Berkeley v. Intel Corporation

• Court: District Court, N.D. California
• Decided: June 27, 2025
• Docket: 5:23-cv-00343
• Status: Unknown

Opinion

1 2 3 4 UNITED STATES DISTRICT COURT 5 NORTHERN DISTRICT OF CALIFORNIA 6 SAN JOSE DIVISION 7 8 Case No. 5:23-cv-00343-EJD GREGG BERKELEY,

9 Plaintiff, ORDER GRANTING MOTION FOR CLASS CERTIFICATION 10 v. Re: Dkt. No. 76 11

12 INTEL CORPORATION, et al.,

13 Defendants. 14 Plaintiff Gregg Berkeley (“Berkeley”) brings this class action against Defendants Intel 15 Corporation and the Administrative Committee of the Intel Minimum Pension Plan (collectively, 16 “Intel”) alleging violations of the Employee Retirement Income Security Act of 1974 (“ERISA”) 17 and breaches of Intel’s fiduciary duties. Compl., ECF No. 1. Before the Court is Berkeley’s 18 motion for class certification. Mot., ECF No. 76; Opp’n, ECF No. 80; Reply, ECF No. 83. The 19 Court held a hearing on May 1, 2025, and heard oral arguments from all parties. ECF No. 87. For 20 the reasons stated below, the Court GRANTS Berkeley’s motion. 21 I. BACKGROUND 22 Berkeley and a proposed class of approximately 1,847 Intel retirees or their surviving 23 spouses allege Intel violated ERISA by converting their single life annuity (“SLA”) to a joint and 24 survivor annuity (“JSA”) using unreasonable actuarial assumptions in the Intel Minimum Pension 25 Plan (“MPP”). Mot. 1. 26 The proposed class is defined as follows: 27 All Plan participants and beneficiaries who are receiving a joint and survivor annuity (or, for beneficiaries whose spouses died 1 before commencing benefits, a pre-retirement survivor annuity) which is less than the value of the single life annuity converted to 2 a joint and survivor annuity using the interest rates and mortality tables set forth in 26 U.S.C. § 417(e) with an annual stability and 3 August lookback period. 4 Id. at 10. 5 An SLA is an annuity payable monthly upon retirement for the duration of the participant’s 6 life. MPP § 5(d)(i), ECF No. 76-7. SLAs are the default benefit for unmarried participants. Id. 7 §§ 5(d)(ii)–(iii). A JSA is an annuity payable monthly where different portions of retirees’ 8 monthly payments continue to their surviving spouse or designated beneficiary after their death. 9 Id. Because a JSA provides a stream of payments over the course of two lives, the monthly 10 payment may be adjusted downward. Expert Report of Ian Altman (“Altman Report”) 7–8, ECF 11 No. 86-1. 12 Under ERISA, JSAs for married retirees must be “actuarily equivalent” to SLAs for single 13 retirees. 26 C.F.R. § 1.401(a)-11(b)(2). To accomplish this, pension plan administrators may use 14 certain “reasonable actuarial factors” to convert SLA payments to JSA payments, i.e., mortality 15 rates and interest rates. Id. The MPP currently converts SLAs to JSAs using the mortality table 16 published by the Society of Actuaries in 1983 (“GAM-83 mortality table”), customized with an 17 adjusted gender blend and interest rates set by the Pension Benefit Guaranty Corporation (“PBGC 18 interest rates”). Id. § 2(a); First Am. to Plan Document, ECF No. 76-8. 19 Berkeley contends that the GAM-83 mortality table and PBGC interest rates are outdated 20 and unreasonable actuarial assumptions that do not create “actuarily equivalent” benefits. To 21 comply with ERISA, Berkeley argues the MPP must convert SLAs to JSAs using the higher 22 interest rates and mortality tables set forth in 26 U.S.C. § 417(e). In connection with § 417(e), the 23 government periodically publishes updated mortality studies prepared by the Society of Actuaries 24 and interest rates reflecting contemporary market conditions.1 Altman Report 19–20. 25

26 1 The Court notes that this case presents a relatively new theory of liability. The first cases alleging that a plan used unreasonable actuarial assumptions to convert SLAs to JSAs under 26 27 C.F.R. § 1.401(a)-11(b)(2) arose in 2018, and there are few cases examining class certification on this theory. See, e.g., Urlaub v. Citgo Petroleum Corp., No. 21-cv-5133, 2024 WL 2209538 II. LEGAL STANDARD 1 Rule 23 sets forth the two-step process for certifying class actions. First, a plaintiff must 2 establish: 3 (1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the 4 class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the 5 representative parties will fairly and adequately protect the interests of the class. 6 Fed. R. Civ. P. 23(a). Second, the plaintiff must show that the proposed class fits into one of the 7 three categories of Rule 23(b). Relevant here, the first category in Rule 23(b)(1) is satisfied if: 8 prosecuting separate actions by or against individual class 9 members would create a risk of: (A) inconsistent or varying adjudications with respect to individual class members that would 10 establish incompatible standards of conduct for the party opposing the class; or (B) adjudications with respect to individual class 11 members that, as a practical matter, would be dispositive of the interests of the other members not parties to the individual 12 adjudications or would substantially impair or impede their ability to protect their interests. 13 Fed. R. Civ. P. 23(b)(1). 14 To meet their obligations under Rule 23, plaintiffs “must actually prove—not simply 15 plead—that their proposed class satisfies each requirement of Rule 23” by a preponderance of the 16 evidence. Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 31 F.4th 651, 664–65 17 (9th Cir. 2022) (en banc) (citation omitted), cert. denied, 143 S. Ct. 424 (2022). Courts must 18 conduct a “rigorous” analysis of the Rule 23 factors that will often “entail some overlap with the 19 merits of the plaintiff’s underlying claim.” Ellis v. Costco Wholesale Corp., 657 F.3d 970, 980 20 (9th Cir. 2011) (quoting Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 351 (2011)). However, 21 “Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification 22 stage. Questions of merit may be considered to the extent—but only to the extent—that they are 23 24

25 (N.D. Ill. May 16, 2024); Berube v. Rockwell Automation, Inc., No. 20-C-1783, 2023 WL 8468522, at *3 (E.D. Wis. Apr. 3, 2023); McAlister v. Metro. Life Ins. Co., No. 18-cv-11229, 26 2023 WL 5769491 (S.D.N.Y. Sept. 7, 2023); Thorne v. U.S. Bancorp, No. CV 18-3405 (PAM/KMM), 2021 WL 1977126, at *2 (D. Minn. May 18, 2021); Torres v. Am. Airlines, Inc., 27 No. 4:18-CV-00983-O, 2020 WL 3485580, at *9 (N.D. Tex. May 22, 2020). 1 relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” 2 Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 466 (2013). 3 III. DISCUSSION 4 The Court finds that Berkeley has carried his burden under Federal Rules of Civil 5 Procedure 23(a) and (b)(1) and GRANTS his motion to certify the class for the reasons stated 6 below. 7 A. Rule 23(a) Requirements 8 As the Court recited above, Rule 23(a) requires that Berkeley establish: (1) numerosity, (2) 9 commonality, (3) typicality, and (4) adequacy. Intel only challenges commonality, typicality, and 10 adequacy. 11 1. Commonality 12 Commonality requires there to be “questions of law or fact common to the class.” Fed. R. 13 Civ. P. 23(a)(2). A common question is one that “is capable of classwide resolution,” meaning 14 that answering the question “will resolve an issue that is central to the validity of each one of the 15 claims in one stroke.” Wal-Mart, 564 U.S. at 350. The key factor is whether the common 16 question generates “common answers.” Id. However, plaintiffs need not show that all questions 17 are common; they need only identify a “single significant question of law or fact.” Stockwell v. 18 City & Cnty. of San Francisco, 749 F.3d 1107, 1111 (9th Cir. 2014). 19 The Court finds the commonality requirement satisfied. The overarching questions 20 relevant to all claims include (1) whether ERISA requires “reasonable” actuarial assumptions for 21 SLA to JSA conversions, and (2) whether the actuarial assumptions Intel used in the conversion 22 calculation failed to provide actuarial equivalence between the two forms of benefit. To prove 23 these claims and to measure losses, Berkeley presents evidence that he will compare the JSAs 24 class members received under the MPP’s current actuarial assumptions to the JSAs they would 25 have received if the MPP used the assumptions set forth in § 417(e). See Altman Report 24–27. 26 This is sufficient to establish commonality. See, e.g., McAlister v. Metro. Life Ins. Co., No. 18-cv- 27 1229, 2023 WL 5769491, at *3 (S.D.N.Y. Sept. 7, 2023) (finding common questions including 1 which actuarial assumptions should be used to determine whether the JSAs were actuarially 2 equivalent to the SLAs, whether the conversion factors improperly reduced the JSAs, and whether 3 ERISA requires that JSAs be “actuarily equivalent”); Urlaub v. CITGO Petroleum Corp., 348 4 F.R.D. 319, 324 (N.D. Ill. 2024) (finding common questions including whether assumptions 5 resulted in reasonable SLA-to-JSA conversion for all class members).2 6 Intel’s arguments to the contrary are unpersuasive. First, Intel argues that there is no 7 common answer to whether the MPP’s assumptions were reasonable for the entire 30-year class 8 period because reasonableness requires context. Regarding mortality rates, Intel presents evidence 9 that mortality tables are constructed with a significant margin to account for future mortality 10 improvements—though the GAM-83 mortality table was created in 1983, it was designed to be 11 applicable well past that date. Therefore, Intel argues, there will be different answers to whether 12 the mortality table was reasonable in 1995, 2005, and 2015. Similarly for interest rates, Intel 13 presents evidence that § 417(e) itself was based on PBGC interest rates up until 2000, after which 14 time the IRS changed its approach; therefore, there will also be a different answer to whether the 15 PBGC rates were reasonable before and after 2000. 16 The Court finds that these issues do not defeat commonality. Evidence at trial may show 17 that Intel did not violate ERISA or breach its fiduciary duties throughout the entire class period, 18 but these context-specific inquiries do not present individualized issues. The question applicable 19 to the entire class remains whether it is unreasonable to use the GAM-83 mortality table and 20 PBGC interest rates to convert SLAs to JSAs, and if so, at what point in time it became 21 unreasonable. Though answers may differ over time, they will not differ according to the facts of 22 individual class members. 23 Next, Intel argues that there is no common answer to whether any class member was 24 “underpaid” because there are multiple variations of the applicable interest rate. Berkeley seeks to 25

26 2 The Court finds the distinctions Intel draws between these cases and Berkeley’s motion immaterial. Though the defendants there suffered from a failure to meet their burden of proof, 27 the Court still finds that, even when supported by evidence, Intel’s arguments do not defeat commonality. 1 impose interest rates as defined under § 417(e), which has fifteen variations determined by the 2 “stability period” and “lookback month” chosen by plan sponsors. 26 C.F.R. § 1.417(e)- 3 1(d)(4)(iii-iv); Altman Report 20. Berkeley asks the Court to certify a class using a § 417(e) 4 interest rate with an annual stability period and August lookback month. This means that when 5 participants retire, they will receive the interest rate from August the year prior. Intel argues that, 6 depending on the chosen stability period and lookback month, the rate used to calculate a lump 7 sum could vary up to 3.3% for class members. At oral argument, Intel also contended that the 8 Court will be unable to determine whether Intel’s conduct was unreasonable based on just one § 9 417(e) variation because it could have been reasonable under an alternative variation. 10 The Court finds that the variations in stability periods and lookback months also do not 11 defeat commonality. Berkeley’s expert decided to use this particular stability and lookback period 12 because it is the one currently adopted in the MPP, which provides: “For purposes of Sections 13 5(a)(v), 6(c), 9(h) and 9(p), the actuarial assumptions used are the Applicable Mortality Table and 14 the Applicable Interest Rate for the month of August preceding the calendar year in which the 15 Annuity Starting Date occurs.” MPP Amendment § 1(a), ECF No. 76-8 (emphasis added); 16 Altman Report 20–21. The Court finds that choosing to maintain the same stability and lookback 17 period is logical, and more notably, it isolates the issue at the center of Berkeley’s allegations—the 18 impact of the actuarial assumptions. See Urlaub, 348 F.R.D. at 327 (rejecting arguments 19 regarding the stability and lookback period variations in part because the plaintiff provided logical 20 reasons to choose the particular periods). Intel may present evidence at trial that its actuarial 21 assumptions were reasonable or that the damages are reduced when applying alternative stability 22 and lookback periods, but the common questions identified above remain. See, e.g., Pedersen v. 23 Kinder Morgan Inc., 345 F.R.D. 302, 320 (S.D. Tex. 2024) (“[T]he common question of whether 24 the GAM-83 mortality table and 8% interest rate violate ERISA's actuarial equivalence 25 requirement pervades the subclass. The fact that each subclass member may have suffered a 26 different degree of harm as a result of Defendants' actuarial assumptions does not defeat 27 commonality; the source of their harm was the same.”); McAlister, 2023 WL 5769491, at *5 1 (“[T]he possibility that the adoption of Defendants' alternative methodology would result in a 2 varied impact among purported class members does not defeat certification, even if it may 3 ultimately necessitate a different result on the merits of Plaintiffs' claims.”) (internal quotation 4 marks omitted). 5 The Court also finds these facts distinguishable from Thorne, which Intel relies on for the 6 proposition that alternative assumptions defeat commonality. There, the plaintiffs’ expert offered 7 six different models of actuarial equivalence, yet “no model result[ed] in higher benefits for all 8 class members and each model result[ed] in lower benefits for some class members.” Thorne v. 9 U.S. Bancorp, No. CV 18-3405 (PAM/KMM), 2021 WL 1977126, at *2 (D. Minn. May 18, 10 2021). Here, Berkeley has proposed one model based on the existing stability and lookback 11 period in the MPP to show that applying the new actuarial assumptions in § 417(e) would result in 12 an increased benefit to all class members. Altman Report 24–27. While Intel asserts that other 13 reasonable permutations of § 417(e) create different cutoffs for actuarial equivalence, 14 “[u]ncertainty regarding class members’ damages does not prevent certification of a class as long 15 as a valid method has been proposed for calculating those damages.” Lambert v. Nutraceutical 16 Corp., 870 F.3d 1170, 1182 (9th Cir. 2017), rev’d on other grounds and remanded, 586 U.S. 188 17 (2019). 18 2. Typicality 19 To satisfy Rule 23’s typicality requirement, class representatives must show that their 20 claims are “reasonably coextensive with those of absent class members,” though their claims 21 “need not be substantially identical.” Just Film, Inc. v. Buono, 847 F.3d 1108, 1116 (9th Cir. 22 2017). But when “there is a danger that absent class members will suffer [because] their 23 representative is preoccupied with defenses unique to it,” a class representative is not typical. 24 Hanon v. Dataprods. Corp., 976 F.2d 497, 508 (9th Cir. 1992) (citation omitted). 25 The Court finds typicality satisfied. Berkeley and the proposed class’s claims arise from 26 the same course of events because each person is receiving a JSA that Intel converted from an 27 SLA using the same challenged actuarial assumptions. Berkeley and each class member also 1 make the same legal arguments to prove Intel’s liability—alleging that the benefits they receive 2 are less than the benefits they are entitled to under ERISA’s actuarial equivalence requirement; 3 that Intel illegally caused them to forfeit benefits; and that Intel’s conduct was a breach of its 4 fiduciary duties. 5 Intel argues that Berkeley’s claims are not typical of the class because he has two unique 6 defenses. First, Intel contends that Berkeley will be preoccupied by the defense that his alleged 7 underpayment is approximately equal in value with the JSA. Berkeley’s alleged underpayment is 8 just 2.6% of his JSA benefit, but federal regulations provide that an SLA and JSA with an 9 actuarial present value within 5% of one another are “approximately equal in value.” Altman 10 Report 16; 26 C.F.R. § 1.417(a)(3)-1(c)(2)(iii)(C). Second, Intel contends Berkeley will be 11 preoccupied by the defense that Berkeley benefited from one aspect of the plan he now seeks to 12 change—setting interest rate at termination rather than benefit commencement. Although 13 Berkeley argues in his motion that the MPP unreasonably sets an interest rate at termination rather 14 than benefit commencement, Intel argues that Berkeley’s benefits would have been less had the 15 MPP applied the interest rate at benefit commencement rather than termination. 16 The Court finds that these issues do not defeat typicality. “Typicality refers to the nature 17 of the claim or defense of the class representative, and not to the specific facts from which it arose 18 or the relief sought.” Ellis, 657 F.3d at 984. While it is true that these defenses may not apply to 19 each class member’s claims, they apply to more class members than just Berkeley. See Lyon v. 20 U.S. Immigr. & Customs Enf’t., 308 F.R.D. 203, 213 (N.D. Cal. 2015) (“[I]f other class members 21 may be subject to the same defenses as the representative plaintiff, such defenses are not ‘unique’ 22 and therefore do not defeat typicality.”); see also Wolin v. Jaguar Land Rover N. Am., LLC, 617 23 F.3d 1168, 1175 (9th Cir. 2010) (“Typicality can be satisfied despite different factual 24 circumstances . . . . ”); Ellis, 657 F.3d at 984 (“Typicality refers to the nature of the claim or 25 defense of the class representative, and not to the specific facts from which it arose or the relief 26 sought.”). Berkeley presents evidence that 43% of class members’ underpayments fall under the 27 5% threshold and argues that other class members could also theoretically maximize their 1 individual damages by utilizing the interest rate at the time of their terminations rather than benefit 2 commencement. Sutter Decl. ¶ 2, ECF No. 83-1. The Court also finds that these defenses are not 3 so substantial that they risk becoming the focus of this case. Unlike the cases Intel cites, Berkeley 4 is a participant in the same plan as every other class member and raises the same theories of 5 liability such that the defenses Intel identifies do not make Berkeley atypical of the class. Contra, 6 e.g., Berube v. Rockwell Automation, Inc., No. 20-C-1783, 2023 WL 8468522, at *3 (E.D. Wis. 7 Apr. 3, 2023) (finding no typicality because the plaintiff was only a member in one of two plans in 8 the class definition). 9 3. Adequacy 10 The adequacy analysis under Rule 23(a)(4) requires that the class representatives and class 11 counsel “fairly and adequately protect the interests of the class.” Fed. R. Civ. P. 23(a)(4). “The 12 proper resolution of this issue requires that two questions be addressed: (a) do the named plaintiffs 13 and their counsel have any conflicts of interest with other class members and (b) will the named 14 plaintiffs and their counsel prosecute the action vigorously on behalf of the class?” In re Mego 15 Fin. Corp. Sec. Litig., 213 F.3d 454, 462 (9th Cir. 2000), as amended (June 19, 2000) (citing 16 Lerwill v. Inflight Motion Pictures, Inc., 582 F.2d 507, 512 (9th Cir. 1978)). This inquiry often 17 merges with other Rule 23(a) requirements, such as commonality and typicality, as they all 18 concern whether the named plaintiff and the named plaintiff’s claims are so interrelated with those 19 of the class members that they will be protected in their absence. See In re ConAgra Foods, Inc., 20 90 F. Supp. 3d 919, 974 (C.D. Cal. 2015), aff’d sub nom., Briseno v. ConAgra Foods, Inc., 674 F. 21 App’x 654 (9th Cir. 2017), and aff’d sub nom., Briseno v. ConAgra Foods, Inc., 844 F.3d 1121 22 (9th Cir. 2017) (citing Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 158 (1982)). 23 The Court finds that Berkeley and his counsel have no conflicts of interest with other class 24 members. Berkeley alleges that all class members, including himself, received lower pension 25 payments than they would have otherwise received had Intel converted their SLAs into JSAs using 26 the actuarial assumptions under § 417(e). Berkeley also proposes a theory and loss methodology 27 that he argues will result in increased pension payments to all class members. See, e.g., Urlaub, 1 2024 WL 2209538, at *5 (finding no conflicts because the class members “share ‘the same 2 interest’ and the ‘same injury’” (citation omitted)); Kernan, 2023 WL 3620884, at *7–8 (finding 3 Rule 23(a) factors met because under the plaintiffs’ methodology, the named plaintiffs showed 4 that the JSAs available were not actuarially equivalent to the SLAs). The Court further finds that 5 Berkeley and his counsel will prosecute this action vigorously on behalf of the class. Berkeley has 6 actively participated in this litigation by reviewing the complaint, responding to document 7 requests, sitting for a deposition, and remaining in communication with counsel; and Berkeley’s 8 counsel has a demonstrated knowledge of ERISA and history of bringing similar actuarial 9 equivalence cases across the country. See Berkeley Decl. ¶¶ 5–8, ECF No. 76-3; Yau Decl. ¶¶ 3– 10 19, ECF No. 76-1; see also Californians for Disability Rts., Inc. v. Cal. Dep’t of Transp., 249 11 F.R.D. 334, 349 (N.D. Cal. 2008) (“The threshold of knowledge required to qualify a class 12 representative is low; a party must be familiar with the basic elements of her claim, and will be 13 deemed inadequate only if she is ‘startingly unfamiliar’ with the case.” (citation omitted)). This is 14 sufficient to establish adequacy. 15 Intel’s arguments to the contrary are unpersuasive. Intel argues that Berkeley is an 16 inadequate class representative because: (1) Berkeley’s preferred assumptions would have harmed 17 some class members; (2) Berkeley’s stability and lookback selection would have harmed some 18 class members; and (3) other various conflicts. The Court addresses each argument in turn. 19 a. Preferred Mortality and Interest Rate Assumptions 20 Intel argues that Berkeley’s preferred mortality and interest rate assumptions would harm 21 some class members when applied to the entire JSA conversion calculation because the MPP is an 22 offset plan. To contextualize Intel’s arguments, the Court will briefly summarize the two phases 23 of the MPP’s JSA conversion calculation: (1) calculate the net SLA using the floor-offset plan, 24 and (2) convert the SLA to a JLA. See Expert Report of Thomas S. Terry (“Terry Report”), ECF 25 No. 80-4. 26 Phase one is a three-step calculation of net SLA benefits using the MPP’s floor-offset plan. 27 First, the MPP calculates the minimum SLA gross benefit based on years of service and pay. Id. 1 ¶¶ 48–53, 114–17. For Berkeley, this is $6,424 per month after the age of sixty-five. Id. Next, 2 the MPP calculates the “RC Plan offset.” Id. The RC Plan is a retirement contribution plan 3 designed to make up the difference if an employee’s RC Plan drops below a specified minimum 4 value. Id. The MPP calculates the RC Plan offset using the GAM-83 mortality table and PBGC 5 interest rates. Id. For Berkeley, he had an RC Plan account balance of $870,293. Id. After 6 applying the GAM-83 mortality table and PBGC rate to calculate the value of his RC Plan if it 7 were an SLA, this would be valued at $5,002 per month. Id. Finally, the RC Plan offset is 8 subtracted from the SLA gross benefit to calculate the SLA net benefit. Id. For Berkeley, the 9 amount calculated in step two ($5,002 monthly) would be subtracted from the gross benefit 10 calculated in step one ($6,424). Id. This results in a “net benefit” SLA of $1,422 monthly. Id. 11 In phase two, the MPP converts the SLA to a JSA by applying the same GAM-83 mortality 12 table and PBGC interest rates used in phase one. Id. For Berkeley, this would convert his $1,422 13 SLA net benefit to a $1,294 JSA benefit. Id. 14 When examining the calculations in both phases, Intel argues that because the RC Plan 15 offset is also calculated using the GAM-83 mortality table and the PBGC interest rates, Berkeley’s 16 proposed assumptions (longer mortality, higher interest rates) would result in a higher offset in 17 phase one—and therefore a decrease in net JSA benefits in phase two. Intel further argues that 18 this is the reason class members asked Intel to not change the actuarial assumptions in 2020. At 19 that time, the PBGC announced that it would no longer publish the PBGC interest rate and instead 20 adopt the higher interest rates in § 417(e). Plan participants expressed concern that this would 21 eliminate most or all of their benefits, and in response, Intel announced that the MPP would be 22 amended so the new PBGC interest rates methodology would not impact the MPP. 23 Intel is correct that if Berkeley sought to change the GAM-83 mortality table and PBGC 24 interest rates in both phases of the calculation, this would injure the class and make him an 25 inadequate class representative. However, that is not the remedy Berkeley seeks—Berkeley only 26 asks that Intel use the § 417(e) assumptions in the SLA to JSA conversion in phase two, keeping 27 the GAM-83 mortality table and PBGC interest rates the same in the floor-offset plan in phase 1 one.3 2 Berkeley contends that this would be a simple remedy. Although the MPP contains a 3 single, plan-wide definition of “actuarial equivalence” that applies the same GAM-83 mortality 4 table and PBGC interest rate in all calculations, it also includes a list of exceptions when required 5 by law: “For purposes of Sections 5(a)(v), 6(c), 9(h) and 9(p), the actuarial assumptions used are 6 the Applicable Mortality Table and the Applicable Interest Rate for the month of August 7 preceding the calendar year in which the Annuity Starting Date occurs.” MPP Amendment § 1(a). 8 At the hearing, Berkeley suggested amending the MPP to provide: “For purposes of Sections 9 5(a)(ii), 5(a)(iii), 5(a)(v), 6(c), 9(h) and 9(p) . . . . ” See 5/1/2025 Tr. 8–9, ECF No. 90. Sections 10 5(a)(ii) and (iii) are the provisions defining JSAs. MPP §§ 5(a)(ii), (iii). Berkeley argues that 11 Intel must add these sections to the list because it is also required by law—while using the GAM- 12 83 mortality table and PBGC interest rates in phase one may be appropriate, using these outdated 13 assumptions in phase two violates ERISA because they do not generate actuarially equivalent 14 JSAs. 15 To this point, Intel argues that the Court cannot create an arbitrary carve out to the MPP’s 16 definition of actuarial assumptions in only one phase of the calculation. Intel cites to Torres v. 17 Am. Airlines, No. 18-cv-983, 2020 WL 3485580 (N.D. Tex. May 22, 2020), where the court 18 declined to consider changing the actuarial assumptions in just one part of the plan while leaving 19 the allegedly outdated assumptions in other sections. However, the plan in Torres contained a 20 single definition of actuarial equivalence with no exceptions, and the plaintiff argued that the 21 existing actuarial assumptions were antiquated and unreasonable when applied to any part of the 22 plan. Id., at *9. Here, the MPP already excludes several provisions from the general definition, 23 and Berkeley seeks to simply add the JSA conversion to that list. Further, Berkeley concedes that 24 the GAM-83 mortality table and PBGC interests rates may be considered reasonable in phase one, 25 but when used in phase two, Berkeley argues they violate ERISA by creating JSAs that are not 26

27 3 Indeed, Berkeley highlights that he is unable to change the definition in phase one because the SLA is statutorily protected from reduction. See 29 U.S.C. §§ 1053(a), 1054(g). 1 actuarily equivalent to the SLAs. Though the Court makes no findings regarding the merits of 2 Berkeley’s remedy, the Court finds that it does not conflict with the interests of the class. 3 b. Preferred Stability and Lookback Selection 4 The Court discussed in detail above Intel’s issues with Berkeley’s preferred stability and 5 lookback selections. Specific to adequacy, Intel also argues that Berkeley’s choice presents a 6 conflict with the class because some members may benefit more from a different variation. 7 The Court also finds that this issue does not defeat adequacy. There is no evidence to 8 suggest that Berkeley chose to maintain the MPP’s existing lookback and stability period because 9 it benefited him more than other class members, and there is nothing in the record to suggest that 10 this choice impacts his ability to represent the class. See, e.g., In re: College Athlete NIL 11 Litigation, No. 20-cv-3919, 2023 WL 8372787, at *18 (N.D. Cal. Nov. 3, 2023) (“That [the] 12 model allocates slices of the damages pie in ways that benefit some class members more than 13 others is not problematic because [the expert] provides a rationale for this proposed allocation that 14 a reasonable [factfinder] could accept.”); Monaco v. Bear Stearns Cos., Inc., No. 09-cv-5438, 15 2012 WL 10006987, at *11 (C.D. Cal. Dec. 10, 2012) (“[T]hat some individual plaintiffs would 16 fare better under alternative theories and that some class members might receive no recovery 17 under one of those alternatives does not render their claims ‘antagonistic or conflicting.’”); Vogt v. 18 State Farm Life Ins. Co., 963 F.3d 753, 768 (8th Cir. 2020) (“[E]ven if there are slightly divergent 19 theories that maximize damages for certain members of the class, ‘this slight divergence is greatly 20 outweighed by shared interests in establishing [defendant’s] liability.’”). 21 c. Other Conflicts of Interest 22 The Court similarly rejects Intel’s additional arguments that “other conflicts” of interest 23 defeat adequacy. Opp’n 24. The Court finds that Berkeley’s alleged 2.6% underpayment, the 24 statute of limitations for the breach of fiduciary duty, and the choice to set interest rates at the 25 beginning of employment rather than termination do not present conflicts of interest or undermine 26 Berkeley’s and class counsel’s willingness to prosecute this action vigorously on behalf of the 27 class. 4. Numerosity 1 Although not challenged, the Court also finds numerosity met. There is no exact 2 numerical cut-off for when a proposed class is “so numerous that joinder of all the class members 3 is impracticable,” but courts often will presume that this requirement is satisfied when the class 4 exceeds forty members. Fed. R. Civ. P.23(a)(1); see, e.g., Arroyo v. Int’l Paper Co., No. 17-cv- 5 6211, 2019 WL 1508457, at *2 (N.D. Cal. Apr. 4, 2019). Here, the class includes approximately 6 1,847 members, making joinder impracticable. The Court finds this sufficient to meet the 7 numerosity requirement. 8 B. Rule 23(b)(1) Requirements 9 Intel also does not challenge Berkeley’s arguments regarding Rule 23(b). As the Court 10 recited above, Berkeley must show that the class can be certified under one of Rule 23(b)’s three 11 categories. The first category in Rule 23(b)(1) is generally satisfied when prosecuting separate 12 actions would create a risk of either inconsistent adjudications or adjudications that would impact 13 the interests of other members who are not parties to the litigation. Fed. R. Civ. P. 23(b)(1). 14 ERISA actions are typically “paradigmatic” Rule 23(b)(1) cases. Kanawi v. Bechtel Corp., 254 15 F.R.D. 102, 112 (N.D. Cal. 2008) (citation omitted); see also McAlister, 2023 WL 5769491, at *8 16 (“The language of subdivision (b)(1)(A), addressing the risk of ‘inconsistent adjudications,’ 17 speaks directly to ERISA suits, because the defendants have a statutory obligation, as well as a 18 fiduciary responsibility, to ‘treat the members of the class alike.’”); Urlaub, 2024 WL 2209538, at 19 *6–7 (certifying Rule 23(b)(1) class because remedy would reform the plans’ method of 20 calculating JSA benefits for all class members). 21 The Court finds certification proper under Rule 23(b)(1). Separate actions could result in 22 conflicting determinations as to how to calculate JSA benefits, and any adjudication by one class 23 member will necessarily impact all class members due to the nature of the MPP modifications 24 Berkeley seeks. The Court need not examine Berkeley’s arguments under the remaining Rule 25 23(b) categories. 26 27 IV. CONCLUSION Based on the foregoing, the Court GRANTS Berkeley’s motion for class certification. 2 IT IS SO ORDERED. 3 Dated: June 27, 2025 aM. 5 EDWARD J. DAVILA 6 United States District Judge 7 8 9 10 11 12

© 15 16

= 17

Z 18 19 20 21 22 23 24 25 26 27 28 || Case No.: 5:23-cv-00343-EJD ORDER GRANTING MOTION FOR CLASS CERTIFICATION