STATE OF NEW YORK et al v. UNITED STATES DEPARTMENT OF LABOR et al, No. 1:2018cv01747 - Document 79 (D.D.C. 2019)

Court Description: MEMORANDUM OPINION. Signed by Judge John D. Bates on 3/28/19. (lcjdb1)
Download PDF
STATE OF NEW YORK et al v. UNITED STATES DEPARTMENT OF LABOR et al Doc. 79 UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA STATE OF NEW YORK, et al. Plaintiffs, v. Civil Action No. 18-1747 (JDB) UNITED STATES DEPARTMENT OF LABOR, et al. Defendants. MEMORANDUM OPINION Eleven states and the District of Columbia have sued the Department of Labor (“DOL”), 1 alleging that its final rule interpreting the definition of “employer” in the Employee Retirement Income Security Act of 1974 (“ERISA”), 88 Stat. 829, 29 U.S.C. § 1001 et seq., is unlawful under the Administrative Procedure Act (“APA”), 5 U.S.C. § 706. DOL’s interpretation of the term “employer,” found at Definition of “Employer” Under Section 3(5) of ERISA—Association Health Plans, 83 Fed. Reg. 28,912 (June 21, 2018) (hereinafter “Final Rule”) (codified at 29 C.F.R. pt. 2510), A.R. at 1–53, 2 impacts the treatment of certain healthcare plans under both ERISA and the Patient Protection and Affordable Care Act (“ACA”), Pub. L. No. 111-148, 124 Stat. 119 (2010). 3 The States charge that DOL’s Final Rule stretches the definition of “employer” beyond what ERISA’s text and purpose will bear. For the reasons that follow, the Court agrees. 1 The named defendants in this case are DOL, R. Alexander Acosta in his official capacity as Secretary of Labor, and the United States of America. This opinion will collectively refer to defendants as “DOL.” 2 Citations to “A.R.” refer to the administrative record. See Joint Appendix [ECF Nos. 74-1, 74-2, 74-3, 744, 74-5, 74-6, 74-7, 74-8]. 3 For the purposes of this opinion, the term “ACA” refers not only to the original enactment of the Patient Protection and Affordable Care Act but also to subsequent amending legislation, including the Health Care and Education Reconciliation Act, Pub. L. No. 111-152, 124 Stat. 1029 (2010), and the Protecting Affordable Coverage for Employees Act, Pub. L. No. 114-60, 129 Stat. 543 (2015). 1 Dockets.Justia.com ERISA governs employee benefit plans arising from employment relationships. It provides that some employer associations acting “in the interest of” employer members are sufficiently employer-like to fall within the statute’s scope. Health plans offered by these associations may qualify as single ERISA plans, a designation that confers regulatory advantages under the ACA. For decades, DOL has interpreted these provisions narrowly so as to allow only so-called “bona fide associations” with close economic and representational ties to their employer members to qualify as “employers” under the statute. In 2018, DOL abruptly reversed course, issuing the Final Rule challenged in this case. The Final Rule allows virtually any association of disparate employers connected by geographic proximity to qualify as single ERISA plans. These associations no longer have to be viable apart from offering an association health plan (“AHP”) and may form solely for the purpose of creating an AHP. In addition, the Final Rule brings sole proprietors without any employees within ERISA’s scope by counting them as both “employers” and “employees.” Because the ACA defines terms key to its implementation—including “employer” and “employee”—according to the definition of these terms in ERISA, the Final Rule expands AHPs in a way that allows small businesses and some individuals to avoid the healthcare market requirements imposed by the ACA. The Final Rule is clearly an end-run around the ACA. Indeed, as the President directed, and the Secretary of Labor confirmed, the Final Rule was designed to expand access to AHPs in order to avoid the most stringent requirements of the ACA. Exec. Order 13,813, 82 Fed. Reg. 48,385 (Oct. 12, 2017), A.R. 6970; Final Rule, 83 Fed. Reg. at 28,912 (citing Executive Order); see also Alexander Acosta, A Health Fix for Mom and Pop Shops, Wall St. J., June 18, 2018. But equally important for the analysis that follows, the Final Rule does violence to ERISA. The Final Rule scraps ERISA’s careful statutory scheme and its focus on employee benefit plans arising from 2 employment relationships. It purports to extend ERISA to cover what are essentially commercial insurance transactions between unrelated parties. In short, the Final Rule exceeds the statutory authority delegated by Congress in ERISA. For the reasons that follow, the Final Rule’s provisions defining “employer” to include associations of disparate employers and expanding membership in these associations to include working owners without employees are unlawful and must be set aside. BACKGROUND Statutory schemes created by ERISA and the ACA shape the content and context of the Final Rule. First, therefore, it will help to describe relevant parts of ERISA and the ACA, explain the structure and function of the Final Rule, and set the stage for the provisions challenged in this case. I. ERISA AND THE ACA ERISA is the key statute at issue in this case. It regulates employee benefit plans, including welfare plans and pension plans, arising out of employment relationships. Congress enacted ERISA in 1974 “following almost a decade of study[]” of employment benefits and pension systems and after making “detailed findings which recited, in part, ‘that the continued well-being and security of millions of employees and their dependents are directly affected by [employee benefit] plans.’” Nachman Corp. v. Pension Benefit Guarantee Corp., 446 U.S. 359, 361–62 (1980) (quoting 29 U.S.C. §1001(a)). ERISA states that its purpose is to address the “growth in size, scope, and numbers of employee benefit plans” across the country and to protect “the interests of participants in employee benefit plans and their beneficiaries.” 29 U.S.C. § 1001(a)–(b). The ACA is a statutory scheme that regulates health insurance markets more broadly. The ACA, among other things, establishes standards that apply differently to individual, small-group, and large-group health insurance markets. Congress targeted the individual and small-group 3 healthcare markets for special heightened protections. Individual and small-group healthcare plans are required by the ACA to provide ten essential health benefits to insured individuals. 42 U.S.C. §§ 300gg-6, 18022(a). Large-group market participants face a choice: They may decline to provide these essential health benefits and instead pay a tax—the so-called “employer shared responsibility payment.” I.R.C. § 4980H, 26 U.S.C. § 4980H. Congress differentiated small employers from large employers—for the purpose of placing them in small- or large-group markets—by the number of employees these employers employed. See 42 U.S.C. § 300gg91(e)(2). The ACA absorbs key ERISA definitions into the ACA statutory scheme. Under ERISA, an employer is “any person acting directly as an employer, or indirectly in the interest of an employer, in relation to an employee benefit plan; and includes a group or association of employers acting for an employer in such capacity.” ERISA § 3(5), 29 U.S.C. § 1002(5). An employee under ERISA is simply “any individual employed by an employer.” ERISA § 3(6), 29 U.S.C. § 1002(6). The statutory definitions of “employer” and “employee” in ERISA have remained unchanged since ERISA’s enactment in 1974. Congress codified many of the ACA’s key provisions in the Public Health Service Act (“PHS Act”), 42 U.S.C. § 201 et seq. At the time that Congress passed the ACA, the PHS Act already defined the term “employer” as having “the meaning given such term under [ERISA § 3(5)], except that such term shall include only employers of two or more employees,” and defined “employee” as having the meaning given to the same term under ERISA § 3(6) without exception. 42 U.S.C. §§ 300gg-91(d)(5)–(6). 4 Congress preserved these definitions of “employer” and 4 Congress initially crafted these definitions of employer and employee in the PHS Act through the enactment of the Health Insurance Portability and Accountability Act (“HIPAA”), Pub. L. No. 104-191, 110 Stat. 1936, 1974 (1996). 4 “employee” when it passed the ACA, and thus ERISA’s definitions of those terms were incorporated into the ACA’s statutory scheme. II. ASSOCIATION HEALTH PLANS AND THE FINAL RULE AHPs are group health plans offered through an association of employers, such as an industry group. DOL has always permitted some AHPs meeting stringent criteria to qualify as a single ERISA employee benefit plan, as if the plan was sponsored by a single employer for its employees. Under DOL’s longstanding sub-regulatory guidance, only so-called “bona fide associations” could sponsor an AHP under ERISA. Bona fide associations had to display certain employer-like characteristics, because “the Department’s regulation of employee benefit plans [was] focused on employment-based arrangements, as contemplated by ERISA, rather than merely commercial insurance-type arrangements that lack the requisite connection to the employment relationship.” Final Rule, 83 Fed. Reg. at 28,914. The “overall structure” of ERISA “contemplates employment-based benefit arrangements,” and since these AHPs qualify as ERISA plans, they had to fit within an employment context, notwithstanding the fact that they were sponsored by employer associations rather than directly by single employers. Id. at 28,913. Of most relevance to the issues raised in this case, DOL’s sub-regulatory guidance analyzed bona fide associations based on three criteria: “(1) Whether the group or association [was] a bona fide organization with business/organizational purposes and functions unrelated to the provision of benefits; (2) whether the employers share[d] some commonality and genuine organizational relationship unrelated to the provision of benefits; and (3) whether the employers that participate[d] in a benefit program, either directly or indirectly, exercise[d] control over the program, both in form and substance.” Id. at 28,914. This opinion will refer to these three criteria as requirements for purpose, commonality of interest, and control, respectively. 5 Passage of the ACA raised the regulatory stakes. The majority of AHPs—which were not sponsored by associations qualifying under DOL’s bona fide association test—were “treated as the mechanism by which each individual employer obtains benefits and administrative services for its own separate plan.” Id. (citing CMS Ins. Standards Bulletin, “Application of Individual and Group Market Requirements under Title XXVII of the Public Health Service Act when Insurance Coverage Is Sold to, or through, Associations,” Sept. 1, 2011, at 1–3, A.R. at 2211–13). In other words, the fact that the insurance came from an AHP did not matter for purposes of determining which ACA market standard applied. A small employer purchasing coverage through an AHP had to meet the ACA’s requirements for small employers, individuals purchasing coverage through an AHP had to meet the ACA’s individual market requirements, and so on. Id. Most AHPs providing coverage to individuals and small groups, then, had to satisfy the ACA’s essential health benefits requirement. However, in the “rare instances” in which AHPs were sponsored by associations meeting DOL’s bona fide association criteria, the “association coverage [would be] considered a single group health plan,” and the number of total employees of all employer members would be counted to determine whether small or large group rules applied. CMS Ins. Standards Bulletin at 3, A.R. at 2213. If a bona fide association served as “employer” of “an average of at least 51 employees” over the past year, then the bona fide association would qualify as a “large employer,” and its AHP would be subject to “large group market” rules. 42 U.S.C. § 300gg-91(e)(2)–(3). 5 Hence, AHPs sponsored by these bona fide associations avoid the more comprehensive requirements for healthcare coverage in the individual and small group markets. 5 States have the option to expand the definition of small employers to cover employers of up to 100 employees, and at least four states have done so. Protecting Affordable Coverage for Employees Act § 2(b), 129 Stat. 543, 543 (amending PHS Act § 2791(e)(7) and 42 U.S.C. § 18024(b)(3)); Ctrs. for Medicare & Medicaid Servs., Market Rating Reforms: State Specific Variations, https://www.cms.gov/cciio/programs-and-initiatives/healthinsurance-market-reforms/state-rating.html (last updated June 2, 2017). For these states, the same rules apply to large employers, except that these employers must have 101 employees to qualify. 6 The Final Rule loosens the requirements for associations to qualify as ERISA-covered “bona fide associations,” thereby allowing the AHPs they sponsor to qualify as single ERISA plans and avoid the ACA’s individual and small-group market requirements. President Trump prompted DOL to undertake this change. In October 2017, President Trump issued an Executive Order titled “Promoting Healthcare Choice and Competition Across the United States,” which directed DOL to “[e]xpand[] access to AHPs” by “allow[ing] more small businesses to avoid many of the [ACA’s] costly requirements.” Exec. Order 13,813, 82 Fed. Reg. at 48,385; see also Final Rule, 83 Fed. Reg. at 28,912 (citing Executive Order). More specifically, the Executive Order suggested that DOL “expand[] the conditions that satisfy the commonality-of-interest requirements” in DOL’s then-existing guidance and “consider ways to promote AHP formation on the basis of common geography or industry.” Exec. Order 13,813, 82 Fed. Reg. at 48,386. DOL did as instructed. In June 2018, DOL promulgated the Final Rule, which significantly relaxed two of the three key criteria for qualifying as a bona fide association: the commonality of interest and purpose requirements. See Final Rule, 83 Fed. Reg. at 28,912. Associations can satisfy the new “commonality of interest” test if their members are either in the same trade or business or in the same geographic area (the same state or same metropolitan area, even if that area includes multiple states). 29 C.F.R. § 2510.3-5(c). Before the Final Rule, “geography, alone, was not sufficient to establish commonality.” Final Rule, 83 Fed. Reg. at 28,928 n.40. The Final Rule also allows an association to qualify as a “bona fide association” even if its primary purpose is “to offer and provide health coverage to its employer members and their employees” so long as it has “at least one substantial business purpose” unrelated to the provision of health care. 29 C.F.R. § 2510.3-5(b)(1). This represents a “departure from” DOL’s prior guidance, which required that 7 associations be viable organizations even without providing an AHP. Final Rule, 83 Fed. Reg. at 28,917. The Final Rule also adds an entirely new provision allowing working owners (i.e., sole proprietors) without any common-law employees to “qualify as both an employer and employee” for two key ERISA purposes. 29 C.F.R. § 2510.3-5(d). First, working owners may join bona fide associations of employers, including for purposes of “the requirement . . . that each employer member of the group or association participating in the group health plan must be a person acting directly as an employer of one or more employees who are participants covered under the plan.” Id. Second, working owners qualify as both employer and employee for purposes of satisfying the requirement that AHPs—as ERISA health benefit plans—may only offer health coverage “to employer members through the association” for qualifying employees and beneficiaries. Id. The effect of this provision is that working owners with no employees may join these newly expanded bona fide associations and receive health benefits through the association’s AHP, a single ERISA employee benefit plan. Under DOL’s prior guidance, “working owners without common law employees were not permitted to participate” in an AHP. Final Rule, 83 Fed. Reg. at 28,928 n.40. Although the Final Rule primarily interprets ERISA, the preamble to the Final Rule describes at least one other important implication of this new interpretation under the ACA. The preamble notes that the definition of the Final Rule “will apply solely for purposes of Title I of ERISA and for determining whether health insurance coverage of the AHP is regulated by the . . . PHS Act . . . provisions that apply to the individual, small group, or large group market, and not, for example, the purposes of taxation under the Internal Revenue Code.” Id. at 28,915. In other words, DOL indicates that the Final Rule permits AHPs to qualify as large employers under the ACA yet avoid the choice that other large employers must make between providing essential health 8 benefits or paying a shared responsibility payment. Under the Final Rule, AHPs avoid both ACA requirements for essential health benefits as described in the PHS Act and the employer shared responsibility payment under portions of the ACA codified in the Internal Revenue Code. 6 The Final Rule applied to fully-insured AHPs in September 2018 and to not-fully-insured existing AHPs in January 2019; it will apply to newly created AHPs beginning in April 2019. 29 C.F.R. § 2510.3-5(f). III. PROCEDURAL HISTORY Eleven States and the District of Columbia sued DOL over the Final Rule, raising claims under section 706 of the APA. Compl. [ECF No. 1]. The States allege that the Final Rule’s bona fide association test and working owner provision are “not in accordance with law” under ERISA and the ACA, that the Final Rule does not “carry out” Congress’s intent in enacting ERISA, and that the Final Rule is arbitrary and capricious. Id. ¶¶ 108–45. The States moved for summary judgment. Pls.’ Mem. of Law in Supp. of Mot. for Summ. J. (“Pls.’ Mot.”) [ECF No. 31-17]. DOL moved to dismiss the complaint on jurisdictional grounds for lack of standing and crossmoved in the alternative for summary judgment. Mem. of P. & A. in Supp. of Defs.’ Mot. to Dismiss, or, in the Alt., for Summ. J. (“Defs.’ Mot.”) [ECF No. 47-1]. The parties have fully briefed their positions on the motion to dismiss and on the motions for summary judgment, and the Court heard argument on the motions on January 24, 2019. The issues presented are now ripe for the Court’s consideration. 6 In a provision of the ACA codified in the Internal Revenue Code’s section on employer shared responsibility payments, Congress provided that “[a]ny term used in this section which is also used in the Patient Protection and Affordable Care Act shall have the same meaning as when used in such Act.” 26 U.S.C. § 4980H(c)(6). This section of the Internal Revenue Code includes repeated references to “employers,” including key provisions that require certain large employers to pay a tax. To the extent that the Final Rule purports to require that the agency’s interpretation of “employer” apply only to ERISA and portions of the ACA integrated in the PHS Act—but not to portions of the ACA integrated in the Internal Revenue Code—this outcome appears to be foreclosed by statutory language to the contrary. Compare id. with 42 U.S.C. § 300gg-91(d)(6). 9 LEGAL STANDARD A court must “hold unlawful and set aside agency action . . . found to be . . . arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law” or “in excess of statutory jurisdiction, authority, or limitations, or short of statutory right.” 5 U.S.C. § 706(2)(A), (C). “[W]hen a party seeks review of agency action under the APA, the district judge sits as an appellate tribunal.” Am. Bioscience, Inc. v. Thompson, 269 F.3d 1077, 1083 (D.C. Cir. 2001). “The ‘entire case' on review is a question of law,” id., and summary judgment is the proper mechanism for review, Ctr. For Food Safety v. Salazar, 898 F. Supp. 2d 130, 138 (D.D.C. 2012). ANALYSIS The Court first considers the issue of standing raised in DOL’s motion to dismiss. Finding that the States have standing to challenge the Final Rule, this opinion then considers the merits of the States’ challenges to the Final Rule. I. STANDING As a threshold matter, DOL urges dismissal of all claims on jurisdictional grounds. DOL argues that the States do not have standing to sue because they have not suffered a legally cognizable injury. DOL explains that “[t]he Final Rule applies to AHPs and not to states; it does not command any state to take, or refrain from taking, any action.” Defs.’ Mot. at 14. The States respond that they have several different forms of injury, each of which confers standing. Pls.’ Mem. of P. & A. in Opp’n to Defs.’ Mot. (“Pl.’s Opp’n”) [ECF No. 54-1] at 2–15. “Standing is a structural, constitutional restraint on the subject matter jurisdiction of the federal judiciary.” Air Alliance Houston v. EPA, 906 F.3d 1049, 1057 (D.C. Cir. 2018). “To establish Article III standing, an injury must be ‘concrete, particularized, and actual or imminent; fairly traceable to the challenged action; and redressable by a favorable ruling.’” Clapper v. Amnesty Int’l U.S.A., 568 U.S. 398, 409 (2013) (quoting Monsanto Co. v. Geertson Seed Farms, 10 561 U.S. 139, 149 (2010)). A court “presume[s] that federal courts lack jurisdiction ‘unless ‘the contrary appears affirmatively from the record.’” Renne v. Geary, 501 U.S. 312, 316 (1991) (quoting Bender v. Williamsport Area School Dist., 475 U.S. 534, 546 (1986)). “[T]he presence of one party with standing is sufficient to satisfy Article III’s case-or-controversy requirement.” Rumsfeld v. Forum for Acad. & Inst. Rights, Inc., 547 U.S. 47, 52 n.2 (2006). “Since they are not mere pleading requirements but rather an indispensable part of the plaintiff’s case, each element must be supported in the same way as any other matter on which the plaintiff bears the burden of proof, i.e., with the manner and degree of evidence required at the successive stages of the litigation.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 561 (1992). At the summary judgment stage, a plaintiff must set forth by affidavit or other evidence specific facts necessary to support standing. Swanson Grp. Mfg., LLC v. Jewell, 790 F.3d 235, 240 (D.C. Cir. 2015). First, although the parties agree that the Final Rule does not directly preempt state law, the States allege harm to their sovereign interests in making and enforcing a legal code. They express concern that the Final Rule sets the stage for possible future preemption—more specifically, that DOL might “enact future regulations to preempt State insurance laws as to AHPs . . . if States go ‘too far’ in regulating them.” Compl. ¶ 101 (citing Final Rule, 83 Fed. Reg. at 28,937). The Court agrees that States have sovereign interests in “the exercise of sovereign power over individuals and entities within the relevant jurisdiction,” which can be articulated as “the power to create and enforce a legal code.” Alfred L. Snapp & Son, Inc. v. Puerto Rico (“Snapp”), 458 U.S. 592, 601 (1982). The D.C. Circuit accordingly has recognized federal preemption of state law as a concrete injury that can give rise to standing. See Alaska v. U.S. Dep’t of Transp., 868 F.2d 441, 444 (D.C. Cir. 1989) (recognizing state standing to sue the Department of Transportation where its 11 regulations on airline pricing preempted state consumer protection laws). But the problem here is that the Final Rule does not preempt or otherwise constrain state law. The preamble to the Final Rule expresses DOL’s intention to leave AHPs to regulation by the states. For fully-insured AHPs, the Final Rule provides that State laws that regulate the maintenance of specified contribution and reserve levels (and that enforce those standards) may apply, and State insurance laws are generally saved from preemption when applied to health insurance issuers that sell policies to AHPs and when applied to insurance policies that AHPs purchase to provide benefits . . . [, and] it is the view of [DOL] that ERISA section 514(b)(6) clearly enables States to subject AHPs to licensing, registration, certification, financial reporting, examination, audit and any other requirement of State insurance law necessary to ensure compliance with the State insurance reserves, contributions and funding obligations. Final Rule, 83 Fed. Reg. at 28,936. For AHPs that are not fully insured, “section 514(b)(6)(A)(ii) of ERISA [provides that] any State law that regulates insurance may apply to the AHP to the extent that such State law is ‘not inconsistent’ with ERISA.” Id. The States point to no law that will be preempted, instead alleging that DOL might “enact future regulations to preempt State insurance laws as to AHPs . . . if States go ‘too far’ in regulating them,” and that “no statement in the Final Rule disclaim[s DOL’s] intent” to preempt other non-insurance state laws. Compl. ¶ 101 (citing Final Rule, 83 Fed. Reg. at 28,937). The Court finds that the possibility of future preemption is too speculative—and the concern about a possibility of preemption of unidentified non-insurance laws is too nebulous—to constitute an injury-in-fact. See Clapper, 568 U.S. at 401; Whitmore v. Arkansas, 495 U.S. 149, 158 (1990). Hence, the States do not have standing based on the possible future preemption of state law. The States next allege harm to their quasi-sovereign interests based on their “responsibility to protect the health, safety and welfare of their citizens.” Compl. ¶ 2. They predict that the Final Rule will harm state insurance markets “in States where state law does not duplicate ACA 12 requirements for individual and small group plans” because the Final Rule will allow healthier individuals to “leave the traditional market in states without sufficiently protective state laws.” Id. ¶ 104. 7 These healthier individuals allegedly will leave individual and small-group insurance markets for AHPs, which are predicted to be less expensive but also less comprehensive. This in turn will allegedly cause insurance premiums in the individual and small-group markets to rise, potentially pricing consumers out of coverage and/or pushing insurers out of these markets. Id. ¶¶ 104–05. The States express particular concern about employees earning less than four-hundred percent of the federal poverty level and working for small employers, because if these small employers were to offer insurance through an AHP—even if that insurance did not cover the employee’s full insurance needs—the employee could lose eligibility for premium tax credits under the ACA and thereby be priced out of a plan in the individual market offering all essential health benefits. Id. ¶ 106; Pls.’ Opp’n at 10 n.10. A state may have standing to sue under the doctrine of parens patriae when a state “express[es] a quasi-sovereign interest,” including “in the health and well-being—both physical and economic—of its residents in general.” Snapp, 458 U.S. at 607. And here the States have raised such interests. But because a state cannot claim superior sovereignty to the federal government, a state “does not have standing as parens patriae to bring an action against the Federal Government.” Id. at 609 n.16 (citing Massachusetts v. Mellon, 262 U.S. 447, 485–86 (1923)); see also Pennsylvania v. Kleppe, 533 F.2d 668, 677 (D.C. Cir. 1976) (explaining that “the state can not have a quasi-sovereign interest because the matter falls within the sovereignty of the Federal Government”). Here, the States’ general responsibility for their citizens’ health and welfare— including the threat of increased premiums, loss of insurance coverage, or loss of tax credits— 7 Plaintiff New York notes that its state laws are sufficiently protective “to prevent these harms.” Id. at ¶ 104 n.49. 13 cannot directly support State standing because the underlying harms would be suffered by the States’ citizens. Any theory of standing on this ground would derive from the State’s status as parens patriae, which is not available to States in a suit against the federal government. See Snapp, 458 U.S. at 609 n.16; Mellon, 262 U.S. at 485–486. The States accordingly do not have standing on this ground. The States also allege that the Final Rule will cause them three forms of proprietary or economic harm. First, they predict a rise in “uncompensated care costs” because an “individual who formerly had access to coverage through the [ACA] marketplace will now be underinsured or uninsured due to the Final Rule,” in which case the States “will . . . become responsible for providing care to individuals who cannot afford coverage or who are underinsured.” Compl. ¶ 106. Second, five States—New Jersey, Delaware, California, Washington, and Massachusetts—allege that under the Final Rule they will lose “tax revenue or administrative fees paid to state agencies for small group and individual plans obtained on a state insurance exchange.” Id. ¶ 102; Tr. of Jan. 24, 2019, Mot. Hr’g (“Tr.”) [ECF No. 77] at 31:20–23. Third, three States allege that the Final Rule will cause “a substantially increased regulatory burden on the States” as they “substantially ramp up enforcement against a new type of plan[] or face a wave of fraud and abuse similar to what occurred under [multiple employer welfare arrangements (“MEWAs”)] in past decades.” Compl. ¶ 103. The preamble to the Final Rule acknowledges that DOL “anticipates that the increased flexibility afforded AHPs under this rule will introduce increased opportunities for mismanagement or abuse, in turn increasing oversight demands on the Department and State regulators.” Final Rule, 83 Fed. Reg. at 28,953. At least two of these alleged economic injuries— 14 loss of tax revenues and increased regulatory burden—are sufficiently concrete, imminent, and direct to support standing for the States. 8 Lost tax revenues may serve as a cognizable injury-in-fact for standing purposes where an action caused “a direct injury in the form of specific tax revenues,” Wyoming v. Oklahoma, 502 U.S. 437, 448 (1992), and the state articulates a “fairly direct link between the state’s status as a collector and recipient of revenues and the legislative or administrative action being challenged,” Kleppe, 533 F.2d at 672. A state must provide evidence of a “specific loss of tax revenues,” not merely point to a source of revenue that might be affected by a federal policy or program. Wyoming v. U.S. Dep’t of Interior, 674 F.3d 1220, 1234–35 (10th Cir. 2012). Here, the States have met their burden to show a “fairly direct link” between the Final Rule’s intended expansion of self-insured AHPs and a decrease in specific tax revenues. See Kleppe, 533 F.2d at 672; Wyoming, 502 U.S. at 448. Unlike in the case of fully-insured AHPs, in which associations purchase insurance from insurance companies, an association sponsoring a self-funded AHP assumes the risk for all claims and pays benefits directly. The Final Rule’s expansion of self-funded AHPs will decrease state tax revenues because the affected States will not collect premium taxes when individuals select coverage through a self-insured AHP. The States have provided evidence that at least New Jersey, Delaware, and Washington stand to lose specific tax revenues under the Final Rule. 9 For example, Marlene Caride, Commissioner of the New Jersey Department of Banking and Insurance, has attested that health insurance companies in New Jersey pay one percent of group insurance premiums and two percent 8 For the third theory of economic harm—increased costs of uncompensated care—the record contains little concrete evidence of harm. Because the Court finds standing on the other two economic theories that the States present, it is unnecessary to decide whether uncompensated care costs might also confer standing. 9 The States represented to the Court at the January 24, 2019, motions hearing that California and Massachusetts also stood to lose tax revenue under the Final Rule, but the record does not contain sufficient information to reach a conclusion about those States. See Tr. 31:20–23. 15 of individual premiums to the State, and health maintenance organizations pay a corporate business tax and a two-percent assessment on premiums. Decl. of Marlene Caride in Supp. of Pls.’ Mot. (“Caride Decl.”) [ECF No. 31-2] ¶ 15. Any person currently insured in the traditional market in New Jersey who moves to a self-insured AHP or an out-of-state insured AHP will cause a corresponding decrease in tax revenues paid to New Jersey. Id. Declarations from representatives of Delaware and Washington similarly demonstrate that the Final Rule will decrease their tax revenues. 10 The preamble to the Final Rule also corroborates the States’ anticipated loss of premium taxes. In fact, avoidance of premium taxes is a feature of the Final Rule: self-insured AHPs can provide less costly insurance coverage in part because they “may avoid the potentially significant cost” of state requirements, including payment of “premium taxes.” Final Rule, 83 Fed. Reg. at 24,943. Hence, the Final Rule’s intended expansion of self-insured AHPs will cause a direct loss of premium tax revenues and related revenues from fees in at least three States. If the Final Rule is invalidated, these tax revenues are likely to be restored. Accordingly, the States have demonstrated an injury-in-fact that is caused by the Final Rule and is redressable, and hence they have standing to challenge the Final Rule as it relates to self-insured AHPs. See Rumsfeld, 547 U.S. at 52 n.2. 10 Delaware Insurance Commissioner Trinidad Navarro has attested that insurance companies in Delaware “pay an annual premium tax based on their net premium income,” and “[a]n increase in self-insured AHPs, made possible through the Final Rule, will result in a decrease in insurance companies’ premium income and, therefore, a decrease in the amount of premium tax collected by the State.” Decl. of Trinidad Navarro, Del. Ins. Comm’ner, in Supp. of Pls.’ Mot. (“Navarro Decl.”) [ECF No. 31-9] ¶ 10. Pam MacEwan, chief executive officer of the Washington Health Benefit Exchange (“WAHBE”), has attested that any decrease in the number of enrollees on the state exchange “will lower WAHBE’s revenues” because insurance carriers pay a two-percent premium tax plus “a flat per-member per-month assessment for enrollees on the Exchange.” Decl. of Pam MacEwan in Supp. of Pls.’ Mot. [ECF No. 31-7] ¶ 16. “[A]ny decline in enrollment will reduce the Exchange revenue” in Washington, which is also used “to support the Exchange’s costs for enrolling Medicaid applicants” and thereby will require expenditure of “additional state general fund dollars to replace those premium tax funds.” Id. ¶ 17. 16 The States’ direct regulatory costs also support standing. “Monetary expenditures to mitigate and recover from harms that could have been prevented absent [an agency action] are precisely the kind of ‘pocketbook’ injury” that constitute an injury to a proprietary interest for standing purposes. Air Alliance Houston, 906 F.3d at 1059–60. Such an injury must be direct and concrete, not speculative or attenuated. Clapper, 568 U.S. at 409. Here, the States have standing based on the Final Rule’s direct imposition of an increased regulatory burden on them. Several state regulators have attested that they have already incurred costs in hiring staff and designating staff time to regulation and enforcement of state and federal laws because of the Final Rule. For example, in Delaware, state regulators have already begun expending regulatory resources to answer “multiple inquiries” about “the formation and licensing requirements of AHPs” under the Final Rule. Navarro Decl. ¶ 7. 11 New York has also already incurred costs due to the Final Rule. 12 Invalidation of the Final Rule would not restore monies already expended, but it would halt the need for future state expenditures. For example, in the months ahead, Delaware anticipates using “additional State resources for the policing of, and enforcement actions taken against, AHPs” as well as a “35% increase in work for current . . . staff related to the licensing, oversight and enforcement actions for AHPs.” Navarro Decl. ¶ 11. Similarly, New Jersey will “expend additional resources and monies to enforce applicable state laws against . . . AHPs that are fraudulent and/or underfunded” and will “hire additional employees and devote additional funding 11 The Delaware Department of Insurance has also “already had to reassign and reprioritize assignments in order to prepare” for full implementation of the Final Rule and anticipates the need to “hire additional staff, including at least one full-time position and several part-time positions, to ensure that AHPs comply with applicable state and federal laws and regulations.” Id. ¶ 11. 12 New York has “devoted staff time to analyzing the AHP Rule” for associations inside and outside of the State and “advise[d] insurers and other licensees of New York requirements.” Decl. of Maria T. Vullo in Supp. of Pls.’ Mot. (“Vullo Decl.”) [ECF No. 31-13] ¶ 18. New York has also “already . . . taken action to prevent a potential influx of plans that purport to be authorized by the AHP Rule and ERISA but would violate New York” laws and regulations. Id. ¶ 17. 17 to ensure that AHPs are not improperly and impermissibly marketed” in the state. Caride Decl. ¶¶ 10, 12. Massachusetts, Pennsylvania, New York, Maryland, Oregon, and the District of Columbia have also described future costs that will be incurred due to the role their regulators will play in the ongoing implementation and enforcement of the Final Rule. 13 Finally, the States anticipate incurring expenses for consumer education initiatives. For example, Delaware intends to employ a “media campaign” at a cost of “tens of thousands of dollars” and “a minimum of 100 hours” of staff time, all of which the state “would not have incurred if not for the Final Rule.” Navarro Decl. ¶ 12. 14 It is notable that these regulatory expenditures are not merely incidental to the federal agency action. See Arpaio v. Obama, 27 F. Supp. 3d 185, 202–03 (D.D.C. 2014), aff’d, 797 F.3d 11, 20 (D.C. Cir. 2015), cert. denied 136 S. Ct. 900 (2016) (concluding that injuries related to a purported regulatory burden did not confer standing where the alleged injury was largely speculative and based on attenuated predictions of future illegal third-party conduct). In fact, the Final Rule expressly “depends on state insurance regulators for oversight and enforcement to, among other things, prevent fraud, abuse, incompetence and mismanagement, and avoid unpaid health claims.” Final Rule, 83 Fed. Reg. at 28,960. DOL expressly anticipated that the Final Rule would increase “oversight demands on . . . State regulators.” Id. at 28,953. And in setting the Final Rule’s staggered applicability dates, DOL considered the time needed for States to adopt 13 Massachusetts will “incur new costs . . . , such as additional staff time dedicated to enforcement [of state insurance standards].” Decl. of Audrey Morse Gasteier (“Gasteier Decl.”) [ECF No. 31-4] ¶ 8. Pennsylvania expects to “expend significant efforts to investigate consumer complaints.” Decl. of Christopher R. Monahan [ECF No. 318] ¶ 36. New York will expend “significant staff time, expenses for current and additional staff, diversion of staff from other priorities, budgetary planning, and travel and other expenses for investigative and visitorial activities” because of the Final Rule. Vullo Decl. ¶ 20. See also Decl. of Patricia F. O’Connor in Supp. of Pls.’ Mot. [ECF No. 31-10] ¶ 13 (describing costs to be incurred and resources to be mobilized for regulation and enforcement in Maryland); Decl. of Andrew Stolfi [ECF No. 11] ¶¶ 9–10 (same for Oregon); Decl. of Stephen C. Taylor in Supp. of Pls.’ Mot. [ECF No. 12] ¶ 16 (same for District of Columbia). 14 Massachusetts also anticipates incurring costs to distribute educational materials to consumers. Gasteier Decl. ¶ 8. 18 “new AHP-specific legislation and/or modification of existing regulations and enforcement programs.” Id.; see also Notice of Legislation 1 [ECF No. 72] (noting that the District of Columbia Council enacted emergency legislation in response to the Final Rule). State regulators are a central, essential piece of the Final Rule’s enforcement scheme. The economic harm due to this increased regulatory responsibility constitutes an injury-in-fact that is caused by the Final Rule and redressable by its invalidation. See Air Alliance Houston, 906 F.3d at 1059–60. Hence, this increased regulatory burden confers standing on the States to challenge the Final Rule. II. CHALLENGES TO THE FINAL RULE The States challenge multiple dimensions of the Final Rule under section 706 of the APA, arguing that the Final Rule’s bona fide association and working owner provisions conflict with the text and purpose of both the ACA and ERISA and exceed DOL’s statutory authority. The States also maintain that the Final Rule is arbitrary and capricious under the APA. DOL disagrees with the contention that this case has anything to do with the ACA and urges the Court to consider the Final Rule narrowly as an interpretation of ERISA. Defs.’ Mot. at 1–2; Tr. at 65:16–18 (“[T]his is a case about ERISA[.]”). Because Congress granted DOL statutory authority to interpret ERISA, DOL argues that the Court should defer to DOL’s reasonable interpretation under the narrow judicial review permitted under the APA. Defs.’ Mot. at 3, 25–26, 33, 44, 48. The Court agrees that this case is fundamentally one about ERISA and that DOL’s interpretation of ERISA should normally receive deference to the extent that interpretation is reasonable. However, the Court concludes that DOL has failed to reasonably interpret the statute. The Final Rule’s bona fide association standard fails to establish meaningful limits on the types of associations that may qualify to sponsor an ERISA plan, thereby violating Congress’s intent that only an employer association acting “in the interest of” its members falls within ERISA’s scope. The Final Rule’s working owner provision similarly exceeds ERISA’s scope because it seeks to 19 extend ERISA’s coverage to plans arising outside of any employment relationship. For the reasons that follow, then, the Court concludes that these provisions of the Final Rule exceed DOL’s authority and must be set aside. A. The Chevron Framework Applies to DOL’s Interpretation of ERISA As a threshold matter, the Court agrees that the deferential standard announced in Chevron v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), applies to DOL’s interpretation of “employer” in ERISA. Courts generally apply the two-step Chevron framework when reviewing an agency’s construction of a statute that it administers. See id. at 842. Under that framework, courts first determine whether the statute is ambiguous, because if “Congress has directly spoken to the precise question at issue [and] . . . the intent of Congress is clear, that is the end of the matter.” Id. If the statute is ambiguous, however, then the court considers whether the agency’s interpretation is reasonable—that is, “whether the agency’s answer is based on a permissible construction of the statute.” Id. at 843. Such interpretive regulations are “binding in the courts unless procedurally defective, arbitrary or capricious in substance, or manifestly contrary to the statute”—even if “the agency’s chosen regulation seems unwise.” United States v. Mead Corp., 533 U.S. 218, 227, 229 (2001). Here, Congress tasked DOL with administering ERISA. 29 U.S.C. § 1135 (delegating authority to the Secretary of Labor to “prescribe such regulations as he finds necessary or appropriate to carry out the provisions of [ERISA]”); see also Black & Decker Disability Plan v. Nord, 538 U.S. 822, 831 (2003) (deferring to a DOL interpretation of an ERISA provision). ERISA’s definition of “employer” is ambiguous. DOL noted the statute’s ambiguity in its preamble to the Final Rule, see Final Rule, 83 Fed. Reg. at 28,914, and this conclusion is corroborated by other courts that have considered the issue, see Matincheck v. John Alden Life Ins. Co., 93 F.3d 96, 100 (3d Cir. 1996) (finding ERISA statutory definitions of employer and 20 employee ambiguous); Dodd v. John Hancock Mut. Life Ins. Co., 688 F. Supp. 564, 569 (E.D. Cal. 1988) (finding ambiguity in ERISA terms employer, employee, and person). Because Congress delegated authority to DOL to interpret this ambiguous statute, the Chevron framework applies, and the question before the Court is therefore whether DOL’s interpretation is reasonable. B. DOL’s Regulatory Interpretation of ERISA Is Not Reasonable DOL’s Final Rule does not reasonably interpret ERISA. ERISA regulates benefits in the context of employment relationships. DOL acknowledges as much in the preamble to the Final Rule, explaining that the agency’s authority under the statute is constrained because “ERISA section 3(5) does not broadly extend to arrangements established to provide benefits outside the employment context and without regard to the members’ status as employers.” Final Rule, 83 Fed. Reg. at 28,916. Because the Final Rule stretches the definitions of “employer” beyond what the statute can bear, the Final Rule is unlawful under the APA. i. ERISA Is Limited to Employee Benefit Plans Arising from Employment Relationships The text and purpose of ERISA limit its scope to benefit plans arising from employment relationships. This is not a situation in which a court must divine congressional intent from legislative history and presidential signing statements; rather, Congress inscribed its findings and declaration of policy on the face of the statute. Congress enacted ERISA “following almost a decade of study[]” of employment benefits and pension systems. Nachman Corp., 446 U.S. at 361. By its text, ERISA addresses the “growth in size, scope, and numbers of employee benefit plans” across the country and aims to protect “the interests of participants in employee benefit plans and their beneficiaries.” 29 U.S.C. § 1001(a)–(b). Congress expressed its interest in regulating “employee benefit plans” because these plans, among other things, implicate interstate commerce and directly affect “the continued well-being and security of millions of employees and their 21 dependents” as well as “the stability of employment and industrial relations.” Id. More specifically, ERISA remedies “lack of employee information and adequate safeguards concerning . . . operation” of employee benefit plans through disclosure requirements, as well as plans’ failure to pay promised benefits by setting minimum standards and “providing . . . appropriate remedies, sanctions, and ready access to the Federal courts.” Id. §§ 1001(a)–(b). In short, ERISA concerns benefit plans arising from employment relationships and accordingly regulates only those plans. Notably absent from ERISA’s statement of policy is any expression of an intent to expand citizen access to healthcare benefits outside of an employment relationship or to directly regulate commercial healthcare insurance providers. Congress does regulate in these areas, but it does so through other statutory schemes—including the ACA. ERISA’s concern with benefits arising from employment relationships is further evinced by its “lexicographic topography”—that is, its definitions section. See MDPhysicians & Assocs., Inc. v. State Bd. of Ins., 957 F.2d 178, 182 (5th Cir. 1992). ERISA’s fundamental unit of analysis is the “employee benefit plan,” which may take the form of an “employee welfare benefit plan,” an “employee pension benefit plan,” or both. ERISA §§ 3(1)–(3); 29 U.S.C. §§ 1002(1)–(3). An employee welfare benefit plan is a “plan, fund, or program . . . established or maintained by an employer or by an employee organization, or by both, to the extent that such plan, fund, or program was established or is maintained for the purpose of providing for its participants or their beneficiaries.” Id. §§ 3(1); 29 U.S.C. § 1002(1). A benefit plan “participant” includes “any employee or former employee of an employer, or any member or former member of an employee organization.” Id. § 3(7); 29 U.S.C. § 1002(7). And an “employee” is “any individual employed by an employer.” Id. § 3(6); 29 U.S.C. § 1002(6). These definitions have been further refined through DOL regulations. See 29 C.F.R. §§ 2510.3–1, 3–2, 3–3. But the broad contours of the 22 statutory text make clear that Congress intended that only benefit plans arising from employment relationships would fall within ERISA’s scope. If Congress had intended ERISA to regulate ordinary commercial insurance relationships existing outside of the employment context, one would not expect it to have framed ERISA’s scope in terms of employee benefit plans, created or maintained by employers or employee organizations for the benefit of current and former employees and their beneficiaries. See ERISA § 3; 29 U.S.C. § 1002. ii. ERISA Extends the Definition of Employer Only to Associations Acting in the Interest of Employers ERISA authorizes some employer associations to qualify as “employers” for the purpose of sponsoring an employee benefit plan, so long as the “group or association of employers” acts “in the interest of an employer.” ERISA § 3(5), 29 U.S.C. § 1002(5) (emphasis added). This congressional mandate forms the core of the dispute in this case. The Final Rule is lawful only to the extent that it carries out Congress’s intent to limit associations qualifying as employers to those acting “in the interest of” an employer. This provision leaves open to agency interpretation the precise contours of who might fit within this definition. But the statutory text is not infinitely elastic. The phrase “in the interest of an employer” distinguishes employer associations that stand in the shoes of an “employer” for the purpose of sponsoring an ERISA plan from every other employer association. The D.C. Circuit has warned that Congress did not intend for “every . . . agent who discharges some responsibility in regard to a corporation’s employee benefit plan” to be “swept within the definition” of “employer” under ERISA. Int’l Bhd. of Painters & Allied Trades Union v. George A. Kracher, Inc., 856 F.2d 1546, 1548 (D.C. Cir. 1988). The legislative history of ERISA makes clear that entrepreneurial ventures selling insurance for a profit to unrelated groups are unequivocally outside of ERISA’s scope. Wayne Chem., Inc. v. Columbus 23 Agency Serv. Corp., 567 F.2d 692, 699 (7th Cir. 1977); MDPhysicians, 957 F.2d at 183–84; Gruber v. Hubbard Bert Karle Weber, Inc., 159 F.3d 780, 786 (3d Cir. 1998); see also Activity Report of the House Committee on Education and Labor, H.R. Rep. No. 1785, 94th Cong., 2d Sess. at 48 (1977) (noting that the committee does “not believe that the statute and legislative history [of ERISA would] support the inclusion of what amount[] to commercial products within the umbrella of the definition [of an ERISA plan under section 3(3), 29 U.S.C. § 1002(3)]”). Relatedly, because ERISA plans arise from the special relationship between employers and employees, a plan is not an ERISA plan unless the entity providing benefits and the individuals receiving the benefits demonstrate the “economic or representation[al]” ties or protective “nexus” that characterizes an employment relationship. Wis. Educ. Ass’n Ins. Trust v. Iowa State Bd. of Pub. Instruction (“WEAIT”), 804 F.2d 1059, 1063 (8th Cir. 1986); MDPhysicians, 957 F.2d at 186. iii. The Final Rule’s Expansive Test for Bona Fide Associations Is Not Reasonable The Final Rule’s bona fide associations provision is not reasonable because it unlawfully expands ERISA’s scope. The Final Rule adopted the same three overall criteria 15 that DOL previously employed for determining which associations are “bona fide” and thus act in the interest of an employer for the purpose of establishing an AHP: purpose, commonality of interest, and control. However, the Final Rule departs significantly from DOL’s prior sub-regulatory guidance in the way it measures these criteria. This is the heart of the States’ challenge to the Final Rule. 15 The Final Rule also imposes additional requirements on bona fide associations, including requirements for an organizational structure and nondiscrimination. 29 C.F.R. §§ 2510.3-5(b)(3), (7). The Final Rule clarified eligibility to participate in an AHP and prohibited commercial health insurers from running AHPs. Id. §§ 2510.35(b)(6), (8). Plaintiffs do not challenge these additional requirements imposed on AHPs, and the Court will not address these requirements further because they are collateral to the three main requirements under the Final Rule. 24 These three criteria form the primary basis for DOL’s examination of which associations qualify as ERISA “employers,” and they must be viewed holistically to determine whether in sum they limit bona fide associations to those contemplated by Congress in ERISA. The analysis that follows describes each criterion individually, then considers whether together they place reasonable constraints on the types of associations that act “in the interest of” employers under ERISA. a. Purpose Test The purpose test allows an association to sponsor an AHP so long as the association has “at least one substantial business purpose” unrelated to the provision of health care, even if its primary purpose is “to offer and provide health coverage to its employer members and their employees.” 29 C.F.R. § 2510.3-5(b)(1). The rule does not define “substantial business purpose,” but it creates a safe harbor “if the group or association would be a viable entity in the absence of sponsoring an employee benefit plan.” Id. The preamble to the Final Rule provides examples of “substantial business purpose[s],” including “convening conferences or offering classes or educational materials on business issues of interest to the association members,” acting “as a standard-setting organization that establishes business standards or practices,” or engaging “in public relations activities such as advertising, education, and publishing on business issues of interest to association members unrelated to sponsorship of an AHP.” Final Rule, 83 Fed. Reg. at 28,918. The problem with the purpose test as DOL now interprets it is that it fails to set meaningful limits on the character and activities of an association. The Final Rule permits associations to form for the primary purpose of establishing an AHP, and hence the only limitation imposed by the purpose test is its undefined “substantial business purpose” requirement. The possible scope of qualifying substantial business purposes ranges from the resource intensive—e.g., setting business 25 standards and practices—to the de minimis—e.g., publishing a newsletter on business issues. Of course, the latter is something that most associations already do and thus is not a defining characteristic of a subset of organizations that would fall within ERISA’s scope. The Final Rule’s “safe harbor” provision reveals how flimsy the purpose test really is. The safe harbor provision specifies that an association that “would be a viable entity in the absence of sponsoring an employee benefit plan” will satisfy the purpose test. 29 C.F.R. § 2510.3-5(b)(1). This phrasing suggests, however, that at least some associations would still meet the purpose test even if they were outside of this safe harbor—that is, not viable but for sponsoring an AHP. Sponsoring an AHP, then, may serve not only as the “primary purpose” of an association under the first part of the purpose test—sponsoring an AHP may serve as its only real purpose. The “substantial business purpose” test, then, is only an ex post facto, perfunctory requirement— merely a box to check—that virtually any association may fulfill on the side and thereby qualify to sponsor an AHP under the Final Rule. This business purpose does not, in fact, need to be “substantial” in the ordinary sense of that term, because it need not make the association viable in the absence of the association’s AHP. This requirement therefore is more aptly called the “other business task” test. It sets such a low bar that virtually no association could fail to meet it. Indeed, it is likely that nearly every employer association that exists today already satisfies the Final Rule’s substantial business purpose requirement. In short, the Final Rule’s purpose test provides no meaningful limit on the associations that would qualify as “bona fide” ERISA “employers.” It does no work towards narrowing extant associations to only those that act “in the interest of” employers. Although it describes its requirement in terms of a “substantial business purpose,” this requirement gives way under the slightest pressure. 26 b. Commonality of Interest Test To form an association sponsoring an AHP under the Final Rule, employers also must display a “commonality of interest.” 29 C.F.R. § 2510.3-5(b)(5). This requirement may be met through two alternative routes: employers must either share a common “trade, industry, line of business, or profession,” or else each employer must have “a principal place of business in the same region that does not exceed the boundaries of a single State or a metropolitan area (even if the metropolitan area includes more than one State).” 29 C.F.R. § 2510.3-5(c)(1). The States object to the latter, which deems employers to be united in interest solely because of common geographical location. The commonality of interest test is arguably the most important of the three criteria because it most directly relates to the core concern of the statute: employers’ interests. Before an association can act “in the interest of” an employer member, that interest must be defined. But common geography does not necessarily correlate with any common interest. The Final Rule thereby permits unrelated employers in multiple, unrelated industries to associate and be deemed to act “in the interest of” the employer members, notwithstanding the fact that the interests of these employer members may be very different or even conflicting. DOL does not provide a rationale that would connect geography and common employer interest. The preamble to the Final Rule explains the desired effect of the geography requirement, which is that this requirement will “provide employer groups and associations with important flexibility and allow more employers to join together to secure lower cost healthcare coverage for themselves and their employees through AHPs.” Final Rule, 83 Fed. Reg. at 28,924. In other words, the geography requirement is desirable because it lowers barriers to AHP formation. In this way, the Final Rule responds to the President’s direction to “consider ways to promote AHP formation on the basis of common geography.” Id. at 28,912; see also Exec. Order 13,813, 82 27 Fed. Reg. at 48,386. However, the preamble to the Final Rule fails to explain how geography furthers the statutory requirement that associations act in the interest of employers or why employers with a place of business in a state would be expected to share common interests. See id. at 28,923–26; see also Definition of “Employer” Under Section 3(5) of ERISA—Association Health Plans, Proposed Rule, 83 Fed. Reg. 614 (Jan. 5, 2018) [A.R. at 6947–69]. There is nothing intrinsic in common geography that would generate the types of economic or reputational ties that courts have deemed essential for a plan to be covered by ERISA. Hence, courts interpreting ERISA’s requirements have declined to extend ERISA’s protections and privileges to organizations based simply on geographic proximity. For example, in both MDPhysicians and WEAIT, the plans at issue were geographically focused (in the Texas panhandle and in Iowa, respectively), yet both courts concluded that the plans fell outside of ERISA’s scope. See MDPhysicians, 957 F.2d at 180; WEAIT, 804 F.2d at 1060. ERISA imposes a common interest requirement, not merely a something-in-common requirement. Accordingly, courts have “consistently rejected the contention that heterogenous businesses that share nothing more than a common size and a high regard for the ‘entrepreneurial spirit’” satisfy ERISA’s commonality of interest requirement. Gruber, 159 F.3d at 788. Geography, similarly, is not a logical proxy for common interest, and substituting shared geography for the statutory requirement of common interest improperly expands ERISA’s scope. 16 DOL’s decision to select geography as a proxy for common interest is more perplexing when one considers other designators of commonality that DOL rejected. These rejected characteristics included “ownership characteristics (e.g., an association of owners who are women, 16 DOL could certainly craft a bona fide association rule taking geography into account. The problem with the rule is not that any consideration of geography is categorically impermissible, but rather that geography alone does not logically predict a commonality of interest. 28 minorities, or veterans), business models or structures (such as businesses owned by ESOPs, franchises, or not-for-profits), size of business (e.g., small businesses), [and] shared religious and moral convictions.” Final Rule, 83 Fed. Reg. at 28,926. DOL noted some commenters’ belief that “employers within these relationships often share unique bonds, interests, needs, and regulatory schemes, and may have significantly more commonality of interest than those in the same industry or region due to these shared traits.” Id. But DOL rejected these characteristics because “a test that would treat all nationwide franchises, all nationwide small businesses, or all nationwide minority-owned businesses[] as having a common employment-based nexus—no matter the differences in their products, services, regions, or lines of work—would not be sufficient to establish commonality of interest for a national group or association and AHP because it would be impossible to define or limit . . . and, in the Department’s view, would eviscerate the genuine commonality of interest required under ERISA.” Id. The same concerns that animated DOL’s decision to reject these other designators of commonality apply with equal force to DOL’s geography test. Take, for example, an association composed of employers with principal places of business in California. The association could form for the primary purpose of creating an AHP, so long as participating employer members elected directors and the association published a quarterly newsletter on business issues of interest to California-based businesses. The employer members (of varying sizes) might include a restaurateur in Oakland, a physicians practice group in the Hollywood Hills, an almond farmer in the Central Valley, an importer in Long Beach, a technology company headquartered in San Diego but doing business primarily in New York, and a Fresno fast-food franchise. These employers share no “unique bonds, interests, needs, [or] regulatory schemes” and exhibit a wide range of “differences in their products, services, regions, [and] lines of work.” See id. They would be no 29 more or less united in interest if their principal places of business happened to be in various states. The Court concludes, therefore, that the geography standard under the Final Rule fails to account in any way for employers’ commonality of interest. This standard effectively eviscerates the genuine commonality of interest required under ERISA, thereby expanding the scope of the statute beyond what ERISA intended. Because the geography test does not, in fact, ensure that associations qualifying to sponsor AHPs under the Final Rule share a “commonality of interest,” it creates no meaningful limit on these associations. In other words, the geography test does no work to focus the Final Rule on the types of associations that Congress intended ERISA to cover. c. Control Test The third criterion under the Final Rule that serves to constrain which associations may sponsor AHPs is the control test, 17 which requires simply that “[t]he functions and activities of the group or association are controlled by its employer members, and the group’s or association’s employer members that participate in the group health plan control the plan.” 29 C.F.R. § 2510.35(b)(4). This employer member “[c]ontrol must be present both in form and in substance.” Id. “Whether the requisite control exists is determined under a facts and circumstances test.” Final Rule, 83 Fed. Reg. at 28,919. DOL will deem employer members to have control where, for example, they can nominate, elect, and remove directors and approve or veto material amendments 17 DOL contends that the Final Rule’s less stringent standards for commonality of interest and purpose are balanced by “new nondiscrimination requirements.” Defs.’ Mot. at 2. This nondiscrimination requirement “prevent[s] associations from charging employer members different premium rates based on the health status of their employees” and “reflects the Department’s reasonable efforts to supply the criteria necessary to elucidate the definition of Section 3(5) in a manner that is consistent with the statute and ERISA’s objective of encouraging the creation of employment benefit plans.” Id. at 28. This nondiscrimination provision still “permit[s] employment-based distinctions to be used within an AHP, provided that such distinctions are not directed at individual participants or beneficiaries based on any health factor.” Final Rule, 83 Fed. Reg. at 28,923 n.31. But this nondiscrimination provision does not constrain which associations qualify under the Final Rule—it only limits how qualifying associations may structure their premiums in an AHP. Because this requirement does not serve to limit which associations qualify as ERISA “employers,” it does not weigh in the Court’s analysis in this case. 30 to the AHP. Id. at 28,920. Employer members are not required to “manage the day-to-day affairs of the group or association or the plan.” Id. The control test “largely duplicate[s] the conditions in the Department’s pre-rule guidance.” Id. at 28,919. DOL has observed that “the control test is necessary to satisfy the statutory requirement in ERISA section 3(5) that the group or association must act ‘in the interest of’ the employer members in relation to the employee benefit plan in order to qualify as an employer.” Id. at 28,919, 28,955. The control test limits the types of associations that qualify as employers under the Final Rule by ensuring that employer members direct the actions and decisions of the association with respect to the AHP. The control test alone, however, does not mean that employer members are united in interest, but rather only that the employer members can steer the association’s decisionmaking when working in concert. Courts have concluded—and this Court agrees—that the control test complements and supplements the commonality of interest test but cannot replace it. For example, the Third Circuit has held that “to qualify as an ‘employer’ for ERISA purposes, an employer group or association must satisfy both the commonality of interest and control requirements.” 18 Gruber, 159 F.3d at 787 (emphasis added); see also MDPhysicians, 957 F.2d at 186 (determining that the subject plan was not an ERISA plan because it failed to meet both the commonality of interest and control requirements). In other words, the control test is only meaningful if employer members’ interests are already aligned. If employer members have opposed interests, the control test does nothing to 18 The Third Circuit in Gruber also distinguished the legal bases of the two tests. It noted that the commonality of interest requirement derived from the text of ERISA itself, as distilled through DOL advisory opinions, while the control requirement derived solely from DOL guidance, which the court found worthy of deference as “a reasonable means of ensuring that the administrators of multi-employer welfare benefit plans in fact act ‘in the interest of’ their employer members.” See 159 F.3d at 787–88. 31 resolve those differences. For example, in the case of an association whose employer members display widely disparate interests, the existence of indicia of control such as election of officers and voting on plan amendments would make it more likely that the association might further the interests of some—perhaps those that are most powerful or most numerous—but not all employers in the organization. * * * The Final Rule is a lawful interpretation of ERISA only so long as it limits the “bona fide associations” that may qualify as employers to those acting “in the interest of” their employer members, as mandated by ERISA’s text. See ERISA § 3(5); 29 U.S.C. § 1002(5). Here, the three criteria that limit association membership fall short of ERISA’s statutory requirements. The Final Rule’s purpose test sets no meaningful limitation on associations because its requirements are largely cursory, describing activities in which an association must participate but not meaningfully addressing whether the association’s activities align with or further employer members’ interests. So, too, the geography pathway for meeting the commonality of interest requirement does little to ensure that the association will act in its employer members’ interests because shared geography does not reflect any common interest. The purpose and commonality of interest criteria—even when combined—fail to establish an interest-based bond between associations and employer members, which in turn undermines the effectiveness of the control test: disparate employers pursuing their own varied interests through an association’s AHP—even if the employers exert a modicum of control over the organization—simply do not come within ERISA’s statutory scheme. The three criteria set by the Final Rule for qualification as a “bona fide association,” then, even acting in combination, ultimately fail to respect the statutory limitations set by Congress because the Final Rule does not functionally constrain bona fide associations to those acting “in the interest of” employers. The Final Rule would permit a group of employers with no common 32 characteristic other than presence in the same state to qualify as a single employer under ERISA so long as that group had an election-based officer structure and some modest business-related side project. Yet as summarized in WEAIT, ERISA is premised on the idea that employers and employees are connected by an employment nexus: “[a]n employee depends on his employer,” and vice versa. WEAIT, 804 F.2d at 1063. This nexus is reproduced at the level of associations and their employer members. Yet there is no reason to believe that the employee of an employer joining such a loosely defined association under the Final Rule would have real ties to that association or that any associational bonds would provide inherent limits on the activities of the association with respect to its employer members or their employees. The Final Rule allows groups that closely resemble entrepreneurial, profit-driven commercial insurance providers to qualify for ERISA’s protections—an outcome that longstanding interpretations of ERISA have explicitly forbidden. See Gruber, 159 F.3d at 786; MDPhysicians, 957 F.2d at 183–84; Wayne Chem., 567 F.2d at 699; Activity Report of the House Committee on Education and Labor, H.R. Rep. No. 1785, 94th Cong., 2d Sess. at 48–49 (1977). Allowing such associations to sponsor AHPs under ERISA “twist[s] the language of the statute and defeat[s] the purposes of Congress,” and therefore conflicts with DOL’s authority under ERISA. See MDPhysicians, 957 F.2d at 185. For these reasons, the bona fide association provision of the Final Rule unlawfully expands ERISA’s scope and conflicts with ERISA’s text. Hence, it is not a reasonable interpretation of ERISA and must be set aside. iv. The Final Rule’s Expansion of “Employer” to Include Working Owners without Employees Is Not Reasonable The Final Rule’s expansion of the definition of employer to include working owners without employees is also contrary to the text of ERISA. The Final Rule allows “[a] working owner of a trade or business without common law employees [to] qualify as both an employer and 33 as an employee of the trade or business” for several key purposes. 29 C.F.R. § 2510.3-5(e)(1). These working owners without employees qualify as both “employer members” of bona fide associations and as their own “employees,” thereby qualifying them to join bona fide associations and participate in the association’s AHP. Id. §§ 2510.3-5(b), (e). The Final Rule specifies that working owners with no employees satisfy the condition for bona fide associations that “[e]ach employer member [participating in an AHP] is a person acting directly as an employer of at least one employee who is a participant covered under the plan.” Id. § 2510.3-5(e). Working owners without employees, per the Final Rule, may also receive benefits from the AHPs of bona fide associations notwithstanding the provision that these AHPs may provide health coverage for only current and former “employee[s] of a current employer member” or that employee’s beneficiaries. Id. § 2510.3-5(b)(6)(i). The States note that for ERISA’s more-than-forty-year history, working owners without employees have been ineligible to join associations under ERISA. Pls.’ Mot. at 2. This is more than a consequence of agency policy, they argue—it is because ERISA cannot accommodate such an interpretation. Id. at 28–30. DOL counters that the definition of “employer” is ambiguous and that nothing in the statute or precedent conclusively forecloses the agency’s interpretation. Defs.’ Mot. at 44–47. DOL acknowledges that ERISA would not apply “to a plan that the working owner is providing to himself” outside of membership in an association, because DOL acknowledges that a working owner’s plan covering only himself “falls outside the scope of ERISA.” Tr. at 73:17– 74:10. However, DOL contends that membership in an association transforms a working owner with no employees into both an employer and employee under ERISA, thereby bringing that working owner without employees within the scope of the statute and the Final Rule. Id. at 74:11– 76:22. More specifically, DOL argues that two sole proprietors without employees may band 34 together in an association, and this two-member association may then sponsor an ERISA plan for the two working owners’ benefit. Id. at 76:1–22. A working owner without employees is plainly beyond ERISA’s scope when he establishes a benefit plan for himself. The Court concludes that a working owner’s membership in an association does not bring him within ERISA. And the contention that two working owners without employees, neither of whom is within ERISA’s scope alone, could associate with one another and thereby come within the statute’s reach is absurd. Congress did not intend for working owners without employees to be included within ERISA—either as individuals or when joined in an employer association. ERISA clearly contemplates regulation of benefits arising from employment relationships. The Final Rule’s attempt to bring sole proprietors with no employees into ERISA’s fold stretches the statute too far. As a practical matter, one does not have an employment relationship with oneself. Notwithstanding the broad sweep of ERISA’s definition of employer, a sole proprietor neither acts “directly as an employer” or “indirectly in the interest of an employer”—i.e., of himself. See ERISA § 3(5); 29 U.S.C. § 1002(5). There is no indication that Congress crafted the statute with the intent of sweeping working owners without employees—who employ no one—within ERISA’s scope through the statutory definition of “employer.” The statutory conflict is even clearer when one considers the statute’s definition of “employee.” Unlike the more expansive definition of “employer,” the definition of “employee” under ERISA is limited: simply an “individual employed by an employer.” ERISA § 3(6); 29 U.S.C. § 1002(6). The text of the statute clearly anticipates a relationship between two parties. Congress constructed the definition of “employee” with a transitive verb (to employ), which necessarily describes a relationship between a subject and object—employer and employee. If 35 such a relationship were not intended, then Congress could have defined “employee” as simply a “working individual.” But it did not. Further, the Supreme Court has read ERISA’s definition of “employee” to “incorporate traditional agency law criteria for identifying master-servant relationships,” Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318, 319 (1992), and “[t]he traditional agency criteria can be applied logically only in situations involving relationships between two different persons, i.e., those who employ persons and those who are so employed,” In re Watson, 161 F.3d 593, 597 (9th Cir. 1998). It is unreasonable to say that Congress drafted the statute with the intent to regulate a person’s internally conflicted relationship with himself. That would threaten to turn ERISA into an exercise in psychoanalysis. Supreme Court precedent also supports finding that a working owner without employees falls outside ERISA. In Raymond B. Yates, M.D., P.C. Profit Sharing Plan v. Hendon, 541 U.S. 1 (2004), the Supreme Court held that “[u]nder ERISA, a working owner may have dual status, i.e., he can be an employee entitled to participate in a plan and, at the same time, the employer (or owner or member of the employer) who established the plan.” 541 U.S. at 16. DOL argues that this sentence in Yates at least opens the door for any sole proprietor—with or without employees—to qualify as dual-status employee and employer under ERISA. Defs.’ Mot. at 45– 47. But DOL’s reading emphasizes a single sentence and ignores the rest of the opinion. The Supreme Court emphasized that Yates’s plan had always included “at least one person other than Yates or his wife.” Id. at 8. The opinion clarified that “[i]f the plan covers one or more employees other than the business owner and his or her spouse, the working owner may participate on equal terms with other plan participants.” Id. at 6 (emphasis added). The presence of employees, then, is a condition precedent to the creation of an ERISA plan because without employees there is no 36 employment relationship to give rise to an employee benefit plan. But once such a plan exists, a working owner may participate alongside his or her employees. As the Supreme Court explained: “Plans that cover only sole owners or partners and their spouses . . . fall outside [ERISA] Title I’s domain,” but “[p]lans covering working owners and their nonowner employees . . . fall entirely within ERISA’s compass.” Id. at 21. 19 Put differently, “if a benefit covers only working owners, it is not covered by [ERISA] Title I.” Yates, 541 U.S. at 21 n.6. Although the Supreme Court holding in Yates does not directly address working owners without employees, the fact that the opinion carefully distinguished between working owners with and without employees and their relative places within and beyond ERISA’s scope further confirms this Court’s conclusion that ERISA does not cover working owners without employees. It is notable that DOL makes no attempt to directly categorize working owners without employees as ERISA “employers”; rather, the Final Rule only categorizes them as employers when they join associations. DOL attempts this strategic move because longstanding interpretations of ERISA make clear that a working owner without employees is not an ERISA “employer” directly and cannot create an ERISA plan for himself. See Tr. at 73:17–74:10 (DOL acknowledging that such a plan cannot fall under ERISA); Yates, 541 U.S. at 21 n.6 (listing cases holding that a plan covering a working owner without employees does not qualify as an ERISA plan). But if a 19 In Yates, the Supreme Court examined both the text of ERISA and DOL’s regulations interpreting the definition of “employee benefit plan,” which prohibited employee benefit plans from including any plan covering working owners without employees. 541 U.S. at 21 (citing 29 C.F.R. § 2510.3-3(b)–(c)). However, the Supreme Court emphasized that plans covering working owners without employees fall outside of ERISA—not just outside of DOL regulations interpreting ERISA. Thus, even after DOL changed its regulations, the Court reads Yates as stating that working owners without employees remain beyond ERISA’s scope. The Supreme Court also noted (Yates, 541 U.S. at 21 n.6) general agreement among circuit courts that sole proprietors without employees were generally excluded from ERISA’s reach. See Slamen v. Paul Revere Life Ins. Co., 166 F.3d 1102, 1104 (11th Cir. 1999) (“[I]n order to establish an ERISA employee welfare benefit plan, the plan must provide benefits to at least one employee, not including an employee who is also the owner of the business in question.”); In re Watson, 161 F.3d at 597 (holding that sole shareholder is not participant in plan that covered only him); Schwartz v. Gordon, 761 F.2d 864, 868 (2d Cir. 1985) (holding that ERISA excludes plans “in which there are no employees and for which the fiduciary protections afforded by” ERISA are inapplicable). 37 working owner without employees cannot qualify as an “employer” in his own right, an association of such working owners is not truly an employer association—at most, it is an association of working owners without employees. An association offering benefits to a sole proprietor without employees most closely resembles an “entrepreneurial venture,” and the relationship between association and working owner is most “similar to the relationship between a private insurance company . . . and the beneficiaries of a group insurance plan.” WEAIT, 804 F.2d at 1063. It is unreasonable to think that Congress imagined that its definition of an association acting “in the interest of” employers would extend to cover these individuals, who do not otherwise qualify as employer or employee. See ERISA § 3(5); 29 U.S.C. § 1002(5). Adding a working owner without employees to an association, then, does not change his status under ERISA—it cannot transform a sole proprietor without employees into either an “employer” or “employee” under the statute. But adding such a working owner to an association that otherwise could qualify as “bona fide” does change the status of that association. As illustrated by cases like WEAIT, an ERISA plan must be established or maintained to provide benefits to employees or former employees (or, in the case of an employee organization, its members) plus their beneficiaries, but never to unrelated third parties. See WEAIT, 804 F.2d at 1063 (analyzing ERISA § 3(1), 29 U.S.C. § 1002(1)). If an otherwise qualifying association provides benefits to such a working owner without employees through its AHP, it has severed the economic and representational ties that bind its “employer” members together because the working owner without employees would be a non-employer member. See id. By offering benefits to unaffiliated individuals through its AHP, the association would not qualify as an ERISA “employer,” and its AHP would not qualify as a single ERISA plan. 38 The Final Rule’s expansion of the definition of employer to include working owners without employees also does not further ERISA’s purposes. For example, allowing individuals without employees to join ERISA plans does not serve “as an incentive to the creation of plans that will benefit employer and nonowner employee alike.” Yates, 541 U.S. at 17–18. Further, the Final Rule does not address any of Congress’s concerns about nondisclosure, insolvency, or abuse in employee benefit plans as laid out in ERISA’s statement of purpose. See 29 U.S.C. § 1001. As the Second Circuit has explained, “[a] self-employed individual[,] . . . unlike a worker employed by another, has complete control over the amount, investment and form of the fund created by him . . . .” Schwartz, 761 F.2d at 868. Accordingly, ERISA’s “remedial scheme . . . for workers employed by others [is] not necessary for the protection of self-employed persons, and Congress accordingly has not changed the definition of ‘employee’ in that Title to include self-employed persons.” Id. In short, the Final Rule’s expansion of the term “employer” under ERISA to include working owners without employees (when organized in an association) is unreasonable because it is contrary to ERISA’s text and purpose. DOL’s interpretation does not “interpret” ERISA, it rewrites it, obliterating the statute’s references to “employers,” employees “employed by” those employers, and plans “established or maintained by an employer . . . for the purpose of providing for” those employees. It is also unsupported by case law interpreting ERISA from the Supreme Court and other courts. For these reasons, the Court concludes that the working owner provision of the Final Rule is contrary to ERISA and must be set aside. C. DOL’s Unreasonable Regulatory Interpretation of ERISA Creates Absurd Results Under the ACA Finally, the Final Rule creates absurd results under the ACA. This conflict further highlights that DOL’s regulatory interpretation of “employer” under ERISA is unreasonable. 39 As noted above, the ACA defines “employer” and “employee” generally based on how those terms are defined in ERISA. The ACA’s definition of “employee” is the same as under ERISA: it is simply an “individual employed by an employer,” 42 U.S.C. § 300gg-91(d)(5), 29 U.S.C. § 1002(6), which the Supreme Court has held to “incorporate traditional agency law criteria for identifying master-servant relationships,” Darden, 503 U.S. at 319. However, the ACA’s definition of “employer” adopts ERISA’s definition of “employer” with a critical exception: “such term shall include only employers of two or more employees.” 42 U.S.C. § 300gg-91(d)(6). Under the ACA, then, an “employer” must have two or more “employees,” and, under Darden, one expects these individuals to be cognizable as such under the common law. See Darden, 503 U.S. at 319. DOL contends that an ERISA “bona fide association” comprised solely of two working owners without employees would qualify as both an “employer” and “employee” under the ACA. See Final Rule, 83 Fed. Reg. at 28,940–41. DOL’s explanation for how these self-employed individuals fit inside ACA’s definition of employer is a magic trick. First, DOL asks the Court to imagine each working owner without employees wearing two hats: an employer hat and an employee hat. Tr. at 75:4–15. When two working owners without employees associate, they keep their employer hats, but the association also gains an employer hat. Id. at 75:16–25. The association counts as an employer under the ACA because the two working owners without employees also wear their employee hats, and therefore the association is an “employer[] of two or more employees.” Id. at 76:1–11; 42 U.S.C. § 300gg-91(6). Note that the working owners without employees must hang on to their employer hats as well, because otherwise the association would not qualify as an employer under ERISA, which requires the association to act in the interest of its employer members. ERISA § 3(5); 29 U.S.C. § 1002(5). As DOL explained, this analysis 40 “squares everything regarding working owners into one very neat package,” and the association qualifies as an employer under the ACA “as long as the association has at least two individuals to whom it is providing benefits, be it individuals, be it small companies, it doesn’t matter.” Tr. at 76:10–17. This logic is clever but ultimately not persuasive. When one counts the employees employed by two self-employed persons without employees, the sum is zero. DOL’s feat of prestidigitation transforms two individuals, neither of whom works for the other, into a total of three employers and two employees. This interpretation strains the ERISA definition of “employee,” which contemplates an individual “employed by” another. It doubly strains the ACA’s express limit of employers to “employers of two or more employees,” which contemplates two individuals employed by another. An association of two working owners without employees has no employers or employees—DOL’s explanation is pure legerdemain. The absurdity of DOL’s interpretation is compounded when one considers an association of fifty-one working owners without employees. Counting the number of employees employed by fifty-one working owners without employees, the Court again reaches a sum of zero. Yet under DOL’s explanation, this group of fifty-one associated individuals, none of whom employ anyone, counts fifty-two employers (counting the association “employer”) and fifty-one employees. This association would qualify not only as an ACA “employer,” but as a “large employer” free from the ACA’s individual and small-group market requirements. See 42 U.S.C. § 300gg-91(e)(2)–(3). The Court cannot believe that Congress crafted the ACA, with its careful statutory scheme distinguishing rules that apply to individuals, small employers, and large employers, with the intent that fifty-one distinct individuals employing no others could exempt themselves from the 41 individual market’s requirements by loosely affiliating through a so-called “bona fide association” without real employment ties. DOL’s explanation of how the Final Rule operates under the ACA relies on a tortured reading of the ACA’s statutory text that undermines the market structure that Congress so carefully crafted. DOL’s regulatory interpretation sows discord among the Final Rule, ERISA, and the ACA, which serves as further evidence that the Final Rule unreasonably interprets ERISA and fails to carry out congressional intent. CONCLUSION Upon consideration of the parties’ positions, as argued orally and in their briefs, as well as the administrative record, the relevant statutes, legal precedent, and the entire record herein, the Court concludes that the bona fide association and working owner provisions of the Final Rule, codified at 29 C.F.R. §§ 2510.3-5(b), (c) and (e), are unreasonable interpretations of ERISA. The Final Rule was intended and designed to end run the requirements of the ACA, but it does so only by ignoring the language and purpose of both ERISA and the ACA. DOL unreasonably expands the definition of “employers” to include groups without any real commonality of interest and to bring working owners without employees within ERISA’s scope despite Congress’s clear intent that ERISA cover benefits arising out of employment relationships. Accordingly, these provisions are unlawful and must be set aside, pursuant to this Court’s authority under the APA, 5 U.S.C. § 706. The Final Rule’s bona fide association and working owner provisions are therefore vacated. The Final Rule includes a severability provision. Under it, if a provision is found entirely invalid then “the provision shall be severable from [the Final Rule] and shall not affect the remainder thereof.” 29 C.F.R. § 2510.3-5(g). In light of this provision, the Court remands the Final Rule to the agency for consideration in the first instance of how the severability provision affects the remaining portions of the Final Rule. 42 A separate order will issue on this date. /s/ JOHN D. BATES United States District Judge Dated: March 28, 2019 43