
Several problems with Glassdoor's position are apparent.
First, although the United States may have a pro-free-speech policy, the only source of that policy that Glassdoor has identified is the First Amendment. The First Amendment doesn't apply to foreign citizens outside U.S. territory, so it doesn't reflect a U.S. policy of protecting free speech around the world.
Second, there's good reason to tread lightly in applying U.S. free-speech principles abroad. Our country's commitment to free speech isn't universally shared; and even in other countries that protect free speech, a different balance is often struck between the right to free speech and the right to protect one's reputation, with the latter right usually receiving more weight. See Developments in the Law – The Law of Media , 120 Harv. L. Rev. 1031, 1037 (2007) (noting that British Commonwealth countries "value reputation over speech," while the United States is " ‘a recalcitrant outlier to a growing international understanding of what the freedom of expression entails’ " (quoting Frederick Schauer, The Exceptional First Amendment, in American Exceptionalism and Human Rights 30 (Michael Ignatieff ed., 2005))). "No other nation—not even freedom-loving countries like Canada, England, Australia, New Zealand and Israel—has protections of free speech and free press like those enshrined in the First Amendment." Minority Television Project, Inc. v. F.C.C. , 736 F.3d 1192, 1211 (9th Cir. 2013) (Kozinski, J., dissenting).
"[L]ibel tourism" has arisen because of these differences: plaintiffs sue American authors and publishers in foreign jurisdictions with "more favorable" defamation laws. Trout Point Lodge, Ltd. v. Handshoe , 729 F.3d 481, 487 (5th Cir. 2013) (simplified). In 2010, Congress passed the SPEECH Act, 28 U.S.C. §§ 4101 –05, to counteract this trend. Under the Act, foreign defamation judgments are "unrecognizable and unenforceable" in the United States unless "the law of the foreign forum ... provides free-speech protection that is coextensive with relevant domestic law" or "the facts, as proven in the foreign proceeding, are sufficient to establish a defamation claim under domestic law." Trout , 729 F.3d at 487–88 (citing 28 U.S.C. § 4102(a)(1) ). The SPEECH Act isn't implicated here—Zuru isn't attempting to enforce a foreign defamation judgment in the United States—but the Act illustrates that countries have struck different balances between free-speech and reputational rights.
Third, and related to the second point, the fact that the United States generally prioritizes free-speech rights over reputational rights doesn't mean the United States's hierarchy of rights should control when federal courts evaluate § 1782(a) applications. Section 1782(a) "is a provision for assistance to tribunals abroad." Intel , 542 U.S. at 263, 124 S.Ct. 2466. And the Supreme Court has made clear that the statute does not require the applicant to show "that United States law would allow discovery in domestic litigation analogous to the foreign proceeding." Id.
Fourth and finally, the above analysis should be kept in perspective. The Court isn't turning a blind eye to a baseless defamation claim; it already considered Zuru's allegations and concluded that they could support a defamation claim in New Zealand. Glassdoor says the Court should do more—that the Court should require Zuru to offer competent evidence to prove each element of its claim—but the heightened standard Glassdoor invokes is used by courts to protect First Amendment rights, rights that the proposed defendants don't have.
The Court won't require Zuru to satisfy Highfields ’s "real evidentiary basis" standard. Zuru has done enough to show that its defamation claim is plausible.
C.
Zuru's defamation claim, while plausible, almost surely won't make it off the ground without Glassdoor's help. Glassdoor knows who wrote the reviews, Zuru doesn't. And if Glassdoor doesn't identify the reviewers, Zuru can't sue them and will be left without a means by which to "vindicate [its] good name." Milkovich v. Lorain J. Co. , 497 U.S. 1, 12, 110 S.Ct. 2695, 111 L.Ed.2d 1 (1990). The discovery that Zuru seeks, in other words, is critical to the success of the company's proposed defamation claim.
Glassdoor has an interest in keeping the reviews on its website anonymous, so there's a conflict between what Zuru needs and what Glassdoor wants. But ultimately, Glassdoor hasn't convinced the Court that in New Zealand, the relevant jurisdiction, Glassdoor's interest in preserving anonymity would be prioritized over Zuru's interest in protecting its reputation. Because Glassdoor hasn't made that showing, and because Zuru's defamation claim is plausible, the Court will require Glassdoor to reveal who wrote the reviews.
Copies of five of the six reviews are in the record. See Dkt. 1-5. All five contain one or more statements that could be deemed defamatory. As to the sixth, Zuru didn't submit a copy and says it "does not currently possess" one. Dkt. 11 at 11. Because the sixth review isn't in the record, the Court can't assess whether it is plausibly defamatory and so won't presently require Glassdoor to identify who wrote it. Zuru's subpoena, however, asks for copies of the reviews, in addition to the reviewers’ identities, see Dkt. 1-3 at 8 (document subpoena, request no. 1), so Glassdoor must provide Zuru with a copy of the missing review; and if the content of the review is similar to that of the other reviews, Glassdoor will need to identify the reviewer.
III .
Glassdoor must reveal who wrote the reviews. Shifting its sights to another issue, the company argues that it shouldn't be required to identify how many people have seen the reviews, because that information isn't relevant, is commercially sensitive, and would be impracticable to produce. Glassdoor's relevancy and commercial-sensitivity arguments aren't persuasive, but its burden argument is.
The number of people who have seen the reviews is relevant. The more eyes on the reviews, the more harm that has been done to Zuru's reputation. Under New Zealand law, a corporate entity's loss of goodwill due to defamation is actionable. See Tairawhiti [2002] NZAR at [24]; Chinese Herald [2004] 2 NZLR at [53]. To prove a cognizable loss of goodwill, Zuru will likely need to do more than identify how many people have seen the reviews, but that number would be a start and is relevant.
As for commercial sensitivity, Glassdoor asserts that its website-user statistics are confidential. Perhaps so, but Glassdoor's interest in protecting that information from widespread disclosure can be safeguarded by a protective order, which the parties can stipulate to and which would prevent Zuru from using Glassdoor's confidential data outside this litigation and the New Zealand defamation case. Glassdoor's commercial-sensitivity concern can be adequately ameliorated, so it doesn't outweigh Zuru's interest in obtaining relevant information.
As for burden, Glassdoor says it can't produce the data Zuru seeks because it doesn't have it. Glassdoor explains that it "does not log the unique number of users who view each specific employer review on its site." See Dkt. 8-1, Freeman Decl. ¶ 3. The Court won't require Glassdoor to produce data the company doesn't have. But Glassdoor presumably does keep some statistics about how many people use its website and view the reviews posted there; and it is possible that some of those statistics could be relevant to Zuru's claim. The parties must meet and confer to further discuss what data is available. If certain user statistics are available and could advance Zuru's claim, Zuru may ask Glassdoor to produce them.
IV.
Glassdoor's motion to quash, with two exceptions, is denied. The company must comply with Zuru's document subpoena, which among other things requires Glassdoor to produce documents that identify who wrote the reviews in question. The motion is denied in the following limited ways: because one of the reviews isn't in the record, Glassdoor needn't identify its author; and because Glassdoor, without undue burden and expense, can't identify the number of people who have seen or engaged with the reviews, Glassdoor needn't produce that information.