Doe v. Google, Inc.

• Court: California Court of Appeal
• Decided: September 22, 2020
• Docket: A157097
• Status: Published

Opinion

Filed 9/21/20

CERTIFIED FOR PUBLICATION

IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA

FIRST APPELLATE DISTRICT

DIVISION FOUR

JOHN DOE et al., Plaintiffs and Appellants, A157097

v. (City & County of San Francisco GOOGLE, INC., et al., Super. Ct. No. CGC-16-556034 & Coordination Proceeding 4939) Defendants and Respondents.

Google, Inc. and Alphabet, Inc. (collectively, Google), and Adecco USA, Inc. (Adecco) require their employees to comply with various confidentiality policies. John Doe, David Gudeman, and Paola Correa, who are current and former Google and Adecco employees, sued Google and Adecco under the Labor Code Private Attorneys General Act of 2004 (PAGA) (Lab. Code, § 2698 et seq.), alleging the employers’ confidentiality policies restricted their employees’ speech in violation of California law. The trial court sustained defendants’ demurrers without leave to amend, concluding plaintiffs’ claims were preempted by the National Labor Relations Act (NLRA or Act) (29 U.S.C. § 151 et seq.) under San Diego Bldg. Trades Council v. Garmon (1959) 359 U.S. 236, 244–245 (Garmon). Plaintiffs contend the trial court erred in finding the NLRA preempted their PAGA claims. They further challenge the trial court’s denial of a petition to coordinate this case with another case pending in a different trial court.

1 We conclude that, although many of plaintiffs’ claims relate to conduct that is arguably within the scope of the NLRA, the claims fall within the local interest exception to Garmon preemption and may therefore go forward. We also conclude that plaintiffs’ challenge to the trial court’s coordination petition is not properly before us. We will therefore reverse the trial court’s orders sustaining defendants’ demurrers without leave to amend and remand for further proceedings. BACKGROUND Because this appeal comes to us on demurrer, the following facts are based on the allegations in plaintiffs’ pleadings and the requests for judicial notice.1 Litigation Regarding Confidentiality Policies Doe works as a product manager in a supervisory capacity at Google. He began work at Google in July 2014, had his employment terminated in April 2016, and was reinstated in June 2016. After being terminated and before being reinstated, Doe sent notice under PAGA to the California Labor and Workforce Development Agency that he intended to file this suit on behalf of himself and other current and former Google employees. Doe alleged that Google required employees to sign a confidentiality agreement

1 Google and plaintiffs have requested judicial notice of various submissions to and rulings by the NLRB’s regional director and general counsel. The requests are unopposed. With one exception, we grant the requests for notice of these documents as official acts or records of the executive department or a court of record of the United States. (Evid. Code, §§ 452, subds. (c)-(d), 459; PG&E Corp. v. Public Utilities Com. (2004) 118 Cal.App.4th 1174, 1220, fn. 38 [taking judicial notice of briefs filed before administrative agency]; Heston v. Farmers Ins. Group (1984) 160 Cal.App.3d 402, 413 [approving of judicial notice of brief filed with the NLRB as court record].) We deny Google’s request for notice of Doe’s unfair labor practice charge as unnecessary, because that document is already in the record.

2 and imposed certain related confidentiality policies on its employees, and that these policies violated the Labor Code. Six months later, Doe filed this case in San Francisco Superior Court. (John Doe et al. v. Google Inc. et al (Super. Ct. S.F. City & County, 2016, No. CGC-16-556034) (Doe).) Gudeman is a former Google employee, and Correa is a former Google employee who also worked for Adecco as a temporary employee placed at Google. Doe’s second amended complaint included them as named plaintiffs, and added claims against Adecco based on Correa’s experience there. Shortly after plaintiffs filed their second amended complaint, Rachel Moniz filed a complaint against Adecco in San Mateo Superior Court alleging claims based on Adecco’s confidentiality policies. (Moniz v. Adecco (Super. Ct. San Mateo County, 2017, No. 17-CIV-01736) (Moniz).) Ten days later, plaintiffs filed their third amended complaint against Google and Adecco. The Harms Alleged Plaintiffs’ third amended complaint alleges 17 causes of action under PAGA based on defendants’ confidentiality policies. Plaintiffs’ confidentiality claims fall into three subcategories; restraints of competition, whistleblowing, and freedom of speech. In their competition causes of action plaintiffs allege that Google’s confidentiality rules violate state statutes by preventing employees from using or disclosing the skills, knowledge, and experience they obtained at Google for purposes of competing with Google. For example, the policies prevent Googlers from disclosing their wages in negotiating a new job with a prospective employer, and from disclosing who else works at Google and under what circumstances such that they might be receptive to an offer from a rival employer. The complaint grounds these PAGA claims on alleged

3 violations of Business & Professions Code sections 17200, 16600, and 167002 and various provisions of the Labor Code (see Lab. Code, §§ 232, 232.5, 1197.5, subd. (k)). Plaintiffs’ whistleblowing causes of action allege that Google’s confidentiality rules prevent employees from disclosing violations of state and federal law, either within Google to their managers or outside Google to private attorneys or government officials. (See Bus. & Prof. Code, §§ 17200 et seq.; Lab. Code, § 1102.5.) They also allege the policies unlawfully prevent employees from disclosing information about unsafe or discriminatory working conditions, or about wage and hour violations. (See Lab. Code, §§ 232, 232.5.) In their freedom of speech claims, plaintiffs allege that defendants’ confidentiality rules prevent employees from engaging in lawful conduct during non-work hours and violate state statutes entitling employees to disclose wages, working conditions, and illegal conduct. (See Lab. Code, §§ 96, subd. (k), 98.6, 232, 232.5, 1197.5, subd. (k).) This lawful conduct includes the exercise of an employee’s constitutional rights of freedom of speech and economic liberty. As a practical matter, plaintiffs argue, they are forbidden even to write a novel about working in Silicon Valley or to reassure their parents they are making enough money to pay their bills, matters untethered to any legitimate need for confidentiality. Google’s confidentiality rules contain a savings clause stating that the company’s rules were not intended to limit employees’ right to discuss wages, terms, or conditions of employment with other employees, or their right to

2The fifth amended complaint expressly grounds the Business & Professions Code section 17200 allegation on violation of Business & Professions Code sections 16600 and 16700.

4 communicate with government agencies regarding violations of law. However, plaintiffs allege these clauses are meaningless and contrary to Google’s policies and practices of enforcement, which threaten employees for disclosing any information at all. Plaintiffs allege Adecco was liable for both its own confidentiality policies and Google’s because Adecco was Correa’s joint employer when she was placed at Google. Adecco admits that in ruling on the demurrers “there is no meaningful difference between [the] claims against Google and those against Adecco.” Demurrers Google demurred to the entire complaint. As relevant here, Google argued the NLRA preempted plaintiffs’ confidentiality claims. The trial court sustained Google’s demurrer to the confidentiality claims without leave to amend. It overruled the demurrer only as to a single remaining cause of action—alleging defendants required employees to sign illegal releases of potential claims as a condition of being hired—and the parties eventually settled that claim. Adecco demurred to the third amended complaint as well, shortly after it filed a similar demurrer in Moniz. The Moniz court overruled the demurrer, but the Doe court sustained Adecco’s demurrer to the confidentiality claims, with leave to amend, for the same reasons that it sustained Google’s demurrer. Proceedings Specific to Adecco Plaintiffs tried to cure the defects identified by the Doe court as to their claims against Adecco by filing a fourth amended complaint. This complaint retains the allegation that Adecco is jointly liable under PAGA for Google’s confidentiality rules, but adds separate claims on behalf of Adecco employees

5 statewide based on Adecco’s own confidentiality rules. The new causes of action against Adecco fall into the same competition, whistleblowing, and free speech categories as the claims against Google in the third amended complaint. Plaintiffs also allege Adecco had an unlawful policy prohibiting temporary employees placed at Google from working directly for Google without Adecco’s consent. Adecco again demurred, and the trial court sustained the demurrer, this time without leave to amend. Plaintiffs then amended their Doe complaint a final time to add an illegal release claim against Adecco, a claim the parties subsequently settled. Before Adecco filed its demurrer to the third amended complaint, it filed with the Judicial Council a petition to coordinate the action with Moniz. After plaintiffs filed their fourth amended complaint and shortly before Adecco demurred to it, the coordination judge continued proceedings on Adecco’s petition until after the ruling on Adecco’s forthcoming demurrer. Then, after the Doe court sustained Adecco’s demurrer to the fourth amended complaint without leave to amend, the coordination judge denied the petition to coordinate, explaining that the sole then-remaining cause of action in Doe (the illegal release claim) was not at issue in Moniz, the claims in Moniz covered more employees than the claim in Doe, and the Moniz litigation had advanced further. Adecco filed a petition for writ of mandate in this court seeking review of the coordination judge’s denial of its coordination petition. Plaintiffs likewise filed a petition for writ of mandate, seeking review of the Doe court’s orders sustaining Google’s and Adecco’s demurrers. This court summarily denied Adecco’s writ and denied plaintiffs’ writ as untimely. (Adecco USA, Inc. v. Superior Court for the City & County of San Francisco (Feb. 6, 2018,

6 A153470) [nonpub. opn.]; Doe et al. v. Superior Court for the City & County of San Francisco (Mar. 29, 2018, A153726) [nonpub. opn.].) The trial court in Doe entered final judgment, and plaintiffs timely appealed. NLRB Files Then Settles Complaint Against Google At the same time as Doe sent the PAGA notices anticipating this case, he also filed an unfair labor practice charge against Google with the National Labor Relations Board (NLRB or Board). Doe alleged Google’s confidentiality rules violated section 8 of the NLRA by prohibiting employees from exercising their rights under section 7 of the Act, which entitles employees to engage in “concerted activities for the purpose of collective bargaining or other mutual aid or protection.” (29 U.S.C. § 157) Doe alleged that Google violated section 8 by terminating him because he exercised his section 7 rights. On the same day that plaintiffs filed their third amended complaint in Doe, the regional director of the NLRB issued a complaint against Google based on Doe’s unfair labor practice charge. However, the regional director’s complaint did not include certain allegations from Doe’s charge, including the allegation relating to Doe’s termination, because the regional director determined Doe had been a supervisor and therefore was not protected by the NLRA. Doe appealed that decision, but the NLRB’s general counsel denied the appeal. After plaintiffs filed their opening brief in this court, the NLRB’s regional director and Google reached an informal settlement on the NLRB’s complaint.3 As part of that settlement, Google agreed to post a notice for 60

3 We discuss the proceedings on the regional director’s complaint that transpired after the trial court entered judgment because they are not in dispute and come to us by way of judicial notice. (Reserve Insurance Co. v. Pisciotta (1982) 30 Cal.3d 800, 813.)

7 days informing employees that they had the right “to discuss wages, hours, and working conditions with other employees, the press/media, and other third parties, and [Google] WILL NOT do anything to interfere with [employees’] exercise of those rights.” The notice further stated that Google would “NOT prohibit [employees] from discussing or sharing information relating to [their] performance, salaries, benefits, discipline, training, or any other terms and conditions of [their] employment and” had rescinded any such limitations in its confidentiality rules. In exchange, the NLRB regional director would withdraw her complaint, but this would not prevent the courts or the Board from proceeding with other cases. DISCUSSION I. The NLRA and Garmon Preemption Plaintiffs contend the trial court erred in finding the NLRA preempts their confidentiality claims. We review this question de novo. (Wal-Mart Stores, Inc. v. United Food & Commercial Workers Internat. Union (2016) 4 Cal.App.5th 194, 201 (Wal-Mart).) Likewise, de novo review applies to a trial court’s decision sustaining a demurrer. (Traders Sports, Inc. v. City of San Leandro (2001) 93 Cal.App.4th 37, 43). As we shall explain, we conclude that these causes of action fall within the local interest exception to preemption. A. Legal Principles Congress intended the NLRA to serve as a comprehensive law governing labor relations; accordingly, “the NLRB has exclusive jurisdiction over disputes involving unfair labor practices, and ‘state jurisdiction must yield’ when state action would regulate conduct governed by the NLRA. (Garmon, [supra, 359 U.S.] at pp. 244–245.)” (Wal-Mart, supra, 4 Cal.App.5th at pp. 200–201.) Because it is for the NLRB to determine, in

8 the first instance, whether conduct is in fact governed by the NLRA, the Act’s preemptive effect may extend beyond conduct that the NLRA directly governs to “activities which ‘arguably’ constitute unfair labor practices under the Act.” (Balog v. LRJV, Inc. (1988) 204 Cal.App.3d 1295, 1303 (Balog); see Garmon, at pp. 244–245.) Such conduct is “presumptively pre-empted.” (Belknap, Inc. v. Hale (1983) 463 U.S. 491, 498 (Belknap).) But Garmon preemption must not be applied in a “ ‘literal, mechanical fashion’ ” (Local 926, Internat. Union of Operating Engineers, AFL-CIO v. Jones (1983) 460 U.S. 669, 676 (Jones)), and it is subject to exceptions where the activity in question is a “merely peripheral concern” of the NLRA, or where “the regulated conduct touche[s] interests so deeply rooted in local feeling and responsibility that, in the absence of compelling congressional direction, we could not infer that Congress had deprived the States of the power to act.” (Garmon, supra, 359 U.S. at pp. 243–244.) Although framed as separate exceptions, these two factors are often analyzed together, as we will do here. (See, e.g., Linn v. United Plant Guard Workers (1966) 383 U.S. 53, 61 (Linn); Balog, supra, 204 Cal.App.3d at p. 1304.) B. Federal and State Interests at Stake Garmon preemption “has its greatest force when applied to state laws regulating the relations between employees, their union, and their employer.” (Sears, Roebuck & Co. v. San Diego County Dist. Council of Carpenters (1978) 436 U.S. 180, 193 (Sears).) However, “the general applicability of a state cause of action is not sufficient to exempt it from pre-emption.” (Farmer v. United Brotherhood of Carpenters and Joiners of America, Local 25 (1977) 430 U.S. 290, 300 (Farmer).) Rather, we conduct a “balanced inquiry” into the federal and state interests at stake and the potential to interfere with the

9 NLRB’s jurisdiction. (Ibid.) With this in mind, we consider the interests at stake in this action. The NLRA “was designed to ‘eliminate the causes of certain substantial obstructions to the free flow of commerce . . . by encouraging the practice and procedure of collective bargaining, and by protecting the exercise by workers of full freedom of association, self-organization, and designation of representatives of their own choosing, for the purpose of negotiating the terms and conditions of their employment or other mutual aid or protection.’ ” (Balog, supra, 204 Cal.App.3d at p. 1301, quoting 29 U.S.C. § 151.) To this end, section 7 of the NLRA gives non-exempt employees the right to self- organize, bargain collectively, and “engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection.” (29 U.S.C. § 157.) The NLRA also defines certain actions as unfair labor practices. (Balog, at p. 1302, citing 29 U.S.C. §§ 158, 160.) As pertinent here, section 8 of the NLRA declares it an “unfair labor practice for an employer . . . [¶] to interfere with, restrain, or coerce employees in the exercise of the rights guaranteed in” section 7. (29 U.S.C. § 158.) The focus of these provisions is on workers joining together for mutual benefit. By contrast here, plaintiffs seek to enforce Labor Code provisions that protect their activities as individuals. For example, one provision prohibits employers from preventing an employee “from disclosing the amount of his or her wages” (Lab. Code, § 232), a statute that was enacted at the urging of women’s groups to protect employees sharing information necessary to the enforcement of laws against sex discrimination. (See, e.g., Sen. Com. on Industrial Relations Staff Analysis of Assem. Bill No. 3193 (1983-1984 Reg. Sess.) as amended March 21, 1984.) Another provision provides analogous protection for an employee disclosing “information about the employer’s

10 working conditions” (Lab. Code, § 232.5), manifesting California’s public policy to “prohibit[] employer restrictions on, or punishment for, speech regarding conditions of employment” (Glassdoor, Inc. v. Superior Court (2017) 9 Cal.App.5th 623, 633). A third protects the rights of any employee to disclose information about a violation of state or federal law to someone with the power to address the problem—“to a government or law enforcement agency, to a person with authority over the employee, or to another employee who has authority to investigate, discover, or correct” the violation. (Lab. Code, § 1102.5.) A fourth provision protects employees who complain about underpayment of wages to the Labor Commissioner. (Lab. Code, § 98.6; see also Lab. Code, § 1102.5 [protecting right to disclose information to state agencies].) And a fifth protects an employee from retaliation for his or her “lawful conduct occurring during nonworking hours away from the employer’s premises” (Lab. Code, § 96, subd. (k)), so employers do not seek to control non-work aspects of their employees’ lives. Plaintiffs allege that defendants’ confidentiality policies violate these provisions of California law. Plaintiffs also allege violations of section 16600 of the Business and Professions Code, which prohibits any contract that would improperly restrain an employee from securing new employment with a competitor. This statute “evinces a settled legislative policy in favor of open competition and employee mobility” (Edwards v. Arthur Anderson LLP (2008) 44 Cal.4th 937, 946), a policy that has been seen as instrumental in the success of California’s technology industry (see Gilson, The Legal Infrastructure of High Technology Industrial Districts: Silicon Valley, Route 128, and Covenants Not to Compete (1999) 74 N.Y.U.L. Rev. 575, 609 [“Silicon Valley’s legal infrastructure, in the form of Business and Profession[s] Code section 16600’s prohibition of covenants not to compete, provided a pole around which Silicon

11 Valley’s characteristic business culture and structure precipitated”]; see also Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (1994) pp. 34–37.) Keeping these very different federal and state interests in mind, we now analyze Garmon preemption in this case. C. Arguably Protected or Prohibited Activity The first step of a Garmon analysis asks whether the conduct at issue is arguably protected or prohibited by the NLRA. (Jones, supra, 460 U.S. at p. 676.) The trial court concluded all of plaintiffs’ confidentiality claims are presumptively preempted in their entirety because they involve policies against disclosure of wages and working conditions (in the case of the competition claims and some freedom of speech claims) or against disclosures intended to affect the terms or conditions of employment (in the case of the whistleblowing and some freedom of speech claims). We do not doubt that some of the conduct at issue at least arguably falls within the NLRA. (See Luke v. Collotype Labels USA, Inc. (2008) 159 Cal.App.4th 1463, 1470 [discussions among workers about working conditions are protected activity under NLRA].) Indeed, the fact that the regional director brought a complaint challenging Google’s confidentiality policies indicates that she so concluded. However, plaintiffs also allege conduct that clearly falls outside the scope of the NLRA. For instance, plaintiffs’ competition claims allege defendants’ confidentiality rules inhibit an employee seeking new employment elsewhere and competing with defendants. They also allege Adecco prevents its employees from working with companies where Adecco has placed them, unless Adecco consents. These matters are, on their face, unrelated to “mutual aid or protection” (29 U.S.C. § 157) of fellow employees

12 at Google or Adecco. Similarly, some of plaintiffs’ whistleblowing causes of action allege defendants’ confidentiality policies prevent them from discussing with the government legal violations unconnected to working conditions, such as an employer’s violations of securities laws, false claims laws, the federal Foreign Corrupt Practices Act, and other laws unrelated to employees’ terms and conditions of employment. The NLRB has authoritatively rejected the argument that whistleblowing about employer conduct unrelated to working conditions is protected activity, so the NLRA does not protect an employee reporting concerns about patient care in a nursing home. (Orchard Park Health Care Center, Inc. (2004) 341 NLRB 642, 645.) But we need not belabor this point because, as we shall next discuss, regardless of whether the challenged policies reach employee conduct that the NLRA arguably protects or prohibits, plaintiffs’ state-law causes of action fall within the local interest exception to Garmon preemption. D. The Local Interest Exception The local interest exception vindicates interests “ ‘deeply rooted in local feeling and responsibility.’ ” (Sears, supra, 436 U.S. at p. 195.) Two factors relevant to the application of this exception, in a case where an employer’s policies are arguably prohibited by the NLRA, are: (1) whether there is “a significant state interest in protecting the citizen from the challenged conduct” and (2) whether “the exercise of state jurisdiction over the tort claim [for trespass] entailed little risk of interference with the regulatory jurisdiction of the Labor Board.” (Id. at pp. 196–197.) The local interest exception has been applied in a range of circumstances. As explained in Inter-Modal Rail Employees Assn. v. Burlington Northern & Santa Fe Ry. Co. (1999) 73 Cal.App.4th 918 (Inter- Modal), “the Supreme Court has declined to preempt a variety of state law

13 claims even though they arose in a labor law context [involving, for example,] trespass by peaceful picking . . . intentional infliction of emotional distress . . . [and] defamation . . .’ ” (Id. at p. 925; see Sears, supra, 436 U.S. at p. 198 [trespass by picketing]; Farmer, supra, 430 U.S. at pp. 299–300 [intentional infliction of emotional distress]; Linn, supra, 383 U.S. at pp. 61–62 [defamation].) The local interest exception has also been applied to a cause of action challenging an employer’s retaliation against employees for raising concerns about workplace safety (Inter-Modal, at pp. 922–923, 925, citing Balog, supra, 204 Cal.App.3d at p. 1304), and to controversies where the NLRB could not have provided relief to the plaintiffs because their injury was not relevant to its functions (Service by Medallion, Inc. v. Clorox Co. (1996) 44 Cal.App.4th 1807, 1815–1816 (Clorox) [service provider’s contract negotiations with company took place against “backdrop” of union campaign]). Defendants do not deny that plaintiffs’ claims grow from deeply-rooted local interests. This is no surprise, as plaintiffs bring this case under PAGA, which means plaintiffs are serving “ ‘as the proxy or agent of the state’s labor law enforcement agencies.’ ” (Kim v. Reins Internat. California, Inc. (2020) 9 Cal.5th 73, 81, italics omitted.) Courts have long recognized the importance of state labor regulation that “provides protections to individual union and nonunion workers alike, and thus ‘neither encourage[s] nor discourage[s] the collective-bargaining processes that are the subject of the NLRA.’ ” (Fort Halifax Packing Co. v. Coyne (1987) 482 U.S. 1, 20–21.) “[P]re-emption should not be lightly inferred in this area, since the establishment of labor standards falls within the traditional police power of the State.” (Id. at p. 21; accord Iskanian v. CLS Transportation Los Angeles, LLC (2014) 59 Cal.4th 348, 388 [“enactment and enforcement of laws concerning wages, hours, and

14 other terms of employment is within the state’s historic police power”— powers that “ ‘ “courts should assume . . . are not superseded ‘unless that was the clear and manifest purpose of Congress’ ” ’ ”].) The state statutes plaintiffs seek to enforce are all labor standards of this sort, statutes that preserve the freedom of all employees to practice their profession or trade (Bus. & Prof. Code, § 16600), to report wage-and-hour violations or unsafe working conditions to government agencies (Lab. Code, § 1102.5), and to speak as they choose about their work lives (Lab. Code, §§ 232, 232.5, 96, subd. (k)). In sum, these statutes establish as a minimum employment standard an employee anti-gag rule. Not only are the interests protected by these statutes matters of traditional local concern, but they may reasonably be seen as peripheral to the NLRA. Nothing about the NLRA manifests a purpose to displace state labor laws regulating wages, hours, and other terms of employment, as the NLRA is “aimed at ‘safeguard[ing], first and foremost, workers’ rights to join unions and to engage in collective bargaining.’ ” (Epic Sys. Corp. v. Lewis (2018) __ U.S. __ [138 S.Ct. 1612, 1630] (Epic); see also Inter-Modal, supra, 73 Cal.App.4th at p. 926 [focus of NLRA is “ ‘an equitable bargaining process[;] . . . Congress did not intend to preempt all local regulations that touch or concern the employment relationship’ ”].) It is thus well established that a state may set minimum employment standards without running afoul of the NLRA. (Castillo v. Toll Bros., Inc. (2011) 197 Cal.App.4th 1172, 1207 [“state wage-and-hour statutes . . . raise no Garmon preemption concerns”].) The state laws plaintiffs assert here govern matters similarly far afield from the concerns underlying the NLRA. Unable to refute the local interests at stake, defendants instead argue that because the NLRB issued a complaint at Doe’s behest, to allow this case

15 to proceed in state court would risk interfering with the jurisdiction of the NLRB. Were this a serious concern, it would render the local interest exception unavailable. (See Sears, supra, 436 U.S. at pp. 196–197; Hillhaven Oakland Nursing etc. Center v. Health Care Workers Union (1996) 41 Cal.App.4th 846, 855 (Hillhaven); Rodriguez v. Yellow Cab Cooperative, Inc. (1988) 206 Cal.App.3d 668, 678–679.) But the NLRB has settled its claim with no admission of wrongdoing by Google and no findings of fact by the Board. Nothing the state court does at this juncture could interfere with the NLRB’s exercise of its primary jurisdiction. Asked about this point at oral argument, counsel for Google responded with two concerns: (1) that the state court could reach “a different finding on the merits,” in that “the NLRB . . . issued a complaint and [Google] entered into a settlement on it, so there could be a different result in state court on liability,” and (2) that state courts cannot impose “a different remedial scheme for NLRA violations,” especially a scheme of punitive remedies as was found preempted in Wisconsin Dept. of Industry v. Gould, Inc. (1986) 475 U.S. 282 (Wisconsin Dept. of Industry). Responding to these concerns in turn, neither is substantial. First, it would be impossible for the state court to reach “a different result . . . on liability,” since the NLRB settled its case without resolving liability issues. The settlement agreement between the Board and Google is informal and of limited scope. It requires Google to post for 60 days a notice informing its employees of their rights under “FEDERAL LAW,” and if Google upholds its end of the bargain then the NLRB promises to take no further action in the case. The reference to federal law is a signal that the question on liability that underlay the NLRB case (i.e., whether defendants violated the NLRA) is completely different from the liability questions in this

16 case (i.e., whether defendants violated California labor laws). Moreover, the agreement expressly “does not prevent . . . the Board and the courts from finding violations with respect to matters” occurring before the agreement was approved, or from “mak[ing] findings of fact and/or conclusions of law with respect to” evidence obtained in the case. With this provision, the Board itself has given courts license to proceed with claims addressing the same or similar facts. The terms of the agreement itself suggest that, whatever California courts would ultimately decide on plaintiffs’ claims, the Board sees in plaintiffs’ case no threat to its own jurisdiction. As for Google’s second concern—duplicative and punitive remedies for an NLRA violation—this argument founders at the outset because none of plaintiffs’ claims requires proof of an NLRA violation. The difference between this case and Wisconsin Dept. of Industry illustrates the point. There, the state of Wisconsin had adopted a law debarring from state contracting any company “found by judicially enforced orders of the National Labor Relations Board to have violated the NLRA” three times in five years. (Wisconsin Dept. of Industry, supra, 475 U.S. at pp. 283–284.) The NLRA preempts this statute “[b]ecause Wisconsin’s debarment law functions unambiguously as a supplemental sanction for violations of the NLRA.” (Id. at p. 288.) By contrast, the California laws that plaintiffs seek to enforce make no reference to the NLRA, the NLRB, or the rights of workers to organize. They do not supplement sanctions for a violation of the NLRA, but instead extend unrelated protections to conduct that may, or may not, also be addressed by the NLRA. In such circumstances, the availability of a remedy in state court that is unavailable under the NLRA may be a reason not to find a case preempted. (Linn, supra, 383 U.S. at pp. 63–64; Clorox, supra, 44 Cal.App.4th at p. 1816.)

17 In sum, analyzing the two factors the United States Supreme Court has identified as dispositive—the significance of the local interest and the risk of interference with the jurisdiction of the Board—we see no basis for preemption here. (See Sears, supra, 436 U.S. at pp. 196–197; Farmer, supra, 430 U.S. at p. 300.) But the parties have argued, citing competing precedents and legal tests ostensibly derived from them, for alternative ways of analyzing the local interest exception, so we now turn to consider these alternatives. 1. Sears, Linn, and the “Critical Inquiry” In Sears, after the Supreme Court set forth the two relevant factors we have just examined, it synthesized them into a single “critical inquiry” for preemption of claims based on arguably prohibited conduct. That inquiry is “whether the controversy presented to the state court is identical to . . . or different from” a controversy that could have been presented to the NLRB. (Sears, supra, 436 U.S. at p. 197.) Answering that question in Sears meant an employer’s state-court trespass case against a union was not preempted— even though the picketing in question might have been protected or prohibited by the NLRA—because the issues involved in the trespass case were “different from” the issues the NLRB would have considered in assessing the legality of the same picketing under federal law. (Id. at pp. 197–198; see also Wal-Mart, supra, 4 Cal.App.5th 194 [same].) By contrast, a controversy “ ‘identical to’ ” one that could have been presented to the NLRB was an attempt to enforce the Pennsylvania Labor Relations Act, whose relevant language was “ ‘almost identical to’ ” language in the NLRA. (Sears, at pp. 192, 197, discussing Garner v. Teamsters, Chauffeurs & Helpers Local Union (1953) 346 U.S. 485, 487–489, fns. 3 & 5 [employer’s attempt to

18 enforce Pennsylvania Labor Relations Act against peaceful union picketing is preempted].) Sears’s focus on whether the legal issue in the two controversies is the same or different also animates the Supreme Court’s decision in Linn. There, the Court held a state-court libel action was not preempted, explaining: “When the Board and state law frown upon the publication of malicious libel, albeit for different reasons, it may be expected that the injured party will request both administrative and judicial relief.” (Linn, supra, 383 U.S. at p. 66.) Under the formulations of either of these cases, plaintiffs’ claims are not preempted. The Board may “frown upon” an employer’s confidentiality policy because it interferes with workers’ rights to undertake concerted action, but California law disapproves such policies for a different reason: because they interfere with every employee’s right to bring workplace issues to the attention of supervisors, state agencies, courts, and the public. (See Linn, supra, 383 U.S. at p. 66.) And, although there may be overlap in the operative facts, whether an employer’s confidentiality policy constitutes an unfair labor practice under the NLRA is a “different” controversy from the question of whether it violates provisions of the state Labor Code. (See Sears, supra, 436 U.S. at p. 197.) Highlighting that the controversy here is different from the controversy that was, or could have been, placed before the NLRB is the Board’s decision in Boeing Co. (2017) 2017 NLRB Lexis 634 (Boeing), which elucidates how the NLRB would evaluate whether Google’s confidentiality policies comply with the NLRA. In Boeing, the NLRB announces a new standard for determining whether an employer’s adoption of a facially neutral workplace rule that potentially interferes with Section 7 rights is an unfair labor

19 practice. The Board concludes it must evaluate and weigh “(i) the nature and extent of the potential impact on NLRA rights,” and (ii) an employer’s “legitimate justifications associated with” business requirements. (Id. at pp. *60–*63.) This process could lead the NLRB to uphold confidentiality rules that risk inhibiting NLRA-protected activity, especially if that activity is peripheral, rather than central, to the NLRA’s concerns, or the risk of intruding on NLRA-protected rights is “ ‘comparatively slight.’ ” (Id. at p. *66.) Not surprisingly, there is no suggestion that a state’s interests underlying its own statutes will figure in this weighing process at all. The issues and concerns before the NLRB in deciding a challenge to defendants’ confidentiality policies would be wholly different from the state-law issues in this case, and by the same token the issues the state court must adjudicate in this case will require no consideration of the Section 7 rights that animate the NLRB. Thus, under the “critical inquiry” enunciated in Sears (Sears, supra, 436 U.S. at p. 197), plaintiffs’ claims are not preempted. 2. Jones and the “Crucial Element” Although Google acknowledges Sears and Linn remain good law, it urges us to focus instead on a subsequent Supreme Court case in which preemption was found, Jones, supra, 460 U.S. 669. In that case, Jones filed an NLRB charge against a union representing employees at his former company, where he had been hired as a supervisor but then quickly let go. Jones alleged that because he was not a member of the union, the union “ ‘procured’ his discharge, ‘and thereby coerced [the Company] in the selection of its supervisors and bargaining representative,’ ” an unfair labor practice under the NLRA. (Id. at p. 672.) The regional director refused to issue a complaint, concluding “there was insufficient evidence to establish that the Union had caused Jones’ discharge;” the union had “merely participated in

20 discussions” about “changes in the Company’s supervisory structure.” (Id. at pp. 672–673.) Rather than appealing this decision to the General Counsel of the NLRB, Jones filed a state-court action alleging the union had interfered with his employment contract. (Id. at p. 673.) The high court held this action was preempted for several reasons, including that Jones was seeking to prove the union coerced his discharge, a claim that was “concededly preempted” as an unfair labor practice (Jones, supra, 460 U.S. at pp. 681–682); that asking the state court to police the line between a coerced or uncoerced discharge would have required the court to adjudicate issues of federal labor law (id. at pp. 682); and that if Jones attempted to prove non-coercive interference with his employment there would be two further problems. He would still need to prove the union had caused his ouster—a “crucial element” of the NLRA claim that the Regional Director had already decided against Jones—and he would be seeking to impose liability for union conduct that the NLRA arguably protects. (Id. at pp. 682–684.) Relying on Jones, defendants argue the local interest exception does not apply in this case because the dispute in this case and a dispute properly before the Board share a “crucial element,” namely, whether defendants’ policies actually restrict employees from discussing wages and working conditions. But we do not read Jones to create a rule that if the state-law controversy shares a factual element—“crucial” or otherwise—with a matter properly before the NLRB, then the case is necessarily preempted. Such a rule would eviscerate the local interest and peripheral concern exceptions, since a court only considers these exceptions if some common set of facts gives rise to both the state-law claims and a dispute arguably within the purview of the NLRB, as with the picketing activity in Sears.

21 The Jones court does not announce any such revision of settled law. Instead, Jones recognizes the continuing force of Sears and seeks to distinguish it on the ground that the focus of the unfair labor practice charge in Sears was unrelated to that of the trespass action challenging the same picketing activity. (Jones, supra, 460 U.S. at pp. 682–683.) Although Jones does use the phrase “crucial element” in explaining one of several reasons that together explain the court’s preemption finding, the Court does not hold the phrase out as any sort of dispositive test, nor attempt to explain how a court would decipher when an “element” is “crucial.” (Id. at p. 682.) Instead, Jones follows Sears and Farmer in directing courts to undertake “a sensitive balancing” of potential harms to the Congressional scheme for regulating labor-management relations and to a state’s power to protect its citizens. (Jones, at p. 676.) Even if we were to attempt application of a “crucial element” test here, we disagree that the “crucial element” in this case is whether defendants’ policies restrict employees from discussing their wages and working conditions. This factual question about the scope of the employers’ policies may be an area of overlap between this case and a dispute properly before the Board, but the question is antecedent to those questions that bring to bear legal considerations that differ for the two disputes. The crucial elements for the state-law confidentiality claims are whether defendants’ policies infringe on an employee’s right to practice a profession or trade, disclose wrongdoing, and exercise free speech as protected by California law. The crucial elements in the Board’s determination of whether the confidentiality policies are an unfair labor practice are the extent to which the policies interfere with NLRA-protected activity, how central any such protected activity is to the organizing and bargaining activities that are the NLRA’s core concerns, and

22 whether the employer’s business justifications offset any interference with NLRA rights. (Boeing, supra, 2017 NLRB Lexis 634, at pp. *60–*63, *66; 29 U.S.C. § 151.) These elements are not common to the two disputes. Our case is different from Jones in other respects as well. First and foremost, there is in our case no issue of federal labor law that the state court would be required to adjudicate. (Cf. Jones, supra, 460 U.S. at p. 682.) California courts can and should decide whether Google and Adecco violated California law without considering whether, in so doing, they also committed unfair labor practices under the Act. Second, the regional director has made no factual determination that is fatal to plaintiffs’ claims, as occurred in Jones. (Id. at p. 682.) Thus, plaintiffs can proceed in state court without ever taking a position inconsistent with one already adopted by the Board or its regional director. Third, neither Google nor Adecco argues that its policies are protected by federal labor law, as the union’s conduct in Jones arguably was. (Id. at pp. 672–673.) This factor is important because federal supremacy is “implicated to a greater extent when labor-related activity is protected than when it is prohibited.” (Sears, supra, 436 U.S. at p. 200; see also Belknap, supra, 463 U.S. at pp. 498–499 [courts must balance state’s interest against interference with NLRB’s jurisdiction and risk that the state will sanction conduct the NLRA protects].) Finally, in this case there is no union. The absence of a union is significant, for the argument in favor of preemption “has its greatest force when applied to state laws regulating the relations between employees, their union, and their employer.” (Sears, supra, 436 U.S. at p. 193; see also Epic, supra, 138 S.Ct. at p. 1630 [NLRA “ ‘safeguard[s] first and foremost, workers’ rights to join unions and to engage in collective bargaining’ ”].)

23 Because our case differs from Jones in all of these substantial ways and because even Jones did not offer “crucial element” as a dispositive test,4 we decline defendants’ invitation to defeat the local interest exemption on this basis. 3. Hillhaven and Bright-Line Rules Google also argues that Hillhaven, supra, 41 Cal.App.4th 846, is “the dispositive precedent” defeating the local interest exemption in this case. In Hillhaven, another division of our court held that the NLRA preempted state- court action against a union alleged to have overrun a nursing home, disrupting patient care and intimidating workers. (Id. at pp. 850, 862.) The union was “the certified bargaining representative of employees at Hillhaven,” and was in the midst of negotiating a new collective bargaining agreement at the time. (Id. at pp. 849–850.) In finding preemption, the Hillhaven court relied on common factual issues between the state-court suit

The concurring and dissenting opinion accuses us of “ignor[ing] 4

Jones’s reasoning” and “the analytical path the Supreme Court has set forth for the local interest exception.” (At pp. 4, 13, post.) But there is nothing “novel” about our analyzing “competing interests.” (At p. 7, post.) The Supreme Court requires that we conduct a “balanced inquiry into such factors as the nature of the federal and state interests in regulation and the potential for interference with federal regulation” (Farmer, supra, 430 U.S. at p. 300), give “careful consideration [to] the relative impact . . . on the various interests affected” (Sears, supra, 436 U.S. at p. 188), and, in the language of Jones, engage in “a sensitive balancing” of harm to the NLRA’s regulatory scheme and to the state’s interest in protecting its citizens. (Jones, supra, 460 U.S. at p. 676). While in Sears the Supreme Court distilled this balancing of competing interests into a single “critical inquiry” (Sears, at p. 197), the concurring and dissenting opinion dismisses that analytical approach as “of only academic interest” based on a comment made in dissent in Jones. (At p. 10, post.) But no dissenting opinion has the power to overrule precedent, and we have shown that plaintiffs’ claims clear the “identical controversy” hurdle Sears sets forth. (Sears, supra, 436 U.S. at p. 197.)

24 and a complaint already settled before the NLRB and also, more importantly, two factors with no parallel in the case before us that go to the heart of the NLRB’s authority. First was the likelihood “that resolution of some of the state court claims would require . . . interpretation of the collective bargaining agreement between the parties.” (Id. at pp. 860–861 [e.g., “number of union representatives allowed to enter the facility, and where those representatives were permitted access” likely turned on interpretation of collective bargaining agreement].) Second was the “real possibility of conflict” between the injunctive relief Hillhaven sought in state court and “NLRB rulings on issues such as union access to employees at their place of work.” (Id. at p. 861.) Obviously, our case involves no collective bargaining agreement, no union, and no risk that the state court will punish or prohibit conduct that NLRB rulings protect. Defendants extract from the facts of Hillhaven a bright-line rule they would have us apply, that where the regional director has filed a complaint addressing conduct that is also the subject of a state-court action, the state- court action is preempted. We think defendants make too much of an observation in Hillhaven that the court was unaware of any decision failing to find preemption once the regional director had issued a complaint. (Hillhaven, supra, 41 Cal.App.4th at p. 859.) Hillhaven itself acknowledges that “simultaneous jurisdiction of the NLRB and state court is possible for conduct arguably prohibited under the” NLRA (ibid., italics omitted), and other courts have indeed adjudicated controversies after the NLRB issued and settled a related complaint (see, e.g., Belknap, supra, 463 U.S. at 496, 508–509; United Food & Commercial Workers Internat. Union v. Wal-Mart Stores, Inc. (2017) 453 Md. 482, 490–491, 508–511).

25 Where the local interest is strong, even the possibility of findings that conflict with an NLRB complaint need not be fatal. In Linn, the regional director of the NLRB declined to file a complaint against a union because factual investigation led him to conclude “the union was not responsible for” the offending conduct—there, the distribution of the allegedly libelous leaflets. (Linn, supra, 383 U.S. at p. 57.) Yet, the Supreme Court allowed Linn’s libel case against the union to proceed based on the peripheral concern and local interest exceptions, untroubled that the factual issue of the union’s responsibility for the leaflets might be decided differently in the state-court case. (Id. at pp. 61–62, 67.) Although Linn does not, as plaintiffs suggest, create an opposite bright-line rule—that state-court actions may always proceed in parallel to NLRB proceedings when an employer’s conduct violates both the NLRA and state law—its reasoning does establish that with a strong local interest and a peripheral NLRA concern, the possibility of conflicting findings does not foreclose a state-court action. 4. Conclusion The first step of a Garmon preemption analysis sweeps broadly, presumptively preempting conduct that may, in the end, be of only peripheral concern to (or even lie outside the scope of) the NLRA. The local interest exception is vital to protecting workers in such cases. And even where certain aspects of a dispute do, or could, attract the enforcement efforts of the NLRB, “defendants should not be able to escape the jurisdiction of California courts simply because, in addition to allegedly undertaking violations of health and safety regulations which are of compelling local importance and interest, they had the good fortune to also undertake the commission of NLRB-defined unfair labor practices.” (Balog, supra, 204 Cal.App.3d at

26 p. 1308 [plaintiff may proceed with wrongful termination claim to extent it is based on theories not preempted by NLRA].) The complaint in this case makes no mention of union organizing or other concerted activity, and it alleges violations of state law that can be proven without considering whether defendants’ actions also amounted to unfair labor practices under the NLRA.5 Because the asserted statutes protecting competition, whistleblowing, and free speech fit comfortably within our state’s historic police powers and address conduct affecting individual employees, as distinct from the NLRA’s focus on concerted activity, and because this state-court action poses no threat to the NLRA’s exercise of its own jurisdiction, our courts retain the power to decide these claims. II. Denial of the Petition to Coordinate In addition to challenging the trial court’s demurrer rulings, plaintiffs argue the coordination judge should not have continued the hearing on Adecco’s coordination petition and then denied Adecco’s petition to coordinate petitioners’ case with Moniz. Defendants argue that we may not review the substance of the order on the coordination petition because such orders are reviewable only via writ petition. We agree with defendants that the coordination order is not properly before us. We begin by examining the procedures for coordination. When an individual wants to coordinate two or more actions pending in different courts, he or she must submit a petition to the Chairperson of the Judicial

5 We disagree with the concurring and dissenting opinion that our decision will require California courts to decide issues of federal labor law— specifically, whether plaintiffs’ evidence involves concerted activity. (At p. 15, post.) That question is immaterial to the state-law causes of action plaintiffs allege, and we see no reason to litigate it here.

27 Council. (Code Civ. Proc., § 404; remaining statutory references are to the Code of Civil Procedure.) The Chairperson assigns the petition “a special title and coordination proceeding number.” (Cal. Rules of Ct., rule 3.550(c).) The Chairperson then assigns (or authorizes a presiding judge to assign) a coordination judge to decide whether coordination is appropriate. (§ 404.) If the coordination judge decides coordination is appropriate, he or she selects an appellate court to review decisions from the coordinated proceeding. (§ 404.2.) The Chairperson of the Judicial Council then assigns (or authorizes a presiding judge to assign) a judge to hear the coordinated actions. (§ 404.3.) After service of notice of entry of an order relating to coordination, “any party may petition the appropriate reviewing court for a writ of mandate to require the court to make such order as the reviewing court finds appropriate.” (§ 404.6.) This framework demonstrates that coordination petitions are not necessarily decided under the jurisdiction of any one of the courts in which the actions potentially subject to coordination are pending. Rather, the coordination proceeding is its own type of special proceeding, with a separate caption and number. (Cal. Rules of Ct., rule 3.550(c).) When the coordination judge grants or denies a petition for coordination, that order is not filed in the trial court on its own; the party that requested coordination must file it in all included actions. (Cal. Rules of Ct., rule 3.529(a).) As a result, a coordination order is not part of the bundle of orders reviewable via appeal from final judgment in any one of the actions for which coordination was sought. The fact that the judge presiding over the Doe action was also assigned to decide the coordination motion here changes nothing. Were the coordination order reviewable after final judgment, as plaintiffs contend, it would be subject to multiple appeals after final judgment in each of the

28 included actions, with the unacceptable potential for inconsistent rulings. Alternatively, if the coordination order were reviewable after final judgment in the included action only if that action is pending in the coordination judge’s own court (even though no statute or rule requires the coordination judge to be one of the judges hearing an included action), then the parties in that action alone would be able to appeal the order after final judgment. The parties in the other actions for which coordination was sought would be relegated to the writ review procedure in section 404.6, an unequal and unjust result. Consequently, although plaintiffs are correct that section 404.6 does not explicitly say so, we hold that section 404.6 is the exclusive method for appellate review of coordination orders. (Cf. Lautrup, Inc. v. Trans-West Discount Corp. (1976) 64 Cal.App.3d 316, 317–318 [coordination orders are not separately appealable orders under section 904.1; appellate review of such orders is via writ of mandate under section 404.6].) Plaintiffs having failed timely to file a petition seeking writ relief from the trial court’s decision not to coordinate this matter with Moniz, they may not take a second bite at the coordination apple on their appeal from the trial court’s orders on demurrer. DISPOSITION The judgment is reversed, as are the orders sustaining defendants’ demurrers without leave to amend. We dismiss as untimely the appeal from the order denying coordination with Moniz, and remand the case for further proceedings consistent with this opinion. Plaintiffs are entitled to their costs on appeal. (See Cal. Rules of Court, rule 8.278(a)(3).)

29 _________________________ TUCHER, J.

I CONCUR:

_________________________ POLLAK, P. J.

Doe et al. v. Google, Inc. et al. (A157097)

30 POLLAK, P. J.— I concur in the lead opinion. I would add that the line between those state law claims that, while based on conduct arguably protected or prohibited by the National Labor Relations Act (NLRA), are nonetheless exempted from preemption, and those that are preempted, is clarified by the fundamental distinction between such cases as Linn v. United Plant Guard Workers (1966) 383 U.S. 53, 61 (Linn) and Sears, Roebuck & Co. v. San Diego County Dist. Council of Carpenters (1978) 436 U.S. 18, 193 (Sears) on one hand, and Local 926, Internat. Union of Operating Engineers, AFL-CIO v. Jones (1983) 460 U.S. 669 (Jones) on the other. For the plaintiff in Jones, to prove his state law claim he would have to prove the very fact that would necessarily constitute a violation of federal law, namely, that the union coerced the employer to breach its employment contract with him.1 The National Labor Relations Board (NLRB) and the state court might have reached different conclusions on that common issue, hence the conflict, and preemption. In Linn and Sears, it was not necessary to prove a violation of federal law in order to prove the alleged violation of state law. Although the state court claims were based on alleged facts common to potential violations of the NLRA—defamation by union officers in Linn, and trespass while picketing in

1 Although the plaintiff argued that he also asserted a claim for uncoerced interference, which would not violate the NLRA, unlike Justice Brown (at pp. 11-12, post), I read the court to have treated his claims as the same, asserting “Even on Jones’ view of the elements of his state-law cause of action, the federal and state claims are thus the same in a fundamental respect.” (Jones, supra, 460 U.S. at p.682.) This conflation was questioned by the dissenting opinion (id. at p. 688 (dis. opn. of Rehnquist, J.)), but nevertheless the majority opinion treated both claims as requiring a determination of whether the defendant had committed acts that constituted an unfair labor practice.

1 Sears—it was not necessary to prove the elements of an unfair labor practice in order to prove the defamation or the trespass. The case before us is comparable to the situation in Linn and Sears, rather than the situation in Jones. To prove that defendants’ nondisclosure policies violate the various provisions of California law on which the complaint is based, it will not be necessary to prove any facts that would constitute an unfair labor practice. Plaintiff’s claims threaten neither duplication nor conflict with any claims within the jurisdiction of the NLRB, nor any interference with the enforcement of the NLRA.

2 BROWN, J., Concurring and Dissenting Plaintiff John Doe filed an unfair labor practice charge against Google, Inc. and Alphabet, Inc. (collectively, Google) with the National Labor Relations Board (NLRB or Board), alleging Google’s confidentiality policies violated the National Labor Relations Act (29 U.S.C. § 151 et seq. (NLRA or Act)). The regional director of the NLRB issued a complaint against Google based on that charge. In a settlement of that complaint, Google agreed to withdraw the portions of its confidentiality policies on which the regional director based her complaint. Based on the same confidentiality policies, Doe, joined by David Gudeman and Paola Correa, nevertheless continues to pursue claims under the Private Attorneys General Act (Lab. Code, § 2698 et seq. (PAGA)), seeking to add significant monetary penalties beyond what the NLRB required of Google. In these circumstances, I conclude that some of plaintiffs’ claims pose a substantial risk of interference with the NLRB’s jurisdiction. I therefore dissent from the majority opinion’s holding that none of plaintiffs’ claims are preempted under San Diego Unions v. Garmon, (1959) 359 U.S. 236, 244–245 (Garmon). Based on how the Supreme Court applied Garmon preemption in Operating Engineers v. Jones (1983) 460 U.S. 669, 676 (Jones), I would instead hold that while many of the theories of liability in plaintiffs’ pleadings survive, some theories are preempted.1 I. Majority opinion’s interests analysis As the majority opinion states, under the first stage of Garmon analysis “there is a presumption of preemption” for state law claims regulating any conduct even arguably covered by the Act. (Wal-Mart Stores, Inc. v. United Food & Commercial Workers Internat. Union (2016) 4 Cal.App.5th 194, 202,

I join in full the majority opinion’s holding that the coordination order 1

is not properly before us. 1 citing Belknap, Inc. v. Hale (1983) 463 U.S. 491, 498.) A claim can only overcome this presumption if, at the second Garmon stage, a court determines “the conduct at issue is only a peripheral concern of the Act or touches on interests so deeply rooted in local feeling and responsibility that, in the absence of compelling congressional direction, it could not be inferred that Congress intended to deprive the state of the power to act.” (Jones, supra, 460 U.S. at p. 676.) I generally agree with the majority opinion’s analysis of the first Garmon stage. The allegations concerning competition that support plaintiffs’ restraint of trade claims are incompatible with the possibility that employees were working together for mutual aid or protection. As these allegations do not even arguably trigger coverage of the Act, I agree that they are not preempted. (Longshoremen v. Davis (1986) 476 U.S. 380, 395 [no preemption where argument for NLRA coverage is “plainly contrary to its language”].) The same rationale extends to plaintiffs’ free speech claims. They are not preempted to the extent they allege defendants’ policies prevent employees from doing things like writing novels based on experiences at Google, because the Act does not even arguably reach such conduct. Likewise, plaintiffs’ whistleblowing claims based on allegations that defendants’ policies prevent employees from raising violations of laws unconnected with working conditions, such as securities laws or the federal Foreign Corrupt Practices Act (15 U.S.C. §§ 78dd-1 et seq.), are not preempted because the Board has definitively ruled that the Act does not protect such activity. (Davis, at p. 395 [no preemption where Board has “ ‘authoritatively rejected’ ” prospect of NLRA coverage of conduct].) However, as the majority opinion recognizes, other allegations in plaintiffs’ pleadings do involve conduct arguably covered by the Act. The

2 allegations in plaintiffs’ whistleblowing and free speech claims concerning the disclosure of wages and working conditions intrude into territory the NLRA arguably—indeed, unquestionably—covers. These allegations implicate a long line of Board authority stating that the NLRA prohibits employers from interfering with employees’ discussions of wages and working conditions among themselves or with third parties, or whistleblowing about violations of law related to their wages and working conditions. (See, e.g., Parexel International (2011) 356 NLRB 516, 518 [“wage discussions among employees are considered to be at the core of Section 7 rights because wages, ‘probably the most critical element in employment,’ are ‘the grist on which concerted activity feeds’ ”]; Victory Casino Cruises II (April 22, 2016) 363 NLRB No. 167 [2016 NLRB LEXIS 300 at *11] [“employees have a Section 7 right to discuss their conditions of employment with third parties, such as union representatives, Board agents, and the public in general, and the Board has invalidated rules prohibiting such third-party communication”]; Trinity Protection Services, Inc. (2011) 357 NLRB 1382, 1383 [“employees’ concerted communications regarding matters affecting their employment with their employer’s customers or with other third parties, such as governmental agencies, are protected by Section 7 and, with some exceptions not applicable here, cannot lawfully be banned”]; see also Eastex, Inc. v. NLRB (1978) 437 U.S. 556, 566 [Board has held the Act “protects employees from retaliation by their employers when they seek to improve working conditions through resort to administrative and judicial forums”].) My disagreement with the majority opinion arises at the second stage of the Garmon preemption analysis, concerning the question of whether the local interest exception saves these aspects of plaintiffs’ claims. To begin with, I disagree with the majority opinion’s analytical approach. The

3 majority opinion first determines that the state interests here are deeply rooted and that those interests are at the periphery of the Act. (Maj. opn. ante, at pp. 9–10, 14–15.) Only then does it go on to examine whether the tests “ostensibly derived from” the Supreme Court’s most recent applicable Garmon precedents lead to the outcome the majority opinion has already reached. (Maj. opn. ante, at p. 18.) The local interest exception analysis is designed to balance competing state and federal interests (see maj. opn. ante, at p. 24, fn. 4). But the high Court has devised and applied a test focused on the degree of overlap between the state and federal laws to guide that analysis, rather than trying to assess the relative importance of the interests freehand (and beforehand), as the majority does. (Sears, Roebuck & Co. v. Carpenters (1978) 436 U.S. 180, 197 (Sears).) This test is not “ostensibly derived” from precedent. (Maj. opn. ante, at p. 18.) It is precedent. (Seminole Tribe of Fla. v. Florida (1996) 517 U.S. 44, 67 [“When an opinion issues for the Court, it is not only the result but also those portions of the opinion necessary to that result by which we are bound”]; see Ramos v. Louisiana (2020) ___ U.S. ___ [140 S.Ct. 1390, 1416, fns. 5 & 6 (conc. opn. of Kavanaugh, J.)] [“[T]he state courts and the other federal courts have a constitutional obligation to follow a precedent of this Court unless and until it is overruled by this Court.” “In the American system of stare decisis, the result and the reasoning each independently have precedential force, and courts are therefore bound to follow both the result and the reasoning of a prior decision”].) We are therefore obligated to follow the analytical path the Supreme Court has set forth for the local interest exception rather than fashioning our own test. Moreover, the majority opinion’s approach of trying to divine the applicability of the local interest exception by first ranking the significance of the state’s interests or

4 categorizing them as lying at the core or periphery of the NLRA will likely lead to unpredictability, as the approach lacks concrete criteria or analytical guideposts and for that reason ends up being entirely subjective. Not only am I skeptical of the order in which the majority proceeds with its analysis of the competing state and federal interests, I am unconvinced by the substance of the majority opinion’s reasoning in its own right. The majority opinion concludes that the statutes underlying plaintiffs’ PAGA claims involve deeply rooted local interests because they involve substantive labor regulation. The state’s desire to regulate employees’ speech vis à vis their employers may be good public policy, but it is substantially different on its face from the state interests the Supreme Court has so far recognized as supporting the exception, which have included addressing violence, threats of violence, libel, infliction of emotional distress, trespass, obstruction of access to property, and breach of contract actions by laid-off replacement employees. (Hillhaven Oakland Nursing etc. Center v. Health Care Workers Union (1996) 41 Cal.App.4th 846, 854.) In my view, these sorts of claims—which involve classic areas of state common law—are qualitatively different from the purely economic regulation underlying the Labor Code statutes at issue here. The majority’s citations to Fort Halifax Packing Co. v. Coyne (1987) 482 U.S. 1, 20–21, and Iskanian v. CLS Transportation Los Angeles, LLC (2014) 59 Cal.4th 348, 388, to support its contrary view are unpersuasive. The former concerned an entirely different form of NLRA preemption arising under Machinists v. Wisconsin Emp. Rel. Comm’n, 427 U.S. 132, 140 (1976). (Fort Halifax, at pp. 19–20.) In the latter, as the trial court recognized, the California Supreme Court held only that PAGA claims are a form of qui tam action that may proceed despite any covered employees’ agreement to arbitrate. Neither case demonstrates that the United States

5 Supreme Court would view the Labor Code provisions here as implicating interests similar to the state’s desire to prevent violence or protect property from trespass. More importantly, the state’s interest in establishing minimum labor standards is irrelevant because the Labor Code provisions about which I disagree with the majority opinion are not minimum labor standards and cannot be said to lie at the periphery of the NLRA. The majority opinion summarizes the relevant statutes as “establish[ing] as a minimum employment standard an employee anti-gag rule” and states that “[n]othing about the NLRA manifests a purpose to displace state labor laws regulating wages, hours, and other terms of employment.” I would have no difficulty holding that the NLRA does not preempt California’s substantive labor standards, such as minimum wage, overtime, or anti-discrimination laws. But the Labor Code provisions on which plaintiffs base their claims are not this sort of law. As the majority opinion’s description of plaintiffs’ claims demonstrates, the statutes at issue regulate the types of information employees can share with each other and third parties as a means to an end: to allow employees to take action to improve their wages and working conditions. (See maj. opn. ante, at pp. 11–12 [noting, e.g., plaintiffs’ allegations relating to Labor Code provisions that are intended to protect employees who share information in order to address employer sex discrimination and underpayment of wages].) This distinction matters. The statutes’ regulation of the process by which employees improve their working conditions (i.e., by sharing information relating to their wages and working conditions), rather than the substance of those working conditions, places plaintiffs’ PAGA claims within the territory at least arguably covered by the Act.

6 II. Sears and the identical controversy test Besides disagreeing with the majority opinion’s choice to first engage in a novel analysis of the competing interests before applying the Supreme Court’s test for the local interest exception, I am also not convinced that it has accurately stated or applied that test. The Supreme Court announced in Sears that the “critical inquiry” for the local interest exception concerning conduct arguably prohibited by the Act is whether the controversy in a state court action is “identical to . . . or different from” the controversy that could be submitted to the Board. (Sears, supra, 436 U.S. at p. 197.) The majority opinion finds the controversies in the regional director’s suit and plaintiff’s claims different because the laws underlying each have different purposes. (Maj. opn. ante, at pp. 19–20.) This distinction is immaterial. The majority cites no apposite authority for the notion that a state law with a purpose different from the NLRA will escape preemption. Linn v. Plant Guard Workers (1966) 383 U.S. 53 (Linn) stated only that when state libel law prohibits conduct and offers different remedies for different reasons than the NLRA, then parties can be expected to pursue both forms of relief, rather than choosing to pursue relief only in state court. (Id. at p. 66.)2 Linn’s discussion of the different purposes for the two laws is also inapposite because, among other things, there the Board had already authoritatively

Relying on Linn, the majority opinion elsewhere observes that the 2

availability of a remedy in state court that is unavailable under the NLRA may be a reason not to find a case preempted. (Maj. opn. ante, at pp. 18–19.) But Linn’s point was merely that the inability of an NLRA claim to address a particular type of harm “vitiate[d]” the need for preemption. (Linn, supra, 383 U.S. at pp. 63–64.) This point has no application here, as the harm plaintiffs seek to remedy with the threat of PAGA penalties is the existence of excessively restrictive confidentiality policies, and the regional director’s complaint has already led Google to withdraw the offending policies. 7 rejected the idea that defamatory statements were covered by the NLRA. (Linn, supra, 383 U.S. at pp. 60–61.) This is not the case here, where the regional director has evidently concluded that the Act does cover defendants’ policies’ restrictions on discussion of wages and working conditions. Google’s settlement of the NLRB complaint and rescission of the aspects of the policies at issue further suggest the regional director has at least an arguable case. Even if the purposes of the laws at issue were relevant, I would conclude the NLRA and the Labor Code statutes supporting plaintiffs’ whistleblowing and free speech claims based on allegations concerning wages and working conditions involve the same fundamental controversy. The NLRB regional director’s complaint concerns whether Google’s confidentiality policies prevented employees from discussing their wages and working conditions with each other or third parties for their mutual aid or protection. The allegations in plaintiffs’ PAGA claims that involve conduct arguably covered by the NLRA concern whether defendants’ policies prevented employees from discussing wages and working conditions or blowing the whistle on workplace misconduct in order to improve employees’ welfare. For example, several of plaintiffs’ causes of action allege defendants’ policies unlawfully prevented employees from disclosing defendants’ failures to pay overtime and other wage and hour violations. The goal of employees’ concerted activity for mutual aid and protection under the NLRA is the same improvement of employee welfare that underlies plaintiffs’ PAGA claims based on allegations concerning wages and working conditions. The majority opinion concludes that “whether an employer’s confidentiality policy constitutes an unfair labor practice under the NLRA is a ‘different’ controversy from the question of whether it violates provisions of the state Labor Code.” (Maj. opn. ante, at p. 19.) This is tautological: the

8 majority concludes the controversies in plaintiffs’ state suit and the regional director’s NLRA complaint are different because the question of whether defendants’ policy violates state law is different from whether they violate the NLRA. To shore up the tautology, the majority opinion notes that to determine whether a confidentiality policy violates the NLRA, the Board engages in a balancing test that does not take into account a state’s interests. (See The Boeing Co. (Dec. 14, 2017) 365 NLRB No. 154.) But the Boeing balancing test is not a preemption test, so there is no reason for it to consider states’ interests. That aside, it is irrelevant that plaintiffs do not need to prove a violation of the NLRA in order to prevail on their PAGA claims or that the Board does not consider state interests. As the trial court here recognized, if a complete overlap of elements were a prerequisite under Garmon, then any state law claim with even a single different element from an NLRA unfair labor practice charge would avoid preemption. The Supreme Court has never taken this sort of formulaic approach to Garmon preemption—an approach that would make preemption easily avoidable by all but the most inept of complaint-drafters. In my view, the breadth of the majority’s conclusion underscores the lack of soundness in its reasoning. The structure of Garmon preemption sweeps broadly by presumptively preempting any claims based on conduct even arguably covered by the Act. (Jones, supra, 460 U.S. at p. 676.) It then excepts certain limited categories of state law claims that will be allowed to proceed. (Ibid.) This expansive approach “not only mandates the substantive preemption by the federal labor law in the areas to which it applies, but also protects the exclusive jurisdiction of the [NLRB] over matters arguably within the reach of the Act.” (Id. at p. 680.) By contrast, the majority opinion would allow virtually any state law claim to proceed, regardless of its effects

9 on the Board’s jurisdiction, so long as it does not refer to the NLRA by name or duplicate its elements. This flips the Garmon framework on its head, transforming it from a doctrine that sweeps widely with a carefully considered exception into a doctrine that allows everything to proceed except for a few, narrowly targeted areas of preemption where a state claim includes all the elements of an NLRA claim. By defining the local interest exception so broadly, the majority opinion allows it to swallow the intentionally wide rule of Garmon preemption and defeat its purpose. III. Jones and the crucial element test In any event, the application of the identical controversy test here is ultimately of only academic interest. Five years after Sears, the Supreme Court in Jones restated the local interest exception test in what “amount[ed] to a substantial reformulation of the Sears requirement that state and federal controversies be identical.” (Jones, supra, 460 U.S. at p. 688 (dis. opn. of Rehnquist, J.).) Because the facts of Jones are the most closely analogous to this case and it is the most recent Supreme Court precedent, it is worth examining Jones in detail. The plaintiff in Jones believed that a union had persuaded a company to fire him from his position as a supervisor. (Jones, supra, 460 U.S. at p. 672.) The plaintiff filed an unfair labor practice charge with the NLRB, but the NLRB regional director refused to issue a complaint. (Id. at p. 672.) The regional director explained in a letter to the plaintiff that the director found insufficient evidence that the union had caused the company to discharge the plaintiff. (Id. at pp. 672–673.) The plaintiff then filed a state court suit, alleging the union interfered with his contract with the company. (Id. at pp. 673–674.) Before the Supreme Court, the plaintiff argued in part that his state suit was not preempted because his state cause of action was

10 “distinct” from the unfair labor practice charge, like the non-preempted claims in Linn and Sears. (Id. at p. 681.) The plaintiff’s theory was that the NLRA only prohibited a union from coercing an employer’s choice of bargaining representative, while his state law claim could succeed if the union coercively or non-coercively caused the company to fire him. (Ibid.) The Court rejected this argument for several independent reasons. (Jones, supra, 460 U.S. at p. 682.) As relevant here, the Court noted the plaintiff conceded that a claim based on coercive conduct would be preempted, and it viewed his complaint as alleging coercive conduct. (Ibid.) Jones then held that even a claim for non-coercive interference with contract could not proceed in state court, because such a claim would require the state court to decide whether the union’s conduct was coercive or not and “[d]ecisions on such questions of federal labor law should be resolved by the Board.” (Ibid.) Jones further reasoned that a state law claim for non-coercive interference with contract was preempted because one element of such a claim—causation—would overlap with an element of an unfair labor practice charge. (Jones, supra, 460 U.S. at p. 682.) The Court explained, “[E]ven if the [state] law reaches non-coercive interference with contractual relationships, a fundamental part of such a claim is that the Union actually caused the discharge and hence was responsible for the employer’s breach of contract. Of course, this same crucial element must be proved to make out [an NLRA] case: the discharge must be shown to be the result of Union influence. Even on [the plaintiff’s] view of the elements of his state-law cause of action, the federal and state claims are thus the same in a fundamental respect, and here the Regional Director had concluded that the Union was not at fault.” (Ibid., italics added.) Because the plaintiff sought to relitigate the

11 question of causation in state court, the Court concluded “[t]he risk of interference with the Board’s jurisdiction [was] thus obvious and substantial.” (Id. at p. 683.) The Court noted that the issues in Sears, by contrast, were “ ‘completely unrelated’ ” and that there was “ ‘no realistic risk of interference with the Labor Board’s primary jurisdiction’ ” because the state law trespass claim in Sears turned only on the location of a union’s picketing, while an unfair labor practice charge based on the same picketing would have examined the union’s motives for the picketing. (Id. at pp. 682– 683.) The Court also stated that its precedents refuted the plaintiff’s argument that the availability of punitive damages or attorneys’ fees was a reason to allow his tort claim to proceed. (Id. at p. 684.) The majority opinion does not interpret Jones to create a crucial element test because such a test would “eviscerate” the local interest exception as set forth in Sears and, presumably, Linn. But Jones compared the crucial elements of the Jones plaintiff’s claim with the elements of an NLRA claim—rather than the overall controversies or the differing purposes of the laws in question, as the majority does—and found preemption based on an overlap of the single element of causation. (Jones, supra, 460 U.S. at p. 682.) Jones therefore construed the identical controversy test from Sears as turning on whether a “crucial element” of the state and NLRA claims is identical. I do not see how it is possible to read this as doing anything but modifying the local interest exception test. Consistent with this conclusion is Justice Rehnquist’s dissent in Jones, in which he noted that the Court’s opinion was a “substantial reformulation” of the Sears test. (Jones, at p. 688 (dis. opn. of Rehnquist, J.).) Reading Jones in this way does not require us to decide that Sears and Linn are no longer good law. In fact, precedent prevents us from so deciding, because the Supreme Court has instructed that

12 only the high Court itself can declare that a decision has been overruled. (Rodriguez de Quijas v. Shearson/Am. Ex., Inc. (1989) 490 U.S. 477, 484.) However, this same principle dictates that it is not our court’s role or within our power to interpret the Supreme Court’s precedents differently from the Court itself, even when we may disagree with those decisions on the merits. We must defer to the Supreme Court’s authority to construe its own caselaw. Sears and Linn now mean what Jones says they mean, and we must analyze the local interest exception as Jones did. Applying Jones’s construction of the local interest exception here is straightforward. Like the overlap found in Jones itself, plaintiffs’ claims all share a crucial element with the regional director’s NLRA complaint: whether defendants’ policies in fact prevent discussion of wages and working conditions with government agencies or other third parties. For example, one of Google’s defenses to both plaintiffs’ whistleblowing claims and the regional director’s complaint would likely be that a savings clause in its confidentiality policies permitted the disclosure of wages and working condition information to the government. The regional director’s issuance of an amended complaint implicitly demonstrates that she concluded the savings clause was insufficient, but the trial court could conclude the opposite. If the trial court were to deny plaintiffs’ PAGA claims based on the savings clause, such a ruling could undermine public confidence in the regional director’s complaint and the resulting settlement. As in Jones, “[t]he risk of interference with the Board’s jurisdiction is thus obvious and substantial.” (Jones, supra, 460 U.S. at p. 683.) Even if I were to ignore Jones’s reasoning, as the majority opinion seems to do, giving precedential effect only to Jones’s outcome would lead to the same result. The majority holds that plaintiffs can proceed with their

13 claims without also proving an NLRA violation because the NLRA only applies if defendants infringed on employees’ concerted actions and plaintiffs can prove their claims even if defendants infringed on non-collective activity. But this approach is analytically indistinguishable from the Jones plaintiff’s argument that his tort claim could proceed because the NLRA applied only if the union acted coercively and he could prove his tort claim by showing the union acted non-coercively. (Jones, supra, 460 U.S. at p. 682.) Jones definitively rejected this type of maneuver. (Ibid.) The majority also asserts that nothing the trial court may do with plaintiffs’ PAGA claims could interfere with the Board’s jurisdiction because the regional director settled her complaint. The same could be said of the Jones regional director’s decision not to issue a complaint at all, however, yet the Court still held the plaintiff’s state law claim preempted. (Jones, at pp. 673, 680–681.) Although it finds the crucial element test not to exist, the majority opinion nonetheless goes on to apply the test and finds insignificant the overlap between plaintiffs’ and the regional director’s complaints on the issue of causation. It concludes causation is merely “antecedent” to different “legal considerations” at issue for each claim—concerted activity for the NLRA and whistleblowing and free speech for plaintiff’s PAGA claims. (Maj. opn. ante, at p. 22.) This reasoning, too, founders on Jones. If this notion was correct, the Supreme Court would not have held the Jones plaintiff’s interference with contract claim was preempted. After all, the legal consideration for the state claim in Jones was whether the plaintiff had a valid employment contract as a supervisor and the union interfered with it, while the legal consideration for the NLRA claim was whether the union interfered with the employer’s selection of its bargaining representative. (Jones, supra, 460 U.S. at p. 681.)

14 I find similarly unconvincing the majority opinion’s four additional reasons for distinguishing Jones. The majority opinion first says Jones is distinguishable because the trial court here would not need to resolve issues of federal law to adjudicate plaintiffs’ complaint. (Maj. opn. ante, at pp. 22, 25, & fn. 5.) This is incorrect, as noted above. The majority opinion concludes plaintiffs’ complaint is not preempted specifically because it does not mention concerted activity. (Maj. opn. ante, at p. 27.) Therefore, to resolve plaintiffs’ claims on the merits while avoiding any intrusion into preempted areas, the trial court would need to decide the federal labor law question of whether plaintiffs’ evidence supporting their claims involves concerted or collective activity, just as the trial court in Jones would have had to rule on the federal issue of coercion to avoid preemption. (Jones, supra, 460 U.S. at p. 682.) The majority opinion next asserts that the regional director here has not made factual findings fatal to the plaintiff’s claim, as did the regional director in Jones. Putting aside the question of whether the regional director’s rejection letter in Jones actually constituted factual findings (Jones, supra, 460 U.S. at pp. 672–673), a conflict in factual findings can arise either from the NLRB rejecting a claim that a state court allows to proceed (as in Jones) or from the Board accepting a claim that a state court rejects (a possibility here, as discussed above). Moreover, even if the outcome of plaintiffs’ suit is ultimately consistent with the regional director’s settlement, “the Garmon rule prevents States not only from setting forth standards of conduct inconsistent with the substantive requirements of the NLRA, but also from providing their own regulatory or judicial remedies for conduct prohibited or arguably prohibited by the Act. [Citation.] The rule is designed to prevent ‘conflict in its broadest sense’ with the ‘complex and interrelated

15 federal scheme of law, remedy, and administration,’ [citation], and [the Supreme] Court has recognized that ‘[c]onflict in technique can be fully as disruptive to the system Congress erected as conflict in overt policy.’ ” (Wisconsin Dept. of Industry v. Gould, Inc. (1986) 475 U.S. 282, 286 (Gould, Inc.).) As a third basis for distinguishing Jones, the majority says defendants’ conduct here was not protected by the NLRA like the union’s conduct in Jones, and federal supremacy is implicated to a greater extent when a state court tries to prohibit what federal law protects. Jones did note that the Act arguably protected the union’s conduct there, but it was a separate basis for finding preemption, not a prerequisite for preemption. (Jones, supra, 460 U.S. at pp. 683–684.) Instead, the Court’s opinion makes clear it would have found preemption based solely on the overlap of the claims on the causation element, regardless of whether the case for preemption were stronger for arguably protected conduct. (Jones, supra, 460 U.S. at pp. 682– 683.) Finally, the majority opinion notes the absence of a union in this case and quotes the statement in Sears that preemption “has its greatest force when applied to state laws regulating the relations between employees, their union, and their employer.” (Maj. opn. ante, at p. 23; see ante, at pp. 10, 25.) But the Sears remark as to the reasoning behind preemption was intended only to compare labor regulations with “certain laws of general applicability which are occasionally invoked in connection with a labor dispute,” not to imply that Garmon preemption operates differently in unionized and non- unionized workplaces. (Sears, supra, 436 U.S. at p. 193.) The majority opinion cites nothing to support its suggestion that Garmon preemption is

16 less necessary when a union is not involved.3 The regional director’s actions in this case unequivocally demonstrate the Act applies to non-unionized employees seeking to improve their working conditions just as it does to unionized employees. (See Eastex, Inc. v. NLRB, supra, 437 U.S. at p. 565 [Congress used the phrase “ ‘mutual aid or protection’ ” in the Act because it “knew well enough that labor’s cause often is advanced on fronts other than collective bargaining and grievance settlement”].) IV. The risk of interference with the NLRB’s jurisdiction I agree with the majority opinion that Jones does not provide any bright-line rules for determining what elements qualify as crucial for purposes of the local interest exception test. (Maj. opn. ante, at pp. 21–22) Instead, Jones directs us to consider whether a state suit poses a “realistic risk of interference with the Labor Board’s primary jurisdiction to enforce the statutory prohibition against unfair labor practices,” with this interference coming “either in terms of negating the Board’s exclusive jurisdiction or in terms of conflicting substantive rules.” (Jones, supra, 460 U.S. at pp. 676, 683.) When considering the risk of such interference here, it bears emphasizing that Doe himself invited the NLRB to take action against Google and the regional director’s settlement has already caused Google to change the same policies about which plaintiffs now complain, a point the majority opinion mentions only in passing in its discussion of the factual

3 Epic Sys. Corp. v. Lewis (2018) __ U.S. __, 138 S.Ct. 1612, 163, is not on point, as it had nothing to do with preemption. Epic sought to reconcile competing interpretations of the NLRA and the Federal Arbitration Act. (Id. at pp. 1629–1630.) It did not discuss the Act’s protections for non- unionized employees and certainly did not hold that only the Act’s provisions applicable to union certification or bargaining can support Garmon preemption, as the majority opinion seems to imply. (Maj. opn. ante, at pp. 15, 23.) 17 background of the case. (Maj. opn. ante, at p. 8.) By ruling against preemption, then, the majority opinion is allowing plaintiffs to seek additional penalties for the same conduct that the regional director has already remedied. Because it could allow plaintiffs to impose monetary penalties for practices the Board decided to remedy via settlement, plaintiffs’ PAGA suit poses a substantial risk of interfering with the NLRB’s jurisdiction. (Gould, Inc., supra, 475 U.S. at p. 287 [states may not impose additional penalties for conduct the NLRA prohibits].) The majority opinion’s responses to this risk are unpersuasive. The majority opinion notes that the regional director’s settlement with Google was informal and required Google to post a notice of employees’ rights under federal law. The settlement required Google to do more than post a notice. The notice stated that Google had rescinded the policies about which the regional director complained—the same policies at issue in plaintiffs’ complaint—and the settlement agreement required Google to comply with that statement. That aside, the settlement was informal only in the sense that it did not result in a Board order. There was still a formal settlement agreement and Doe had an opportunity to appeal that settlement to the Board. (29 C.F.R. §§ 101.7, 101.9(b)(2).) Moreover, any informality would serve only to highlight the risk of interference. The regional director opted for an informal settlement in exchange for Google’s withdrawal of the offending sections of its policies and because of the absence of any significant history at Google of unfair labor practices. Whatever one might think of the merits of this decision, the specter of heavy PAGA penalties threatens to thwart the regional director’s choice of leniency. The majority opinion also notes that the settlement stated it would not prevent the Board or courts from finding violations with respect to matters

18 occurring before the agreement was approved or making findings of fact or conclusions of law regarding evidence obtained in the case. The provision quoted by the majority (on which plaintiffs do not rely) is from a form agreement prepared by the Board, apparently intended to serve as a standard template for all informal settlements. (See 29 C.F.R. §§ 101.7, 101.9(b)(2).) The provision nowhere states that it was intended to affect the reach of Garmon preemption or to allow state claims to proceed that would otherwise be barred. At a minimum, the settlement agreement does not define the scope of the Act’s preemptive force “with unclouded legal significance.” (Garmon, supra, 359 U.S. at p. 246.) In analogous circumstances, the Supreme Court has instructed that a failure “to define the legal significance under the Act of a particular activity does not give the States the power to act.” (Ibid.) In this case, I see no Board actions of sufficient clarity to permit plaintiffs’ state court claims to intrude into areas that threaten to interfere with the reach of the Board’s jurisdiction. For these reasons, I respectfully dissent.

_________________________________ BROWN, J.

19 Trial Court: City & County of San Francisco Superior Court

Trial Judge: Hon. Curtis E.A. Karnow; Hon A.C. Massullo

Counsel for Appellants: Baker Curtis & Schwartz, P.C.; Chris Baker and Deborah Schwartz

Counsel for Respondents: Google Inc. and Alphabet, Inc.: Paul Hastings LLP; J. Al Latham, Jr., Cameron W. Fox, and Ankush Dhupar

Adecco USA, Inc.: Jackson Lewis P.C.; Mia Farber, Adam Y. Siegel, Scott P. Jang, Dylan B. Carp