Commonwealth of Ky. v. Janet Yellen

54 F.4th 325

• Court: Court of Appeals for the Sixth Circuit
• Decided: November 18, 2022
• Docket: 21-6108
• Status: Published

Opinion

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 22a0245p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

┐ COMMONWEALTH OF KENTUCKY; STATE OF TENNESSEE, │ Plaintiffs-Appellees, │ > No. 21-6108 v. │ │ JANET YELLEN, in her official capacity as Secretary of the │ U.S. Department of the Treasury; RICHARD K. DELMAR, in │ his official capacity as Acting Inspector General of the U.S. │ Department of the Treasury; UNITED STATES DEPARTMENT │ OF THE TREASURY, │ Defendants-Appellants. │ ┘

Appeal from the United States District Court for the Eastern District of Kentucky at Frankfort. No. 3:21-cv-00017—Gregory F. Van Tatenhove, District Judge.

Argued: July 21, 2022

Decided and Filed: November 18, 2022

Before: DONALD, BUSH, and NALBANDIAN, Circuit Judges. _________________

COUNSEL

ARGUED: Daniel Winik, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Appellants. Brett R. Nolan, OFFICE OF THE ATTORNEY GENERAL OF KENTUCKY, Frankfort, Kentucky, for Appellees. ON BRIEF: Daniel Winik, Alisa B. Klein, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Appellants. Brett R. Nolan, Barry L. Dunn, Matthew F. Kuhn, OFFICE OF THE ATTORNEY GENERAL OF KENTUCKY, Frankfort, Kentucky, Andrée S. Blumstein, Brandon J. Smith, OFFICE OF THE ATTORNEY GENERAL AND REPORTER OF TENNESSEE, Nashville, Tennessee, for Appellees. Paul D. Clement, KIRKLAND & ELLIS LLP, Washington, D.C., Joseph D. Henchman, NATIONAL TAXPAYERS UNION FOUNDATION, Washington, D.C., Sheng Li, NEW CIVIL LIBERTIES ALLIANCE, Washington, D.C., Drew C. Ensign, OFFICE OF THE ATTORNEY GENERAL OF ARIZONA, Phoenix, Arizona, for Amici Curiae. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 2

BUSH, J., delivered the opinion of the court in which DONALD, J., joined in full, and NALBANDIAN, J., joined in part. NALBANDIAN, J. (pp. 43–51), delivered a separate opinion concurring in part and dissenting in part.

_________________

OPINION _________________

JOHN K. BUSH, Circuit Judge. In response to the grave economic challenges posed by COVID-19, Congress enacted the American Rescue Plan Act of 2021 (“ARPA” or “the Act”). Pursuant to Congress’s spending power, ARPA set aside $195.3 billion in stimulus funds, to be distributed by the Treasury Department to states and the District of Columbia. This appeal concerns a challenge brought by Kentucky and Tennessee (“the States”) to what they allege is an ambiguous, coercive, and commandeering condition attached to those funds. Specifically, to get the money, the States had to certify that they would comply with the Act’s “Offset Provision.” Its terms bar the States from enacting tax cuts and then using ARPA funds to “directly or indirectly offset a reduction in [their] net tax revenue” resulting from such tax cuts. 42 U.S.C. § 802(c)(2)(A). And a related portion of the Act explains that should a State violate the Offset Provision, Treasury may initiate a recoupment action to recover the misused funds. 42 U.S.C. § 802(e)(1)–(2).

What the Offset Provision actually means, however, is the subject of grave dispute. Because money is fungible, enacting any tax cut and then spending ARPA funds could be construed, the States say, as having impermissibly used those funds to “indirectly offset” a revenue reduction from the tax cut. Appellees’ Br. at 12–13. As a result, should the States wish to expend their ARPA funds, they are effectively barred from enacting any tax cuts1—despite their desire to do so—for fear that Treasury could construe the cuts as implicating an “indirect offset” and correspondingly pursue recoupment. Id. at 22–23; 38. Compounding the Act’s indeterminacy, the Offset Provision itself never explains which fiscal year (“FY”) serves as the baseline for calculating a “reduction” in net tax revenue. Id. at 13, 40. That omission allegedly

1 This alleged restriction applies at least during ARPA’s “covered period,” 42 U.S.C. § 802(g)(1), which extends until “the last day of the fiscal year of such State . . . in which all funds received by the State . . . have been expended or returned to, or recovered by, the Secretary,” § 802(g)(1)(B). No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 3

leaves the States in the dark about when Treasury may deem them to have violated the Act. Id. And even though a Treasury regulation has since offered a narrowing construction of the Offset Provision, the States assert that this construction in no way follows clearly from the text of the Offset Provision itself. Id. at 41. Thus, the States object that the Offset Provision failed to provide them with clear notice of whatever conditions it entails. And because of those indeterminacies, they contend that the Offset Provision is unenforceable under the clear- statement rule the Supreme Court has long instructed governs spending legislation.

Worse yet, the States argue, they were coerced into relinquishing this control over their sovereign taxing authority. Amended Complaint ¶74, R. 23. By offering such a massive aid package—promising to confer on the States a sum equal to one-fifth of their annual budgets—in a time of fiscal crisis no less, the federal government made the States an offer they couldn’t refuse. Appellees’ Br. at 4, 12. Given these alleged intrusions upon their sovereignty, the States filed suit against the Treasury Department. They sought an injunction of the Offset Provision’s enforcement and a declaratory judgment that the provision is unenforceable.

Relying on the coercion rationale alone, the district court granted the States a permanent injunction in September 2021. Treasury’s appeal of that order is now before us. It asserts that the States’ challenges are nonjusticiable and that, in any event, their objections to the Offset Provision fail on the merits.

We agree that Kentucky’s challenge is nonjusticiable. At the outset of their suit, both Kentucky and Tennessee had standing to bring their pre-enforcement challenges, since the Offset Provision itself at least arguably proscribed the post-acceptance enactment of any revenue- reducing tax cut. Thus, the Offset Provision at least arguably threatened a significant intrusion upon state taxing authority—an intrusion that arguably offended the Spending Clause because it was not clearly authorized by the Offset Provision itself. But Treasury later promulgated an implementing regulation (“the Rule”) that disavowed this interpretation of the Offset Provision and established certain safe harbors permitting the States to cut taxes. See Coronavirus State and Local Fiscal Recovery Funds, 86 Fed. Reg. 26,786 (proposed May 17, 2021) (interim final rule); see also Coronavirus State and Local Fiscal Recovery Funds, 86 Fed. Reg. 4,338 (Jan. 27, 2022) (final rule); 31 C.F.R. § 35 et seq. In response, Kentucky and Tennessee offered no additional No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 4

evidence of a concrete plan to violate the Rule, so they failed to establish that Treasury will imminently seek recoupment because of any demonstrated policy they wish to pursue. And because Kentucky offered no evidence for any other theory of injury, the Rule mooted its challenge to the Offset Provision. We thus reverse the district court’s conclusion that Kentucky’s claim is justiciable and vacate the injunction to the extent that it bars enforcement of the Offset Provision against Kentucky.

Tennessee, by contrast, did adduce additional evidence of a distinct theory of injury: that Treasury’s Rule (and the underlying Offset Provision it implements) burden the State with compliance costs. See Eley Dec., R. 25-3. These costs represent additional labor and other expenses that Tennessee must incur to ensure that its recent and proposed tax cuts do not violate the Offset Provision; expenses that it would not incur were enforcement of the Offset Provision enjoined. Far from mooting the compliance-costs theory of injury, the Rule in fact exacerbated the harm with its more detailed explanation of the measures required to comply with the Offset Provision. Thus, we hold that Tennessee’s challenge is justiciable.

On the merits of Tennessee’s claim, we affirm the district court’s injunction on the basis that the Offset Provision is impermissibly vague under the Spending Clause. Because the Offset Provision is subject to a range of plausible meanings, Tennessee was deprived of the requisite “clear notice” of ARPA’s conditions when it accepted the funds. Cummings v. Premier Rehab Keller, P.L.L.C., 142 S. Ct. 1562, 1574 (2022) (quoting Arlington Cent. Sch. Dist. Bd. of Educ. v. Murphy, 548 U.S. 291, 296 (2006)). As a result, Treasury cannot use its Rule to impose compliance requirements upon Tennessee that are not clearly authorized by the Offset Provision itself. And because this defect suffices to affirm, we need not consider Tennessee’s additional objections to the Offset Provision.

I.

Congress enacted ARPA in March 2021 to make available almost $2 trillion in COVID- related relief funding. Approximately $195.3 billion of that sum was set aside for distribution to the states and the District of Columbia. “Kentucky’s allotment under the Act is about $2.1 billion,” while Tennessee’s is about $3.7 billion. Amended Complaint ¶¶26–27, R. 23. These No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 5

sums amount to nearly one-fifth of the States’ respective annual general revenues. Id. In the States’ view, “[t]he financial aid the Act offer[ed] . . . is simply unparalleled in size.” Id. ¶28.

That offer also came with several conditions. For instance, the States may spend their ARPA funds in only four particular areas that Congress deemed relevant to economic recovery from the pandemic. Those four areas are as follows:

(A) to respond to the public health emergency with respect to the Coronavirus Disease 2019 (COVID-19) or its negative economic impacts, including assistance to households, small businesses, and nonprofits, or aid to impacted industries such as tourism, travel, and hospitality; (B) to respond to workers performing essential work during the COVID-19 public health emergency by providing premium pay to eligible workers of the State, territory, or Tribal government that are performing such essential work, or by providing grants to eligible employers that have eligible workers who perform essential work; (C) for the provision of government services to the extent of the reduction in revenue of such State, territory, or Tribal government due to the COVID-19 public health emergency relative to revenues collected in the most recent full fiscal year of the State, territory, or Tribal government prior to the emergency; or (D) to make necessary investments in water, sewer, or broadband infrastructure.

42 U.S.C. § 802(c)(1)(A)–(D).

Conversely, the States are specifically forbidden from using ARPA funds for two particular applications. First, “[n]o State or territory may use funds made available under this section for deposit into any pension fund.” § 802(c)(2)(B). And second—the crux of this lawsuit—the States may not use ARPA funds:

to either directly or indirectly offset a reduction in the net tax revenue of such State or territory resulting from a change in law, regulation, or administrative interpretation during the covered period that reduces any tax (by providing for a reduction in a rate, a rebate, a deduction, a credit, or otherwise) or delays the imposition of any tax or tax increase.

§ 802(c)(2)(A). This is the so-called “Offset Provision”—the States have dubbed it the “Tax Mandate”—that has provoked legal challenges across the country. See, e.g., Missouri v. Yellen, 39 F.4th 1063 (8th Cir. 2022); Arizona v. Yellen, 34 F.4th 841 (9th Cir. 2022); West Virginia v. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 6

U.S. Dep’t of Treas., No. 7:21-cv-00465-LSC, 2021 WL 2952863, *1 (N.D. Ala. July 14, 2021); Texas v. Yellen, No. 2:21-CV-079-Z, 2022 WL 1063066, *1 (N.D. Tex. Apr. 8, 2022).

Accompanying the Offset Provision are a couple of related enforcement mechanisms. First is the statute’s reporting requirement, which instructs the states:

To provide to the Secretary periodic reports providing a detailed accounting of— (A) the uses of funds by such State, territory, or Tribal government, including, in the case of a State or a territory, all modifications to the State’s or territory’s tax revenue sources during the covered period; and (B) such other information as the Secretary may require for the administration of this section.

§ 802(d)(2)(A)–(B). And second is the statute’s recoupment procedure. Should a state violate the Act’s requirements, Treasury may initiate a recoupment action to seek reimbursement from a state “equal to the amount of funds used in [the] violation.” § 802(e).

Kentucky and Tennessee were not alone, it turns out, in their apprehensions about this statutory scheme. Several of their sister-states were similarly puzzled by the Offset Provision’s requirements. So they wrote jointly to Secretary Yellen to seek clarification about the precise obligations it imposed. Secretary Yellen—who elsewhere had acknowledged that the “fungibility of money” presented “thorny questions” about the meaning of the Offset Provision—wrote back to explain that the States could expect “further guidance” from Treasury in the near future. Treasury Secretary & Federal Reserve Chair Testimony on COVID-19 Economic Recovery, C-SPAN (Mar. 24, 2021), at 58:00–59:11, available at https://www.c- span.org/video/?510059-1/treasury-secretary-federalreserve-chair-testimony-covid-19-economic- recovery; see Yellen Letter, R. 1-2.2 But she also explained that Treasury did intend to enforce whatever prohibitions the Offset Provision was revealed to entail. See id. Perhaps

2 In particular, Senator Mike Crapo asked Secretary Yellen, “How do you intend to approach the question of what is ‘directly or indirectly offsetting’ a tax cut?” The Secretary responded, “Well, when I said that we have ‘thorny questions’ to work through, you’ve just indicated why we do. We will have to define what it means to use money from this Act as an ‘offset’ for tax cuts. And, given the fungibility of money, it’s a hard question to answer.” Treasury Secretary & Federal Reserve Chair Testimony on COVID-19 Economic Recovery at 58:30–59:05 (emphasis added), available at https://www.c-span.org/video/?510059-1/treasury-secretary-federalreserve-chair- testimony-covid-19-economic-recovery. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 7

unsurprisingly, given these lingering uncertainties, several states filed suit to restrain the Offset Provision’s enforcement.

Kentucky and Tennessee brought their own challenge in April 2021. They each alleged that because of the funds’ irresistible nature in the midst of an economic crisis, they intended to accept their respective funding allotments.3 But they also alleged that the Offset Provision tied to those funds injured the States with a coercive and ambiguous restriction that “unconstitutionally intrud[es] on the [States’] sovereign authority, by interfering with their ordinary management of their fiscal affairs, and by requiring them to forgo their constitutional taxing powers or face an action to return much-needed federal funds after they have already been spent.” Complaint ¶12, R. 1.

In response to this and other suits, Treasury attempted to clarify the Offset Provision by promulgating an Interim Final Rule in May 2021. See 86 Fed. Reg. at 26,786. In relevant part, the Interim Final Rule explained that Treasury did not read the Offset Provision to proscribe all tax cuts during ARPA’s “covered period.” Id. at 26,807. Rather, it views the Provision as proscribing only a tax cut that (1) results in a revenue reduction as compared to revenues for the “fiscal year ending in 2019,” and (2) for which a state fails to identify a permissible, non-ARPA source of additional funds to offset the revenue reduction. Id.; see also id. at 26,810. In particular, Treasury said, it would not initiate a recoupment action even after a state enacted a revenue-reducing tax cut and expended ARPA funds so long as the state could show that the revenue reduction was offset with (1) a state tax increase on some other activity, (2) additional inlays from macroeconomic growth, or (3) a state spending cut in an area the state is not expending ARPA funds. Id.; see also Appellants’ Br. at 5; 31 C.F.R. §§ 35.1–35.12 (codifying the Rule).

The States reacted with an amended complaint in June 2021. The Interim Final Rule notwithstanding, the States reprised their contention that the Offset Provision functionally

3 Kentucky ultimately accepted the funds after the complaint was filed and certified that it would comply with the Offset Provision, while Tennessee accepted the funds only after the district court entered its permanent injunction. See Recording of Oral Arg. at 27:38–27:55. Additionally, Tennessee accepted the funds with a reservation that it considered the Offset Provision invalid. Id. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 8

proscribes all future tax cuts to the extent a state wishes to expend its ARPA funds. Amended Complaint ¶32, R. 23 (alleging that Congress, as a condition of ARPA, required the States to “promise that [they] will not lower taxes on their residents for four years”). But they augmented their complaint with an allegation about compliance costs. See id. ¶12. In addition to their “imminent recoupment” and “sovereign authority” theories of injury, the States complained that “the Tax Mandate w[ould] impose administrative burdens on [them] by obligating them to spend resources on calculation and reporting requirements.” Id. And the States alleged that all of those injuries “are traceable to the Tax Mandate and Defendants’ efforts to enforce it.” Id. Thus, they continued to seek a declaratory judgment that the Offset Provision is unconstitutional and an injunction to restrain Treasury from initiating an enforcement action.

Two days later, the States submitted their corresponding motion for summary judgment and a permanent injunction. They reiterated their view that they have standing to challenge the Offset Provision for the three aforementioned reasons: that it intrudes on state taxing authority, could result in a recoupment action if the States were to pursue their desired tax cuts, and imposes administrative burdens and compliance costs. As to the merits, they argued that the Offset Provision is impermissibly ambiguous under the Spending Clause, not reasonably related to ARPA’s nominal goal of fiscal recovery, and unconstitutionally coercive and commandeering.

Given Treasury’s strenuous objections to justiciability, the corresponding evidence the States submitted in support of their motion for summary judgment deserves particular scrutiny. Kentucky offered merely a confirmation email indicating its acceptance of the ARPA funds. See Submission Confirmation, R. 25-1. By contrast, Tennessee submitted declarations from two state officials. First was a declaration from N. Antonio Niknejad, Policy Director to Governor Bill Lee. See Niknejad Dec., R. 25-2. Niknejad explained that Tennessee has “a long history of cutting taxes and spending in order to spur economic growth,” that Tennessee had recently enacted several tax cuts on gym memberships, professional licensing, agricultural products, and broadband fiber optic cables, and that Tennessee is contemplating several future tax cuts. Id. ¶¶6, 8, 9–11. Yet he explained that uncertainty about how the Offset Provision could be construed has caused policymakers in Tennessee to “defer, slow, or reconsider some of [their] taxing decisions.” Id. ¶14. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 9

The second was a declaration from Commissioner Howard H. Eley of Tennessee’s Department of Finance and Administration. See Eley Dec., R. 25-3. Unlike Niknejad, who focused on anticipated tax cuts, Eley described the administrative burdens and compliance costs the Offset Provision (and Rule) would inflict on Tennessee. We quote three particularly relevant paragraphs from his declaration below:

8. Tennessee is required by its state constitution to enact a balanced budget. General fund expenditures, which include certain reductions in tax revenue due to a statutory or regulatory change, are described by category, agency, program, and the recurring or non-recurring nature of the expenditure. State revenues, which include federal funds and reimbursements, are described by source. The enacted budget appropriates a specific amount from the general fund and other funds to fund the State’s programs and operations. But in determining whether the budget is balanced, the Department of Finance and Administration generally compares total expenditures to total revenues and does not typically connect expenditures to specific revenue sources or “indirect” causes for those revenues. If the State receives federal funds to offset certain state expenditures, the state funds that would have been used to pay for those expenditures are not used and can be returned to the general fund for future appropriation. To comply with the Tax Mandate, the Department will be required to create new accounting processes that specifically track whether federal funds received under the Rescue Plan are being used to “directly or indirectly offset” any state expenditures resulting from a reduction in tax revenue that otherwise would have been funded from state appropriated tax revenues. That will include tracking whether any cost savings resulting from the receipt of federal funds to offset certain state expenditures are ultimately and indirectly used to offset a tax reduction. Establishing these additional processes and preparing the required reports will require at least one budget analyst and one revenue analyst to divert at least some of their work to that task and other state employees to support and review that work. 9. To comply with the Secretary’s Regulations attempting to implement and enforce the Tax Mandate, the State of Tennessee will be forced to expend additional resources adjusting Tennessee’s “baseline” level of tax revenue for inflation each year during the covered period “using the Bureau of Economic Analysis’s Implicit Price Deflator for the gross domestic product of the United States” and then using that adjusted figure to determine whether the State’s tax policies may violate the Secretary’s interpretation of the Tax Mandate. 31 C.F.R. §§ 35.3, 35.8(b). 10. The State of Tennessee would not incur these additional costs to determine whether any revenue reductions could be said to have been “directly or indirectly offset” by funds received under the Rescue Plan or to report its revenue modifications to the Secretary but for the Tax Mandate. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 10

Eley Dec. ¶¶8–10, R. 25-3 (emphases added). Given both these compliance costs and the asserted threat of a recoupment action should the States pursue their desired tax cuts, the States asked the district court to grant them summary judgment and permanently enjoin the Offset Provision’s enforcement.

Treasury cross-moved for summary judgment, or, alternatively, to dismiss the complaint. It introduced no evidence of its own, and thus it did not attempt to controvert Niknejad or Eley’s declarations. Rather, it argued that even taking the declarations as true, the States lacked standing and that their merits challenges failed as a matter of law. See Mot. to Dismiss & Mot. for Summ. J. at 8, 17, R. 32. Concerning a recoupment action, Treasury argued that none was imminent. Id. at 10. For even if the States had established that they wish to cut taxes, they failed to show that not only would such cuts result in revenue reductions, but also that they intended to use ARPA funds to offset those reductions. Id. at 11. As to compliance costs, it argued that (1) any administrative burdens were traceable solely to the reporting requirement, not the Offset Provision, and (2) no injury occurs because the States are permitted to use ARPA funds “to cover administrative costs.” Id. at 14–15. As to the merits, Treasury conceded that its Rule cannot cure potential ambiguities in the Offset Provision for purposes of the Spending Clause. Id. at 30. But it claimed that the text of the Offset Provision itself is unambiguous. Id. It likewise argued that no precedents support the States’ view that the Offset Provision is unduly coercive or commandeers state taxing authority. Id. at 18–26.

The district court rendered its opinion on these motions in September 2021. As to justiciability, it concluded that both Kentucky and Tennessee had satisfied the pre-enforcement- challenge standing test described in Susan B. Anthony List. See id. at 3–5 (citing Susan B. Anthony List v. Driehaus, 573 U.S. 149 (2014)). First, both the States intended to accept ARPA funds and yet asserted that doing so entailed compliance with the arguably unconstitutional “Tax Mandate.” Op. & Order at 4–5. Second, the “Tax Mandate” at least arguably proscribed the States’ desired efforts to cut taxes. Id. at 5. And third, Secretary Yellen had expressed intent to enforce the Offset Provision in her earlier letter to the States, demonstrating a credible threat of enforcement. Id. The district court thus held that both Kentucky and Tennessee had standing to challenge the Offset Provision. Id. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 11

As for the merits, the district court concluded that ARPA violated the Spending Clause because it had coerced the States into relinquishing control over their taxing authority to the federal government. Id. at 11. In essence, it said, the economic crisis made the federal government’s aid offer irresistible, and so it represented an “undue influence” on the States’ authority to tax. Id. at 6 (quoting Nat’l Fed. of Indep. Bus. v. Sebelius, 567 U.S. 519, 576 (2012) (opinion of Roberts, C.J.)); see also id. at 11. The district court thus granted summary judgment to the States and imposed a permanent injunction4 restraining enforcement of the Offset Provision. Id. at 16–17.

II.

The district court’s order granting the States summary judgment and imposing a permanent injunction was a final decision. See, e.g., Reform Am. v. City of Detroit, 37 F.4th 1138, 1147 (6th Cir. 2022) (citation omitted). Thus, 28 U.S.C. § 1291 gives us statutory jurisdiction to handle Treasury’s appeal. We examine Article III jurisdiction over the States’ respective claims below.

As for our standards of review, we consider summary-judgment orders de novo. See Jordan v. Howard, 987 F.3d 537, 542 (6th Cir. 2021) (citation omitted). Thus, drawing all reasonable inferences in favor of the non-movant, we ask whether the party seeking summary judgment demonstrated “that there is no genuine dispute as to any material fact” and that it is “entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). On cross-motions for summary judgment, we apply these same standards to each of the individual motions. See Taft Broad. Co. v. United States, 929 F.2d 240, 248 (6th Cir. 1991) (citation omitted); accord Reform Am., 37 F.4th at 1147; B.F. Goodrich Co. v. U.S. Filter Corp., 245 F.3d 587, 592 (6th Cir. 2001).

Concerning the district court’s decision to grant a permanent injunction, several standards of review are relevant. “[F]actual findings are reviewed under the clearly erroneous standard, legal conclusions are reviewed de novo, and the scope of injunctive relief is reviewed for an abuse of discretion.” Sec’y of Lab. v. 3Re.com, Inc., 317 F.3d 534, 537 (6th Cir. 2003) (quoting

4 By contrast, the district court denied the States’ requested declaratory judgment, reasoning that such relief was subsumed into its order awarding a permanent injunction. Op. & Order at 14, R. 42. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 12

S. Cent. Power Co. v. Int’l Brotherhood. of Elec. Workers, Loc. Union 2359, 186 F.3d 733, 737 (6th Cir. 1999)).

III.

As is our obligation, we consider first whether Kentucky and Tennessee established that their respective challenges to the Offset Provision are justiciable. See, e.g., Arbaugh v. Y&H Corp., 546 U.S. 500, 514 (2006). We hold that Tennessee alone satisfied that showing. We then explain our view that the text of the Offset Provision is insufficiently clear under the relevant Spending Clause jurisprudence for Treasury, through promulgation of its Rule, to impose the specific obligations that Tennessee complains have inflicted compliance costs upon it.

A. Justiciability

1. Kentucky and Tennessee’s Initial Standing to Sue

From the States’ original complaint onward, their central theory of standing has been as follows. First, they said, they both intended to accept ARPA funds. Complaint ¶¶26–27, R. 1. But second, the Offset Provision at least arguably proscribes enacting any post-acceptance tax cut should the States wish to expend their funds. Id. ¶32. Indeed, because money is fungible, spending ARPA funds and then cutting taxes (or vice versa) could arguably be construed as having used those funds to “indirectly offset” a resultant revenue reduction. Id. ¶35. And second, the States alleged, both Kentucky and Tennessee desire to enact (or have enacted) tax cuts. Kentucky, for instance, recently enacted a tax-deferral bill to revitalize an area of Louisville. Id. ¶41. Likewise, Tennessee is considering eliminating its professional-privilege tax, and it has recently enacted cuts to several other taxes. Id. ¶42. But the States complained that such tax cuts “could be construed to come within the Tax Mandate if they result in a revenue decrease.” Id. They thus contended that the Offset Provision constrained their sovereign authority to tax and exposed them to an imminent recoupment action should they wish to pursue their preferred policies.

These original theories sufficed for standing. Whether a party has standing to redress an injury is measured as of the time the injury is first asserted; here, in the original complaint. See No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 13

Lynch v. Leis, 382 F.3d 642, 647 (6th Cir. 2004). As of that moment, therefore, we apply two relevant frameworks to assess whether these “imminent-recoupment” and “sovereign-authority” theories sufficed for standing. The first framework derives from the Supreme Court’s decision in Lujan, which explained that plaintiffs must establish an injury that is (1) actual or imminent and concrete and particularized, (2) traceable to the defendant, and (3) likely to be redressed by a favorable decision. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61 (1992). Here at least, elements (2) and (3) are not subject to serious dispute. A recoupment action initiated by Treasury is no doubt traceable to Treasury, and an injunction restraining such a proceeding would provide the States corresponding relief. But what about an injury in fact? No recoupment action is now pending. So the question is whether a future such proceeding is sufficiently imminent to say the States have suffered a de facto injury for purposes of Article III.

That brings us to the second, more specialized framework, which instructs us how to determine whether an enforcement action is sufficiently imminent to support Article III jurisdiction over a pre-enforcement challenge. See Susan B. Anthony List, 573 U.S. at 158–59. Under that test, we ask whether the States, when they first asserted these injuries, had established (1) an intention to engage in a course of conduct arguably affected with a constitutional interest, (2) that this course of conduct was arguably proscribed by the Offset Provision, and (3) that if the States should pursue such a course of conduct, there was a credible threat that Treasury would pursue a recoupment action. See id. at 161–64. Before the eventual advent of the Rule, we believe, the States had satisfied this tripartite showing.

First, Kentucky and Tennessee alleged that despite their intention to accept and expend ARPA funds, they had either enacted or planned to enact tax cuts that could potentially result in revenue reductions. Complaint ¶¶41–42, R. 1. And their decision to do so was at least arguably affected with a constitutional interest, given that states have a powerful sovereign prerogative under federalism principles to control their own internal taxation policies. See id. ¶40; see also Lane County v. Oregon, 74 U.S. 71, 76 (1868) (describing states’ control over “the power of taxation” as “indispensable” and “an essential function of government”); Dep’t of Revenue of Oregon v. ACF Indus., Inc., 510 U.S. 332, 345 (1994) (“Subsection (b)(4), like the whole of No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 14

§ 11503, sets limits upon the taxation authority of state government, an authority we have recognized as central to state sovereignty.”).

Second, this course of conduct was at least arguably proscribed by the Offset Provision. As we noted before, “money is fungible.” Complaint ¶35, R. 1; see also United States v. Sperry Corp., 493 U.S. 52, 62 n.9 (1989) (“Unlike real or personal property, money is fungible.”); Ransom v. FIA Card Servs., N.A., 562 U.S. 61, 79 (2011) (same); Ark Encounter, LLC v. Parkinson, 152 F. Supp. 3d 880, 904 (E.D. Ky. 2016) (“Because money is fungible, such benefits will to some extent have the incidental effect of allowing the institution’s other funds to be used to advance their [other] purposes if they wish. Indeed any reimbursement, aid, or tax exemption necessarily frees up other funds for other purposes.”). As a result, merely enacting a revenue-reducing tax cut and expending ARPA funds could at least arguably be construed as having used the funds to “indirectly offset” the revenue reduction, given that the ARPA funds could support continued state spending rendered otherwise impossible by the tax cuts.5

Indeed, Treasury acknowledged in the commentary to its own Final Rule that this is at least a plausible interpretation of the statute. For instance, it explained, “because money is fungible, even if [ARPA] funds are not explicitly or directly used to cover the costs of changes that reduce net tax revenue, those funds may be used in a manner inconsistent with the statute by indirectly being used to substitute for the state’s or territory’s funds that would otherwise have been needed to cover the costs of the reduction.” 87 Fed. Reg. at 4,424 (emphasis added). For that matter, the plausibility of the States’ money-is-fungible interpretation is the very reason Treasury had to promulgate its Rule—to disavow that interpretation and attempt to clarify the Offset Provision. See, e.g., id. at 4,423–24. Thus, the States’ desire to cut taxes while spending ARPA funds was at least arguably proscribed by the Offset Provision.

Last, the States had illustrated a credible threat of enforcement. For instance, the States produced Secretary Yellen’s letter indicating that Treasury intended to enforce the Offset

5 The question may arise why a revenue reduction would necessarily make additional spending impossible, since it would seem the States could continue to spend at the same levels by taking on debt. The answer is that this sort of debt-financed spending is restricted under Kentucky and Tennessee’s respective constitutions, which have balanced-budget amendments. See Ky. Const. §§ 49–50, 171; Tenn. Const. Art. II, § 24. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 15

Provision. See Yellen Letter, R. 1-2. The letter reiterated that “[ARPA] funding may not be used to offset a reduction in net tax revenue resulting from certain changes in state law.” Id. at 1. It also explained that Treasury would later promulgate “further guidance” about what sort of changes in state law could provoke a recoupment action.6 Id. at 1–2. Thus, the letter itself acknowledged that (1) the Offset Provision would be enforced, but (2) it was not yet clear, based on the statute alone, how the States could comply with the Provision (and stave off recoupment). So as of the original complaint, the States had satisfied the Supreme Court’s pre-enforcement- challenge test. In addition to traceability and redressability, in other words, they had also established a sufficiently imminent injury for jurisdiction.

2. The Interim Final Rule Complicates the Initial Imminent-Recoupment and Sovereign-Authority Theories of Injury

A little over a month after the States had filed their original complaint, Treasury promulgated its Interim Final Rule (“IFR”) offering its construction of the Offset Provision. Several features of that Rule are relevant to this dispute. First, the IFR supplied the missing baseline for calculating whether a tax cut results in a revenue reduction. It clarified that the revenue baseline would be the state’s “fiscal year 2019 tax revenue adjusted for inflation.” 86 Fed. Reg. at 26,808.7 Second, the IFR attempted to provide guidance about when a state would be understood to have “directly or indirectly offset a reduction in . . . net tax revenue.” 42 U.S.C. § 802(c)(2)(A). Treasury’s commentary explained as follows:

A recipient government would only be considered to have used Fiscal Recovery Funds to offset a reduction in net tax revenue resulting from changes in law, regulation, or interpretation if, and to the extent that, the recipient government could not identify sufficient funds from sources other than the Fiscal Recovery

6 Technically, even this letter explained Secretary Yellen’s position that the Offset Provision does not render tax cuts impermissible per se. See Yellen Letter at 1, R. 1-2. But the letter also acknowledged that the Offset Provision created significant uncertainty about when tax cuts were permissible versus when they were not, which was why Treasury intended to promulgate “further guidance” about when it would pursue recoupment actions. Id. at 2. Thus, in the absence of those clarifying regulations, it was at least arguable that the tax cuts Kentucky and Tennessee had enacted or planned to enact could provoke a recoupment action. Indeed, as this Circuit’s precedent recognizes, a threatened enforcement action should only be understood as too remote to support jurisdiction when the defendants have provided “clear assurances” they will not undertake the enforcement action. See, e.g., Universal Life Church Monastery Storehouse v. Nabors, 35 F.4th 1021, 1035 (6th Cir. 2022). 7 The Final Rule likewise confirms that the fiscal year ending in 2019 is the relevant baseline. See 31 C.F.R. § 35.3; 87 Fed. Reg. at 4,423. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 16

Funds to offset the reduction in net tax revenue. If sufficient funds from other sources cannot be identified to cover the full cost of the reduction in net tax revenue resulting from changes in law, regulation, or interpretation, the remaining amount not covered by these sources will be considered to have been offset by Fiscal Recovery Funds, in contravention of the offset provision. The interim final rule recognizes three sources of funds that may offset a reduction in net tax revenue other than Fiscal Recovery Funds—organic growth, increases in revenue (e.g., an increase in a tax rate), and certain cuts in spending.

86 Fed. Reg. at 26,807. The IFR thus provided the “further guidance” Secretary Yellen had promised in her initial letter to the States. As we explained before, Treasury construed the Offset Provision not to bar a revenue-reducing tax cut so long as a state identifies replacement funds from (1) macroeconomic growth, (2) increased state taxation on some other activity, or (3) state spending cuts in an area where the state is not expending ARPA funds. Id.; see also Appellants’ Br. at 5.

Yet this narrowing construction created apparent justiciability issues for the States’ challenge to the Offset Provision. In their initial complaint, the States had alleged only that they have enacted or plan to enact tax cuts that may result in reduced state revenues. Complaint ¶¶41–42, R. 1. They never additionally alleged that they would then fail to identify a permissible source of revenue—such as from macroeconomic growth or a reduction in certain state spending—to offset the resultant reductions in inlays. And only if that contingency were to occur, according to Treasury’s new Rule, would Treasury pursue a recoupment action against the States. See 86 Fed. Reg. at 26,807; see also Appellants’ Br. at 5. The IFR thus rendered it unclear why there was a reasonable prospect of a recoupment action.

And the States’ sovereign-authority theory now suffered from a similar issue. The States’ apparent view was that they had either been injured (1) in the past from the receipt of an ambiguous or coercive offer, or (2) are being continuously injured because the Offset Provision “prohibit[s] . . . tax relief.” Appellees’ Br. at 46. But an injunction cannot be used to redress a purely past injury. See City of Los Angeles v. Lyons, 461 U.S. 95, 105 (1983). Rather, the States had to show why they were likely to suffer some present or future harm. Id. So that leaves the No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 17

claim that the Offset Provision “prohibits . . . tax relief.” Appellees’ Br. at 46.8 But the IFR subsequently disavowed the States’ interpretation of the Offset Provision, clarifying that they remain free to expend ARPA funds and enact tax cuts resulting in revenue reductions so long as they identify a permissible source of offsetting funds. See, e.g., 86 Fed. Reg. at 26,807. And the Final Rule crystallized precisely the same understanding. See, e.g., 87 Fed. Reg. at 4,426.9 Yet the States never established that they would fail to meet that obligation. Thus, we do not see how the sovereign-authority theory could support injunctive relief when the States identified no specific course of conduct they wish to pursue but against which Treasury will initiate an enforcement proceeding. See Whole Woman’s Health v. Jackson, 141 S. Ct. 2494, 2495 (2021) (“[F]ederal courts enjoy the power to enjoin individuals tasked with enforcing laws, not the laws themselves.” (citing California v. Texas, 141 S. Ct. 2104, 2115–16 (2021)).

3. The States File Their Amended Complaint and Motion for Summary Judgment But Provide No Evidence that They Intend to Violate the Rule

About a month after the IFR’s promulgation, the States filed their amended complaint and corresponding motion for summary judgment and a permanent injunction. Despite the advent of the IFR, the States made no allegations and adduced no specific evidence about how

8 In response to our request for supplemental briefing on mootness, the States emphasized their contention that their “sovereign authority” theory remains live even despite the Rule because the Offset Provision still “limits the range of policy options available to the[m].” Appellees’ Supp. Br. at 4 (quoting Appellees’ Br. at 20–21). Of course, any law could be said to “limit the range,” in an abstract sense, of a plaintiff’s legitimate behavior. But merely because enjoining the law’s enforcement could be said to expand the range of potential behaviors a plaintiff might permissibly engage in does not alone establish the plaintiff’s standing to seek an injunction. Rather, the plaintiff must show that he is “able and ready” to violate the law and that an enforcement action would realistically and likely ensue in response to the violation. Carney v. Adams, 141 S. Ct. 493, 501–02 (2020); Babbitt v. United Farm Workers Nat. Union, 442 U.S. 289, 298 (1979). Thus, as concerns the “sovereign authority” theory, the States still had the burden to establish, with evidence, why they plan to imminently pursue some policy objective outside the range of conduct permitted by the Rule and against which Treasury would correspondingly take action. Lujan, 504 U.S. at 561; cf. Whole Woman’s Health v. Jackson, 141 S. Ct 2104, 2495 (2021) (“[F]ederal courts enjoy the power to enjoin individuals tasked with enforcing laws, not the laws themselves.” (citing California v. Texas, 141 S. Ct. 2104, 2115–16 (2021))). 9 We do not hold today that the interim final rule itself necessarily mooted the imminent-recoupment and sovereign-authority theories. Interim final rules are subject to revision after the notice-and-comment process, so a rule’s content could still change from its interim form to its final form in some way relevant to justiciability. But that concern is absent from this particular case, given that the final rule varied from the interim final rule in no way material to this dispute. Both disavow enforcement in exactly the same way and present exactly the same safe harbors for states and enforcement constraints on the Treasury Department. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 18

they have pursued or intend to pursue a course of conduct that would arguably violate the Rule. In other words, they provided no declarations or other evidence about how they intend to enact tax cuts that (1) would result in net revenue reductions compared to 2019 inlays, and (2) would then fail to identify a permissible funding source (such as from growth or spending cuts) to offset the revenue reduction. Indeed, the only evidence Kentucky adduced in the States’ motion for summary judgment was its notification that it intended to accept the ARPA funds. See Submission Confirmation, R. 25-1.

Those omissions are problematic for justiciability, since the States produced no evidence about why there is a realistic risk of an enforcement proceeding. And justiciability must be established with the degree of evidence required at each respective stage of the suit. See Lujan, 504 U.S. at 561. So the States were obliged to submit evidence—such as a sworn declaration— detailing how they are “able and ready” to pursue a course of action that would run afoul of the Rule. Carney v. Adams, 141 S. Ct. 493, 501–02 (2020). For only then would there be a demonstrated risk of a recoupment action, which a federal court could redress by enjoining such action. See Jackson, 141 S. Ct. at 2495 (citing California, 141 S. Ct. at 2115–16). In the absence of that evidence, we conclude that Treasury’s disavowal of the money-is-fungible interpretation dispelled the States’ claim that they run the risk of an imminent enforcement action—as when, for instance, a prosecutor credibly disavows that he will enforce a challenged statute. See, e.g., Mink v. Suthers, 482 F.3d 1244, 1256–57 (10th Cir. 2007); cf. Commodity Trend Serv., Inc. v. Commodity Futures Trading Comm’n, 149 F.3d 679, 687 (7th Cir. 1998) (explaining that courts will find a credible threat of enforcement when “the Government fails to indicate affirmatively that it will not enforce the statute” (emphasis omitted)).

The only remaining question is what kind of justiciability defect Treasury’s disavowal created. The parties initially framed the issue as one of the States’ “standing.” But we disagree with that characterization. Whether an “intervening circumstance” arising after a suit has been filed causes a plaintiff’s asserted injury to dissipate is really a question of mootness. See Genesis Healthcare Corp. v. Symczyk, 569 U.S. 66, 72 (2013). And whether the Rule mooted the imminent-recoupment and sovereign-authority theories comes down to whether we should credit Treasury’s voluntary disavowal of a broad view of the Offset Provision; in essence, whether No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 19

Treasury established10 that there is no “reasonable possibility” it will act as if the Offset Provision forbids tax cuts per se. Resurrection Sch. v. Hertel, 35 F.4th 524, 529 (6th Cir. 2022) (en banc).

We hold that Treasury satisfied this showing. Its Final Rule resulted from the notice-and- comment process, and thus it may be rescinded only pursuant to that process as well. See 5 U.S.C. § 551(5); Perez v. Mortg. Bankers Ass’n, 575 U.S. 92, 101 (2015). And we have no evidence that Treasury plans to pursue such rescission. Indeed, and more importantly, Treasury has repeatedly taken the position in this litigation that its Rule necessarily follows from the plain text of the Offset Provision itself. See, e.g., Reply Br. at 1; Appellants’ Supp. Br. at 2 n.2. So even without the Rule, according to Treasury, it would pursue recoupment against Kentucky and Tennessee—even if they were to enact a revenue-reducing tax cut and expend ARPA funds— only if the States additionally failed to identify one of the permissible sources of offsetting funds, such as a tax increase or macroeconomic growth.11 Id. On those bases, then, we conclude that Treasury has affirmatively and credibly disavowed the money-is-fungible interpretation of the Offset Provision. Thus, because the States failed to provide evidence that they intend to

10 Aside from how standing and mootness concern the parties’ interests at different stages of a lawsuit, they can also present different burdens of proof. See Cardinal Chem. Co. v. Morton Int’l, Inc., 508 U.S. 83, 98 (1993). The burden to establish jurisdiction rests on the party invoking jurisdiction—here, the States—while the burden to defeat jurisdiction with a mootness objection rests on the party asserting mootness—here, Treasury. Id. Because of this burden-shifting issue, we ordered supplemental briefing to solicit Treasury’s affirmative case as to why the States’ challenge is moot. See Order, ECF No. 45. After agreeing that mootness (rather than standing) is the appropriate framework to assess the impact of the Rule’s advent, Treasury’s supplemental brief once again disavowed the money-is-fungible interpretation of the Offset Provision and disclaimed that Treasury has any intent to pursue recoupment in response to tax cuts per se. See Appellants’ Supp. Br. at 1. We thus understand Treasury, through its supplemental briefing, to have discharged its duty to make an affirmative showing about why at least the imminent-recoupment and sovereign-authority theories are moot. 11 In their supplemental briefing, the States contended that the Rule did not moot the imminent-recoupment and sovereign-authority theories because it contains a reservation of authority. See Appellees’ Supp. Br. at 5 (citing 31 C.F.R. § 35.4(a)). And, true, the Rule provides that “[n]othing in this part shall limit the authority of the Secretary to take action to enforce conditions or violations of law, including actions necessary to prevent evasions of this subpart.” 31 C.F.R. § 35.4(a). But we have no evidence (or even argument) about how the States plan to engage in conduct that Treasury would construe as an “evasion,” much less why Treasury would construe behavior clearly permitted by other parts of the Rule to constitute such an “evasion.” Moreover, Treasury insists that the Rule’s narrowing construction “flows naturally from the text of the Offset Provision itself.” Appellants’ Supp. Br. at 2 n.2. So even if the Rule arguably did not bar Treasury from pursuing recoupment under, for instance, the money-is-fungible interpretation, Treasury has solemnly represented before us that the Offset Provision itself would preclude such an enforcement action. Again, Treasury’s position is that the Offset Provision itself unambiguously dispels that interpretation of the statute. Reply Br. at 1; Appellants’ Supp. Br. at 2. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 20

specifically violate the Rule (and provoke recoupment), and because Treasury established that there is no realistic prospect it will enforce the States’ expansive interpretation of the Offset Provision, we deem the imminent-recoupment and sovereign-authority theories moot.

Moreover, that holding ends the case for Kentucky. Kentucky submitted nothing other than an email indicating its intent to accept the ARPA funds. Submission Confirmation, R. 25-1. It furnished no proof about how it intends to violate the Rule, or about why it suffers a continuing sovereign injury when it identified no desired tax cut that, if enacted, would likely provoke recoupment. It also offered no additional theory of injury, such as compliance costs, that might sustain its challenge even in the absence of an imminent recoupment action. See Recording of Oral Arg. at 30:38–30:45 (conceding that there is no evidence in the record about Kentucky’s budgeting processes). Thus, the district court should have dismissed its challenge to the Offset Provision as moot.12

4. Tennessee’s Challenge Remains Justiciable, However, Under a Compliance-Costs Theory of Injury

The same cannot be said for Tennessee. Recall how, in their amended complaint, the States made allegations about an additional injury the Offset Provision would inflict upon them:

12 Kentucky insists that its challenge to the Offset Provision is justiciable if we determine that Tennessee’s challenge to the Offset Provision is justiciable. See, e.g., Recording of Oral Arg. at 30:15–30:20; Appellees’ Br. at 16 n.4. We disagree. “Standing is not dispensed in gross.” DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 353 (2006) (quoting Lewis v. Casey, 518 U.S. 343, 358 n.6 (1996)). Rather, to win summary judgment and obtain injunctive relief, Kentucky and Tennessee each had to demonstrate, with evidence, why it was suffering particularized continuing or imminent injuries in fact, and why that remained the case even after promulgation of the Rule. Lujan, 504 U.S. at 561; see also id. at 560 n.1 (“By particularized, we mean that the injury must affect the plaintiff in a personal and individual way.”); Trump v. New York, 141 S. Ct. 530, 534 (2020) (“A foundational principle of Article III is that an actual controversy must exist not only at the time the complaint is filed, but through all stages of the litigation.” (cleaned up)). Thus, the district court had no authority to issue an injunction protecting a party that failed to demonstrate that its challenge was even justiciable. Instead, a “remedy must . . . be limited to the inadequacy that produced the injury in fact that the plaintiff has established.” DaimerChrysler Corp., 547 U.S. at 353. The case that Kentucky cites to the contrary—Rumsfeld v. Forum for Academic & Institutional Rights, Inc.—is inapposite. See Appellees’ Br. at 16 n.4 (citing 547 U.S. 47, 52 n.2 (2006)). There, determining the standing of each individual plaintiff (and thus the legitimate scope of injunctive relief) was irrelevant; since plaintiffs had all advanced the same non-meritorious claim, the Court needed to find standing only as to a single plaintiff to deem their shared legal theory erroneous. Id. at 70. But while such cases “give courts license to avoid complex questions of standing in cases where the standing of others makes a case justiciable, it does not follow that these cases permit a court that knows that a party is without standing to nonetheless allow that party to participate in the case.” Nat’l Rifle Ass’n of Am., Inc. v. McCraw, 719 F.3d 338, 344 n.3 (5th Cir. 2013). The proper course is, instead, to limit relief only to those parties who established the district court’s jurisdiction to award it. DaimlerChrysler Corp., 547 U.S. at 353. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 21

compliance costs. Amended Complaint ¶12, R. 23. Unlike Kentucky, Tennessee then submitted uncontroverted evidence of those costs. Eley Dec. ¶¶8–10, R. 25-3. As Eley’s declaration explains, the Offset Provision requires Tennessee to expend time and money that it would not expend but for the Offset Provision to ensure that none of the tax cuts it has enacted or will enact could be construed as having been “indirectly offset” by ARPA spending. Id. ¶8. Likewise, it must also expend resources it would not otherwise have to expend “adjusting [its] ‘baseline’ level of tax revenue for inflation each year during the covered period” to determine whether its tax policies may provoke a recoupment action. Id. ¶9. These injuries were not mooted by the advent of the Rule, since Tennessee must still expend resources to maintain compliance with the Offset Provision (and, for that matter, the Rule as well). See id.

And, unlike with the imminent-recoupment and sovereign-authority theories, we have no similar imminence concern about the compliance-costs argument.13 Tennessee already accepted the funds, so it must undertake compliance efforts at present. Given those facts, we conclude that Tennessee satisfied its obligation to show an actual injury traceable to the defendants and likely redressable by a favorable decision. Indeed, compliance costs are a recognized harm for purposes of Article III. See, e.g., Federal Election Comm’n v. Ted Cruz for Senate, 142 S. Ct. 1638, 1646 (2022); Czyzewski v. Jevic Holding Corp., 137 S. Ct. 973, 983 (2017) (“For standing purposes, a loss of even a small amount of money is ordinarily an ‘injury.’”); see also State Nat. Bank of Big Spring v. Lew, 795 F.3d 48, 53 (D.C. Cir. 2015) (Kavanaugh, J.) (“The Rule also offers a safe harbor, but banks such as State National Bank must incur costs to ensure that they are properly complying with the terms of that safe harbor. . . . Under Lujan, the Bank therefore has standing to challenge the constitutionality of the Bureau.”); Grand River Enters. Six Nations, Ltd. v. Boughton, 988 F.3d 114, 121 (2d Cir. 2021) (collecting cases). Tennessee’s expenditure of those resources, as we explain below, is traceable to the Offset Provision. Its proscriptions are why Tennessee must incur such costs—to maintain compliance with the Offset Provision and stave off a recoupment action.14 And permanently enjoining the Offset Provision’s enforcement

13 We thus assess Tennessee’s standing argument under the ordinary Lujan framework. See supra at 12–13. 14 Supreme Court doctrine, we note, provides that a federal court has jurisdiction over a regulated entity’s pre-enforcement challenge even when no enforcement action is imminent if the enforcement action’s remoteness stems from the regulated entity’s own involuntary efforts to comply with the contested proscription. MedImmune, No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 22

would redress that injury. For if enforcement of the Offset Provision were enjoined, Tennessee would have no reason to continue expending resources to maintain compliance with an unenforceable provision.

Still, however, Treasury disputes several aspects of this analysis. Its objections focus on whether these compliance costs are a legitimate injury in fact, and, even assuming they are, whether such an injury is truly traceable to the Offset Provision. After more fully describing those objections below, we explain why none persuades us that we lack jurisdiction.

a. Treasury’s Objection that Tennessee’s Compliance Costs are not an Injury in Fact

Treasury at points seems to dispute that the compliance costs the Offset Provision (and Rule) inflict on Tennessee even constitute an injury in fact. See, e.g., Reply Br. at 4 n.1. Its argument rests on a portion of the Final Rule—which, we note, acknowledges that ARPA imposes an “administrative burden” on the States—but that permits states to use ARPA funds to defray the costs of complying with ARPA’s reporting requirement. Id.; 87 Fed. Reg. at 4,444. The theory seems to be that Tennessee cannot be injured by ARPA-related compliance costs when ARPA funds may themselves be used to offset administrative expenses.

We perceive two central problems with this argument. The first is that even if the Rule permits states to use ARPA funds to defray the costs of complying with the reporting requirement, 87 Fed. Reg. at 4,444, that is simply beside the point as concerns Tennessee’s compliance-costs argument. Tennessee complained of compliance costs distinct from those imposed by the reporting requirement. See Eley Dec. ¶¶8–9, R. 25-3. Indeed, Eley’s declaration draws an explicit distinction between the costs of reporting Tennessee’s uses of ARPA funds, on the one hand, and the costs of tracking whether any such use could be construed as an “indirect offset,” on the other. Id. ¶10. And that distinction makes perfect sense based on the statute’s text. The reporting requirement explains that states must provide a “detailed accounting

Inc. v. Genentech, Inc., 549 U.S. 118, 128–30 (2007). MedImmune concerned a declaratory judgment, of course, but the Declaratory Judgment Act did not expand federal courts’ jurisdiction beyond that which they already could have exercised to award traditional remedies like money damages or an injunction. Skelly Oil Co. v. Phillips Petroleum Co., 339 U.S. 667, 671–72 (1950). So since the court in MedImmune had jurisdiction to adjudicate a declaratory- judgment action concerning alleviation of a prospective harm, it necessarily would have had jurisdiction to entertain an injunctive suit (such as the one at issue here) as well. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 23

of . . . the uses of [ARPA] funds” and “all modifications to the State’s . . . revenue sources during the covered period,” along with “other information as the Secretary may require[.]” 42 U.S.C. § 802(d)(2)(A)–(B). The Offset Provision, by contrast, presents a different obligation: that such “uses” may not be for “indirect” offsets of revenue-reducing tax cuts. § 802(c)(2)(A). So a state could violate the Offset Provision—indirectly offsetting tax cuts with ARPA funds—while complying with the reporting requirement—by simply telling Treasury that it was using the funds in an impermissible manner. Thus, even if we assumed that the States were not injured by the reporting requirement, given that they may defray associated administrative expenses with ARPA funds, that would in no way dispel Tennessee’s distinct injury from additional costs incurred to comply with the Offset Provision.

Second, and more fundamentally, even if we assumed that the Rule permitted Tennessee to use ARPA funds to defray its Offset Provision-related compliance costs, that would still represent an injury in fact. Tennessee has an independent interest in expending its ARPA funds on other legitimate uses; for instance, spending in one of the four areas that Congress deemed necessary to recovery from the pandemic. 42 U.S.C. § 802(c)(1)(A)–(D). ARPA funds expended on compliance with an invalid Offset Provision are necessarily ARPA funds not productively expended on economic recovery. So a “diversion of resources” from useful areas to compliance with an invalid condition would nonetheless constitute an injury in fact. Dep’t of Com. v. New York, 139 S. Ct. 2551, 2565 (2019). Put simply, Tennessee has a live interest in not wasting its ARPA funds on compliance with an invalid condition.

b. Additional Concerns about Traceability

Next, Treasury reframes the same argument as an objection to traceability—that all of Tennessee’s compliance costs are traceable solely to the unchallenged reporting requirement, rather than to the Offset Provision. Reply Br. at 4. But we reject this argument for the same reason we rejected it above; again, that it is undercut by both law and fact. The reporting requirement contains no prohibition on how funds may be used. It establishes only an obligation that states inform Treasury of which uses a state pursues. The Offset Provision, by contrast, contains a substantive prohibition on use—that the funds cannot be used to “indirectly offset” revenue reductions resulting from tax cuts. 42 U.S.C. § 802(c)(2)(A). And Eley’s No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 24

uncontroverted declaration establishes that these distinct obligations incur distinct sets of compliance costs upon Tennessee.

That discussion appears to exhaust Treasury’s arguments about why the compliance-costs injury is nonjusticiable. But given our independent obligation to ensure that we have jurisdiction, we find a few additional comments about traceability in order. See, e.g., Arbaugh, 546 U.S. at 514. One important issue, we note, is whether Tennessee’s compliance costs are truly traceable to the Offset Provision itself, or whether they are traceable merely to the Rule. In our view, Tennessee’s costs are most proximately traceable to the Rule, rather than to the Offset Provision’s text. If the Rule had never existed, after all, the statute alone arguably might have entailed obligations wholly distinct from those described by the Rule, and thus a distinct set of compliance costs as well. Absent the Rule’s safe harbors about growth, tax increases, and spending cuts, for instance, the Offset Provision itself, under the money-is-fungible interpretation, arguably proscribed all revenue-reducing tax cuts during ARPA’s “covered period.” So compliance costs under that regime would have been much simpler: just don’t cut taxes if you want to expend ARPA funds. But the Rule clarified that Tennessee may cut taxes insofar as it establishes a new tracking procedure to ensure that funds offsetting a tax cut stem from a permissible replacement revenue source (e.g., growth) rather than from ARPA funds. Eley Dec. ¶8, R. 25-3. Likewise, Eley’s declaration explained that Tennessee must expend additional resources to inflation-adjust its revenues for each year of the “covered period” and then compare them with a 2019 baseline to determine whether any year during the “covered period” witnessed a “reduction” in net tax revenue. That specific requirement only became clear from the Rule. Id. ¶9 (citing 31 C.F.R. § 35.3, 35.8(b)).

Yet while these refined obligations’ origin in the Rule may highlight the indeterminacies of the underlying Offset Provision, they do not establish that Tennessee’s injuries cannot be traced to the Offset Provision itself. The Supreme Court recently confronted an analogous issue in Federal Election Commission v. Ted Cruz for Senate. 142 S. Ct. at 1638. There, the Cruz campaign challenged a campaign-finance restriction found in an agency regulation implementing a statute but absent from the underlying statute’s text. Id. at 1648. Thus, the government argued, “[a] challenge to the regulation . . . is separate from a challenge to the statute that authorized it.” No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 25

Id. But the Court declined to endorse this distinction. To the contrary, it held that an injury from a regulation implementing a statute was still traceable to the statute itself. Id. at 1649. “An agency, after all, ‘literally has no power to act’—including under its regulations—unless and until Congress authorizes it to do so by statute.” Id. (citations omitted). And “[a]n agency’s regulation cannot ‘operate independently of’ the statute that authorized it.” Id. (citation omitted). So even if Tennessee’s injuries are most proximately traceable to the Rule, we nonetheless conclude that these injuries also suffice for standing to challenge the Offset Provision itself. For if enforcement of the Offset Provision itself were enjoined, it would necessarily preclude enforcement of the Rule, at least to the extent it implements the Offset Provision.

Last, we address the notion that Tennessee’s compliance costs are a “self-inflicted injury” and are thus traceable solely to its own conduct in accepting the ARPA funds, rather than to some wrongful conduct of the federal government.15 It is in some sense true that Tennessee exposed itself to the risk of compliance costs when it accepted the ARPA funds. Of course, Tennessee did so with a reservation about the Offset Provision, and it also insists that it took the funds under duress. See Recording of Oral Arg. at 27:38–27:55; Appellees’ Br. at 29. But even if we assumed that Tennessee took the money purely of its own volition, that would not make its compliance costs “self-inflicted” in a way that would defeat jurisdiction. The Supreme Court recently rejected a similar argument in, incidentally, Ted Cruz for Senate. 142 S. Ct. at 1647–48. There, the Cruz campaign stipulated that its “sole and exclusive motivation” for violating the campaign-finance restriction was to create a factual basis for challenging the restriction. Id. at 1647. The government accordingly argued that any resultant injury was traceable not to the restriction, but to the Cruz campaign’s willful violation of it. Id. Again, however, the Court unequivocally rejected this theory. As it explained, “[w]e have never recognized a rule of this kind under Article III. To the contrary, we have made clear that an injury resulting from the application or threatened application of an unlawful enactment remains fairly traceable to such application, even if the injury could be described in some sense as willingly incurred.” Id.; see

15 The district court raised this concern at the permanent-injunction hearing, apparently sua sponte. See 9/8/2021 Tr. of Hearing at 4:15-23, R. 41. As the States point out, Treasury has declined to press it either below or before us. See Appellees’ Br. at 18 n.5. We nonetheless address it because, again, we have independent obligation to assure ourselves of jurisdiction. See, e.g., Arbaugh, 546 U.S. at 514. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 26

also id. at 1648 (“That appellees chose to subject themselves to those provisions does not change the fact that they are subject to them, and will face genuine legal penalties if they do not comply.”). So even if Tennessee had voluntarily chosen to subject itself to the Offset Provision, it would not defeat (and, indeed, would establish) Tennessee’s standing to challenge it.

Moreover, Tennessee points out, such a jurisdictional bar would be irreconcilable with the Supreme Court’s broader Spending Clause jurisprudence. See Appellees’ Br. at 18 n.5. In any Spending Clause challenge, it could be argued, states that accepted federal funds assumed the risk that an ambiguous condition could be construed against their interests. Id. For instance, states might have recognized that the term “costs” in the Individuals with Disabilities Education Act (“IDEA”) could be construed to include expert-witness fees. Arlington Cent. Sch. Dist. Bd. of Educ., 548 U.S. at 293–94. That it did was, in fact, the position of three dissenting Justices. See id. at 308 (Breyer, J., dissenting, joined by Stevens, J., and Souter, J.). Or, states might have recognized, the term “appropriate relief” could be construed to include money damages—a position that two dissenting Justices called “self-evident.” Sossamon v. Texas, 563 U.S. 277, 293 (2011) (Sotomayor, J., dissenting, joined by Breyer, J.). But in neither of those cases were the states held subject to such obligations, since the question is not whether states could have conceived of those liabilities when accepting the funds; it is instead whether they assumed an obligation about which the relevant statute conferred notice clearly and unambiguously. Arlington Cent. Sch. Dist. Bd. of Educ., 548 U.S. at 298; Sossamon, 563 U.S. at 289–91. Conversely, therefore, jurisdiction is not defeated by (and the merits of the challenge are established by) a spending law’s omission of clear warnings about the obligations it entails.

B. The Merits

Having concluded that Tennessee’s challenge is justiciable, we now explain why we agree with Tennessee that the Offset Provision did not establish, with the requisite clarity, the putative obligations it was revealed to entail by the Rule. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 27

1. Under the Relevant Spending Clause Jurisprudence, the Offset Provision Fails to Provide States Clear Notice of the Conditions It Entails

At the outset, we note that the States have labeled their challenge as against an “unconstitutionally ambiguous” piece of spending legislation. See, e.g., Amended Complaint ¶57, R. 23. Neither of those terms is entirely accurate. First, as a technical matter, the Offset Provision is more than merely “ambiguous.” Ambiguity refers to situations in which language has at least two definite meanings and a court must select between or among them. See Caleb Nelson, Statutory Interpretation 77–80 (2011). (For instance, the word “bank” without further context might refer to either a riverside or a financial institution.) Vagueness, by contrast, arises when a term is open-ended and lacks inherent or definite content. Id. The Offset Provision is better described as suffering from the latter defect. The States could not have known from the statute itself the reticulated way that Treasury’s Rule would construe the Offset Provision, since that construction was hardly obvious ex ante. Nor could they have reliably predicted which of the several potential baselines Treasury would select to measure a “reduction,” nor when Treasury might deem such a reduction to have “resulted” from a tax cut. To the contrary, the statute’s open-endedness gave Treasury expansive discretion to construe its terms in the particular way Treasury saw fit.

And second, the Offset Provision is not “unconstitutional” under the Spending Clause, strictly speaking, just because of those indeterminacies. Rather, the Supreme Court has explained that because Congress can cajole the states to enact policies indirectly (through a spending inducement) that it could never directly order them to perform with its other enumerated powers, we must employ a federalism-based clear-statement rule when construing spending legislation as a matter of statutory interpretation. See, e.g., South Dakota v. Dole, 483 U.S. 203, 207 (1987); Cummings, 142 S. Ct. at 1570, 1574; Sch. Dist. of City of Pontiac v. Sec’y of U.S. Dep’t of Educ., 584 F.3d 253, 283–84 (6th Cir. 2009) (en banc) (Sutton, J., concurring) (describing the clear-statement rule as a “statutory limitation on Congress’s spending power”); see also Haight v. Thompson, 763 F.3d 554, 568 (6th Cir. 2014) (“One of the distinguishing features of the spending power is that it allows Congress to exceed its otherwise limited and enumerated powers by regulating in areas that the vertical structural protections of No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 28

the Constitution would not otherwise permit.”). In other words, Congress does not necessarily lack the constitutional power to enact vague spending laws in the same way that, for instance, it lacks the power to enact a law “respecting an establishment of religion.” U.S. Const. amend. I. But those laws may be unenforceable in certain circumstances when they fail to provide states with clear notice of a purported funding condition.

So, as we explain below, we do not hold the Offset Provision “unconstitutional” under the Spending Clause. Rather, our holding is this. As a matter of statutory interpretation, we conclude that the Offset Provision does not clearly explain (1) how to calculate a “reduction” in net tax revenue, (2) how to determine whether such a reduction resulted from a tax cut, or (3) how to tell what particular conduct constitutes an “indirect” offset. And Treasury’s attempted liquidation of the Offset Provision via the Rule in no way followed clearly from the Offset Provision’s text. Thus, Tennessee may legitimately discontinue the compliance procedures entailed by the Rule, and if, as a result, it should engage in conduct Treasury deems a violation of the Offset Provision, Treasury may not initiate enforcement proceedings in response.

2. Applying the Spending Clause Clear-Statement Rule

It is undisputed that ARPA was enacted pursuant to the Spending Clause. Unlike ordinary coercive legislation, “legislation enacted pursuant to the spending power is much in the nature of a contract: in return for federal funds, the States agree to comply with federally imposed conditions. The legitimacy of Congress’[s] power to legislate under the spending power thus rests on whether the State voluntarily and knowingly accepts the terms of the ‘contract.’” Pennhurst State Sch. & Hosp. v. Halderman, 451 U.S. 1, 17 (1981). And “[s]tates cannot knowingly accept conditions of which they are ‘unaware’ or which they are ‘unable to ascertain.’” Arlington Cent. Sch. Dist. Bd. of Educ., 548 U.S. at 296 (quoting Pennhurst, 451 U.S. at 17)). As a result, Congress must provide “clear notice” of the obligations a spending law entails. Pennhurst, 451 U.S. at 25. “After all, when considering whether to accept federal funds, a prospective recipient would surely wonder not only what rules it must follow, but also what sort of penalties might be on the table.” Cummings, 142 S. Ct. at 1570. And this clear- statement rule applies with particular force where “a State’s potential obligations under the Act are largely indeterminate.” Pennhurst, 451 U.S. at 24. “Accordingly, if Congress intends to No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 29

impose a condition on the grant of federal moneys, it must do so unambiguously” and with a “clear voice.” Id. at 17. Applying these principles reveals that neither the “indirectly offset” language nor the “reduction in the net tax revenue . . . resulting from” language provided the states the requisite “clear notice” of whatever obligations such language entails. Id. To the contrary, these are “largely indeterminate” provisions susceptible to a range of plausible meanings. Id. at 24; cf. Boechler, P.C. v. Comm’r, 142 S. Ct. 1493, 1498 (2022) (“Where multiple plausible interpretations exist—only one of which is jurisdictional—it is difficult to make the case that the jurisdictional reading is clear.” (citation omitted)).

a. “Indirectly Offset”

In assessing justiciability, we spoke at length about the Offset Provision’s indeterminacies; particularly, the prohibition on “indirect” offsets. So we briefly reiterate those points here. The first core issue with the Offset Provision, again, is that its text does not clearly explain what it means to “indirectly offset” revenue-reducing tax cuts with ARPA funds. “Indirectly” means “not directly; obliquely; not straightforwardly, or the like; in an indirect, roundabout, or subtle manner.” Indirectly, Webster’s New International Dictionary 1267 (2d ed. 1960); see also Indirectly, New Practical Standard Dictionary 677 (1956) (“Not in direct relation; not tending to a result by the shortest or plainest course; inferential.”); Indirectly, Compact Edition of the Oxford English Dictionary 1418 (1971) (“By indirect action, means, connexion, agency, or instrumentality; through some intervening person or thing; mediately.”). And “offset,” in the relevant sense, simply means “to counterbalance” or “compensate” for something. Offset, Webster’s New International Dictionary at 1691; see also Offset, New Standard Practical Dictionary at 917 (“Anything regarded or advanced as a counterbalance or equivalent; set-off.”); Offset, Compact Edition of the Oxford English Dictionary at 1981 (“To set off as an equivalent against something else.”). So, as Treasury contends, this statutory language apparently stands for the general proposition that states may not circumvent the use restriction “with formalities.” Appellants’ Supp. Br. at 3. Beyond that general notion, however, what this language actually obliges the States to do is difficult to say.

For instance, the States contend that an “indirect offset” could plausibly occur whenever a state enacts a revenue-reducing tax cut and expends ARPA funds—no matter whether the state No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 30

pours the ARPA funds into the precise area it cut taxes. See Appellees’ Br. at 38–39. This is the money-is-fungible interpretation of the Offset Provision that we described above. See id.; see also Amended Complaint ¶35, R. 23. Nothing about ARPA’s text or context suggests that this interpretation is particularly far-fetched. The macroeconomic assumption underlying the Act seems to be that recovery from a recession is best achieved by high levels of spending, rather than static levels of spending accompanied by cuts in taxation. See, e.g., 86 Fed. Reg. at 26,786– 87 (explaining that the “demand for government services is high,” but that “State, local, and Tribal government austerity measures can hamper overall economic growth, as occurred in the recovery from the Great Recession.”). So, chilling tax cuts to facilitate high levels of spending would seem consistent with ARPA’s purpose. Id. And even Treasury’s own Rule acknowledges that the money-is-fungible interpretation is at least a plausible concern with the Offset Provision.16 Once again, it explained, “because money is fungible, even if [ARPA] funds are not explicitly or directly used to cover the cost of changes that reduce net tax revenue, those funds may be used in a manner inconsistent with the statute by indirectly being used to substitute for the state’s or territory’s funds that would otherwise have been needed to cover the costs of the reduction.” 87 Fed. Reg. at 4,424 (emphasis added).17 The plausibility of this interpretation was the very reason that Treasury had to shed so much ink attempting to disavow it with the Rule.

16 In its reply brief, Treasury invoked the canon of constitutional avoidance to suggest that the money-is- fungible interpretation of the Offset Provision is implausible (though, we note, it raised this point in its objections to justiciability rather than the merits). See Reply Br. at 3 (citing Jennings v. Rodriguez, 138 S. Ct. 830, 842 (2018)). We find this canon of minimal importance to either justiciability or the merits for two key reasons. First, “[f]or standing purposes, we accept as valid the merits of appellees’ legal claims.” Ted Cruz for Senate, 142 S. Ct. at 1647; see also Warth v. Seldin, 422 U.S. 490, 500 (1975) (“[S]tanding in no way depends on the merits of the plaintiff’s contention that particular conduct is illegal.”). So it would be inappropriate for us, at the justiciability stage, to render a merits interpretation of the Offset Provision and to then declare based on that merits interpretation that the controversy is not even justiciable. See, e.g., Trump v. Hawaii, 138 S. Ct. 2392, 2416 (2018). After all, if the laws of the United States when “given one construction” would establish jurisdiction and would defeat it when “given another,” then the plaintiff has established jurisdiction. Bell v. Hood, 327 U.S. 678, 685 (1946). Second, concerning the merits, the constitutional-avoidance canon would at most dispel the States’ money-is-fungible interpretation of the Offset Provision. But establishing whatever obligations the Offset Provision does not impose cannot suffice to defeat a Spending Clause challenge, for the critical question would still remain about whatever obligations the Offset Provision does impose—such as those it is claimed to impose in the Rule—and whether it does so clearly and unambiguously. 17 Or, once again, consider Secretary Yellen’s acknowledgement that “[w]e will have to define what it means to use money from this Act as an ‘offset’ for tax cuts. And, given the fungibility of money, it’s a hard question to answer.” Treasury Secretary & Federal Reserve Chair Testimony on COVID-19 Economic Recovery at 58:30–59:05, available at https://www.c-span.org/video/?510059-1/treasury-secretary-federalreserve-chair- testimony-covid-19-economic-recovery. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 31

See, e.g., Yellen Letter, R. 1-1 (explaining that Treasury would promulgate “further guidance” so that states could understand their obligations under the Offset Provision).

True, the Rule—as distinct from the Offset Provision itself—went on to clarify that such an “indirect offset” would not be deemed to have occurred in three particular situations: where the spending cut is offset with macroeconomic growth, another state tax increase, or a state spending reduction. 87 Fed. Reg. at 4,423; see also Appellants’ Br. at 5. But why is the presence of any safe harbor dictated by the underlying statute, much less clearly so? And why do those particular safe harbors reside in the statutory text, much less clearly so? In reality, the statute is silent on those questions. Precisely because the Offset Provision is so indeterminate about what behavior might constitute an “indirect offset,” Treasury was necessarily left with a huge range of discretion about which state behavior it would deem permissible versus impermissible. As a result, the statute itself failed to provide “clear notice” to Tennessee about whichever particular conduct Treasury would permit or proscribe. Arlington Cent. Sch. Dist. Bd. of Educ., 548 U.S. at 296. And no doubt because of the Offset Provision’s lack of inherent content, Treasury found it necessary to promulgate a Final Rule with a hundred pages of commentary in its attempt to establish some concrete “guidance.” See Yellen Letter, R. 1-2.

b. “A Reduction in the Net Tax Revenue . . . Resulting From” a Tax Cut

The Offset Provision’s language concerning “a reduction in the net tax revenue . . . resulting from” a tax cut is similarly indeterminate. 42 U.S.C. § 802(c)(2)(A). Several major issues prevent it from having provided the States “clear notice” of their “obligations” under the statute. Cummings, 142 S. Ct. at 1574.

The first, as we mentioned before, is that this portion of the statute never actually specifies which fiscal year’s revenue inlays serve as the baseline against which to determine whether a state experienced a “reduction” in its revenues. See Appellees’ Br. at 40. And even Treasury’s own Rule acknowledges how critical this omission was: “Measuring a ‘reduction’ in net tax revenue requires identification of a baseline. In other words, a ‘reduction’ can be assessed only by comparing two amounts.” 87 Fed. Reg. at 4,426. In response to that omission, the Rule happened to set the baseline for the Offset Provision as “the fiscal year ending in 2019.” No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 32

31 C.F.R. § 35.3; 87 Fed. Reg. at 4,423. Yet nothing in the Offset Provision’s text clearly dictates why a 2019 baseline applies. For instance, Treasury might have selected a cascading baseline, in which the Offset Provision was construed to “prohibit[ ] the States from cutting taxes in any given year relative to the year prior.” Appellees’ Br. at 40. Or it might have set the baseline as the fiscal year of ARPA’s enactment. And that might have been an especially obvious baseline, given that we typically assess statutory meaning as of “the time Congress enacted the statute.” Wis. Cent. Ltd. v. United States, 138 S. Ct. 2067, 2070 (2018) (quoting Perrin v. United States, 444 U.S. 37, 42 (1979)).

The point is that the statute itself is “indeterminate” with respect to whatever baseline the offer entailed. Pennhurst, 451 U.S. at 24. And that wasn’t because there was some inherent obstacle to Congress’s specification of a baseline. For instance, compare the Offset Provision with the provision just above it—§ 802(c)(1)(C)—which establishes one of the permissible uses of ARPA funds. Congress there explained that states may spend the funds “for the provision of government services to the extent of the reduction in revenue of such State . . . relative to revenues collected in the most recent full fiscal year of the State . . . prior to the emergency.” 42 U.S.C. § 802(c)(1)(C) (emphasis added). So Congress specified a baseline there. Why not specify a baseline for the Offset Provision itself?18

Second, setting aside the baseline issue, the Offset Provision contains further indeterminacies about how states must assess whether a reduction in tax revenue “result[ed] from a change” in state tax policy. § 802(c)(2)(A) (emphasis added). Put simply, the actual effect of a tax cut may be hard to predict ex ante. See 87 Fed. Reg. at 4,406, 4,423, 4,426; see

18 The simultaneous enumeration of a baseline in § 802(c)(1)(C) and the omission of one in § 802(c)(2)(A)—the Offset Provision—creates further clear-notice issues. Treasury interprets the “most recent full fiscal year . . . prior to the emergency” in § 802(c)(1)(C) as imposing a baseline consistent with revenues collected in the fiscal year ending in 2019. 87 Fed. Reg. at 4,426. It also interprets the Offset Provision to impose the same baseline, despite the Offset Provision having omitted the relevant language from § 802(c)(1)(C). Perhaps states were supposed to extrapolate that the § 802(c)(1)(C) baseline also applied to the Offset Provision. See, e.g., 31 C.F.R. § 35.4. But the typical presumption is that when Congress omits specific language in one provision that it includes in another, the omission implies a difference in meaning between the two provisions. See, e.g., Dean v. United States, 556 U.S. 568, 573 (2009) (“[W]here Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.” (quoting Russello v. United States, 464 U.S. 16, 23 (1983)). So the differing text in §§ 802(c)(2)(A) and 802(c)(1)(C) would complicate the notion that states had clear notice that differently worded provisions imposed the same baseline. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 33

Appellees’ Br. at 41–44. So states considering tax cuts must necessarily generate and rely upon estimates of the real-world effects a tax cut will produce when assessing the cut’s potential impact on their budgets. For instance, a state in one fiscal year might collect $10 million per annum from a particular tax. But the state’s budget analysts might forecast that the state could collect the same amount of tax—$10 million per annum—even if the state reduced the relevant tax rate, as doing so might stimulate the occurrence of additional transactions subject to the tax. So imagine that the state, acting upon that assumption, enacted a tax cut in the relevant area. But during the next fiscal year, that same tax generated only $9 million in inlays. Did that fall in inlays result from the tax cut?

The Offset Provision itself does not supply an answer, because it never specifies whether it prohibits a reduction in expected tax revenues, which a state would be able to control ex ante, or whether it prohibits a reduction in actual tax revenues, which a state could potentially determine only ex post. Yet the difference matters. In the above hypothetical, for instance, the state’s tax cut did not reduce its expected tax revenues, since the best information then available to it suggested that the effect of the tax cut would be revenue-neutral. But the tax cut arguably reduced its actual tax revenues, assuming that the only variable that changed from one year to the next was the tax cut.19

Recognizing that the Offset Provision itself is silent on this issue, the Rule, perhaps surprisingly, suggests that whether a revenue reduction “resulted” from a tax cut hinges on whichever accounting method a state uses to determine the effect of the tax cut. 87 Fed. Reg. at 4,406–07. As it explains, “[i]n assessing whether a tax change has had the effect of reducing tax revenue, recipients may either calculate the actual effect on revenue or rely on estimates prepared at the time the tax change was adopted,” so long as those estimates were “based on

19 In real-world applications, of course, determining causation can be much more complicated. For instance, as Treasury acknowledges, many variables “exogenous” to a tax cut itself affect whether the tax cut, even if preceding a reduction in actual revenues, caused the reduction in actual revenues. 87 Fed. Reg. at 4,406. So isolating whether a tax cut ultimately caused a reduction in actual revenues can be extremely difficult. Id. Treasury attempted to resolve this problem by establishing a causation presumption that, whether sound policy or not, has little apparent relationship to ARPA’s plain text: the IFR “included a presumption that all revenue loss is due to the pandemic,” rather than to a tax cut. Id. The Final Rule then explained that this presumption applied to revenue reductions experienced before January 6, 2022, but did not apply to revenue reductions experienced after January 6, 2022. Id. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 34

reasonable assumptions.” Id. (emphases added). So assuming that a state uses an “actual effect” accounting method in the above example, the revenue reduction arguably resulted from the tax cut. But if a state uses a “reasonable expectations” accounting method, the revenue reduction seemingly did not result from the tax cut, as the state did not reasonably expect an actual revenue reduction ex ante. Yet how were the States supposed to know about these critical points based on the Offset Provision alone?

Or consider this issue: the Offset Provision itself never specifies the timespan during which we assess whether a revenue reduction occurred. For instance, imagine a state recorded $100 million in tax revenues in FY 2019, enacted a tax cut in FY 2020 that stimulated the economy and produced $120 million in tax revenues, but then experienced a downturn in FY 2021 resulting in tax revenues of only $95 million (which arguably might have been higher absent the FY 2020 tax cut). Does that scenario count as a “reduction . . . in net tax revenue” resulting from the FY 2020 tax cut, given that the tax cut arguably resulted in reduced inlays in FY 2021? Or did the state actually experience an increase in net tax revenues, since, combining the inlays from FY 2020 and 2021, the state recorded a net gain of $15 million in inlays versus FY 2019? Whatever the answer, it is at least unclear from the Offset Provision itself whether such a “net reduction” is measured across the entire “covered period” or on a year-to-year basis within the “covered period.” 42 U.S.C. § 802(c)(2)(A).

For those reasons, therefore, neither operative portion of the Offset Provision— “indirectly offset” and “reduction in . . . net tax revenue . . . resulting from” a tax cut—provided Tennessee “clear notice” about the measures required to maintain compliance. Cummings, 142 S. Ct. at 1570.20 Nor—as we explain below—can Treasury’s subsequent promulgation of its Rule cure this vagueness defect.

20 Treasury emphasized in its reply brief that no court decision has ever “declared any . . . funding condition to be ‘unconstitutionally ambiguous’ in the abstract, as a facial matter. Rather, the Supreme Court and other courts, including this one, have relied on the clear-statement principle as a tool of statutory interpretation, to be used when adjudicating concrete disputes over the application of particular funding conditions.” Reply Br. at 5. With today’s opinion, the trend continues. Nowhere have we deemed the Offset Provision “unconstitutional” under the Spending Clause because of its indeterminacies. We instead have conducted our analysis as a matter of statutory interpretation. Likewise, that analysis has not unfolded “in the abstract,” but has focused on two concrete obligations imposed by Treasury’s rule: (1) that Tennessee, should it wish to enact revenue-reducing tax cuts, must trace any dollars arguably used to offset those reductions to three particular, ad hoc safe harbors, and (2) that in No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 35

c. Why Treasury’s Rule Cannot Cure Spending-Law Vagueness

The question whether agency regulations construing spending legislation are entitled to deference has generated some occasional academic interest. See, e.g., Peter J. Smith, Pennhurst, Chevron, and the Spending Power, 110 Yale L.J. 1187, 1189–90 (2001). But we note at the outset that this issue is not in the present case: Treasury categorically waived reliance on the Rule to cure a vagueness defect under the Spending Clause. As it told the district court, “agency regulations should have no bearing on the Spending Clause analysis.” Defs.’ Mot. to Dismiss & Mot. for Summ. J. at 30, R. 32 (emphasis added). It argued instead that the Offset Provision itself satisfied the Spending Clause, since at the very least it put the States on notice that the offer came with “a condition”—no matter whether the contours of that condition presented significant indeterminacies as a matter of the statutory text. Id. at 39.

Treasury has reprised these arguments before us. It does not argue that the Rule—even though it was promulgated before the States accepted the ARPA funds—can provide clear notice to the States of their obligations. Rather, it argues that it was the Offset Provision’s text alone that “clearly place[d] States ‘on notice’ that their acceptance of Fiscal Recovery Funds ‘is conditioned upon compliance with’ the requirement not to use those funds to pay for tax cuts.” Reply Br. at 7; see also Appellants’ Supp. Br. at 2 n.2. So again, Treasury acknowledges that whether Tennessee’s Spending Clause challenge succeeds hinges on whether the Offset Provision itself is impermissibly vague about whichever obligations it imposes on the states.

But we note that even if we were bound to independently assess whether Treasury’s Rule could provide clear notice of conditions left otherwise indeterminate by the statute, we still would hold that it could not do so in these particular circumstances.21 Our primary concern here

assessing whether such a revenue reduction occurred, Tennessee must establish and use an accounting procedure to compare its current inlays to its inflation-adjusted revenue in “its fiscal year ending in 2019.” 31 C.F.R. § 35.3. Those are the relevant obligations, we conclude, of which Tennessee did not have clear notice from the Offset Provision itself. 21 We confront the Rule’s effect even despite Treasury’s waiver because we recognize that there are serious and unresolved disputes about, for instance, whether the government may validly waive Chevron deference. Compare Guedes v. BATFE, 920 F.3d 1, 23 (D.C. Cir. 2019) (concluding that Chevron deference cannot be waived if the “underlying agency action” would otherwise merit Chevron deference), with Guedes v. BATFE, 140 S. Ct. 789, 790 (2020) (Gorsuch, J., concurring in denial of certiorari) (arguing that the D.C. Circuit’s conclusion was No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 36

is the legitimate domain of Chevron deference—whether we (or a state) must accept as binding an agency regulation establishing an otherwise-uncertain spending-law condition. For instance, Treasury suggested before us that deference might be appropriate at least if we understood the relevant content of the Rule—the meaning of an “indirect” offset, the baseline for a revenue reduction, and how to tell whether such a reduction “resulted from” a tax cut—to constitute mere “implementation details.” Reply Br. at 7–8. And, to that end, it invoked a couple of circuit decisions where we indeed deferred to agencies’ reasonable views about marginal ambiguities in spending laws: one concerning whether the term “medical devices” includes “incontinence products” and the other concerning whether “records maintained by a law enforcement unit of [an] education agency or institution that were created by that law enforcement unit for the purpose of law enforcement” includes student disciplinary records involving “serious criminal conduct.” See id. (citing Harris v. Olszewski, 442 F.3d 456, 467–68 (6th Cir. 2006), and then citing United States v. Miami Univ., 294 F.3d 797, 814–15 (6th Cir. 2002)).

Yet we find that whether deference was warranted on such arcane topics as those has little relevance to the Offset Provision. It is difficult to see how the Rule represents mere “implementation details” when it supplied content without which the Offset Provision literally could not function. And, in any event, Treasury is wrong to suggest that we should act “as if we were interpreting a statute which has no implications for the balance of power between the Federal Government and the States.” Va. Dep’t of Educ. v. Riley, 106 F.3d 559, 567 (4th Cir. 1997) (en banc).22 Unlike the distribution of incontinence products or the release of disciplinary records, control over taxation is a core aspect of state sovereignty. See Dep’t of Revenue of Or., 510 U.S. at 345; Lane County, 74 U.S. at 76. For Congress to impose conditions in that area, it must do so in clear and unmistakable terms. See, e.g., SWANCC v. U.S. Army Corps of Eng’rs, 531 U.S. 159, 172–73 (2001) (explaining that the Court “would not extend Chevron deference” to an agency interpretation involving “federal encroachment upon a traditional state power.”); see also Ala. Ass’n of Realtors v. Dep’t of Health & Hum. Servs., 141 S. Ct. 2485, 2489 (2021).

erroneous because the “[Supreme] Court has often declined to apply Chevron deference when the government fails to invoke it”). 22 We quote from Judge Luttig’s dissenting panel opinion, the relevant portion of which a majority of the full Fourth Circuit adopted upon en banc rehearing. See Va. Dep’t of Educ., 106 F.3d at 561. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 37

When such a clear-statement rule is in play, it is insufficient merely that an agency reasonably liquidated ambiguities in the relevant statute. Id.; see also Va. Dep’t of Educ., 106 F.3d at 567 (declining to apply Chevron deference to an ambiguous spending statute because “[i]t is axiomatic that statutory ambiguity defeats altogether a claim by the Federal Government that Congress has unambiguously conditioned the States’ receipt of federal monies in the manner asserted”); Carter v. Welles-Bowen Realty, Inc., 736 F.3d 722, 731 (6th Cir. 2013) (Sutton, J., concurring) (“All manner of presumptions, substantive canons and clear-statement rules take precedence over conflicting agency views.”). Rather, in such circumstances, Congress itself must have spoken with a “clear voice.” Pennhurst, 451 U.S. at 17.23

3. Treasury’s Counterarguments

Before we close, we address a couple of Treasury’s counterarguments to our conclusions above. We first discuss the import of our decision in Cutter v. Wilkinson, which affirmed the Religious Land Use and Institutionalized Persons Act (“RLUIPA”) against a Spending Clause challenge. 423 F.3d 579, 585–86 (6th Cir. 2005). We then address Treasury’s claim that Tennessee should have enjoyed clear notice of the Offset Provision’s meaning because the phrase “directly or indirectly” “appears more than a thousand times in the U.S. Code.” Appellants’ Supp. Br. at 3.

a. Cutter v. Wilkinson and the RLUIPA Comparison

Resisting our merits analysis, Treasury asserts that our decision today conflicts with circuit precedent sustaining RLUIPA against a Spending Clause challenge. Cutter, 423 F.3d at 585–86. We agree that RLUIPA is a helpful comparison—just not in the way Treasury thinks.

We begin with some background about RLUIPA itself. From 1963 to 1990, the Supreme Court interpreted the Free Exercise Clause to require state officials to justify even incidental

23 Conversely, though, we note that this analysis has no effect on our earlier mootness determination concerning Kentucky. Even if the Rule cannot fill in missing Spending Clause conditions, it can at least still bind Treasury in how it will administer the statute. And neither Tennessee nor Kentucky sought vacatur of the Rule under 5 U.S.C. § 706. Vacatur at least possibly could have revived the specter of Treasury enforcing the Offset Provision consistent with the money-is-fungible interpretation. But even after vacatur, that possibility would still seem rather remote, given Treasury’s insistence that the text of the Offset Provision alone precludes the money-is- fungible reading. See, e.g., Appellants’ Supp. Br. at 2 n.2. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 38

burdens on religious free exercise under strict scrutiny. See, e.g., Sherbert v. Verner, 374 U.S. 398, 403 (1963). A plaintiff could make out a prima facie case of a constitutional violation if she could establish that she had a sincere religious belief upon which the government had imposed a substantial burden, even if the burden were merely incidental. Id. And if that showing were made, the state then had to prove that its interest in imposing the burden was “compelling,” id., and that it had employed the means least restrictive on religious exercise in achieving its compelling interest. Thomas v. Rev. Bd. of Ind. Emp. Sec. Div., 450 U.S. 707, 718 (1981). During the quarter-century in which this framework prevailed, the Supreme Court produced a sizeable corpus of decisions describing its particular contours. See, e.g., Sherbert, 374 U.S. at 398; Gillette v. United States, 401 U.S. 437 (1971); Wisconsin v. Yoder, 406 U.S. 205 (1972); Thomas, 450 U.S. at 707; United States v. Lee, 455 U.S. 252 (1982); Goldman v. Weinberger, 475 U.S. 503 (1986); Bowen v. Roy, 476 U.S. 693 (1986); Hobbie v. Unemployment Appeals Comm’n of Fla., 480 U.S. 136 (1987); Lyng v. Nw. Indian Cemetery Protective Ass’n, 485 U.S. 439 (1988); Frazee v. Ill. Dep’t of Emp. Sec., 489 U.S. 829 (1989). And the lower courts applied it to hundreds of concrete disputes. See generally James E. Ryan, Smith and the Religious Freedom Restoration Act: An Iconoclastic Assessment, 78 Va. L. Rev. 1407 (1992) (cataloging lower-court applications).

In 1990, however, the Supreme Court functionally overruled this body of precedent, holding that incidental burdens on religious practice merited only rational-basis review. Emp. Div., Dep’t of Hum. Res. of Or. v. Smith, 494 U.S. 872, 878–79 (1990). But Congress, incensed by the Smith decision, twice attempted to overrule it. See 42 U.S.C. § 2000bb et seq. (Religious Freedom Restoration Act); 42 U.S.C. § 2000cc et seq. (RLUIPA). Its second attempt, RLUIPA, restored strict-scrutiny analysis in the land-use and prison-administration contexts. § 42 U.S.C. §§ 2000cc, 2000cc–1. Congress partially rooted its power to enact such a statute in the Spending Clause, 42 U.S.C. § 2000cc-1(b)(1)–(2), and so it conditioned the receipt of certain federal funds on compliance with the old strict-scrutiny framework. See, e.g., § 2000cc-1(a)(1)–(2). Ohio (and various other states) argued that strict scrutiny was too indeterminate to form an enforceable Spending Clause condition. See Gerhardt v. Lazaroff, 221 F. Supp. 2d 827, 841, 844 (S.D. Ohio 2002). But the district court rejected the argument, holding that strict scrutiny was such a well- established framework before RLUIPA that even if it might present marginal indeterminacies in No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 39

certain applications, the states could easily discern what core obligations the statute entailed in the mine-run of cases. See, e.g., id. at 844 (“Courts have been enforcing that exact standard against state action for years.”).

Whether we ever actually adjudicated the correctness of that holding, we note, is uncertain. It appears that Ohio’s officials pressed a different ambiguity argument on appeal to this circuit: that RLUIPA did not clearly specify that it applied to existing federally funded programs, rather than merely to programs established after its enactment. See Cutter, 423 F.3d at 585–86; see also Gerhardt, 221 F. Supp. 2d at 841 (distinguishing between the retroactivity argument and the strict-scrutiny-is-too-indeterminate argument). But we held that RLUIPA imposed that obligation clearly, since its text “explains that the Act applies to ‘any program or activity that receives Federal financial assistance.” Cutter, 423 F.3d at 586 (citing 42 U.S.C. § 2000cc-1(b)(1)). And, given that statutory language, it is difficult to see how a different result could have ensued.

Before us, however, Treasury appears to have erroneously exaggerated Cutter’s precedential effect by claiming that our circuit also rejected the strict-scrutiny-is-too- indeterminate argument. In particular, its reply brief claims that Ohio argued “RLUIPA’s ‘least restrictive means standard’ constituted an ambiguous condition’ that was impermissible under Pennhurst. Cutter, 423 F.3d at 586 (quotation marks omitted). But this Court disagreed, explaining that ‘Congress need not ‘delineate every instance in which a State may or may not comply with the least restrictive means test.’” Reply Br. at 6–7.24 In reality, the portions of Cutter that Treasury quotes were describing a Seventh Circuit decision that had rejected “a similar Pennhurst-based challenge to RLUIPA”—the strict-scrutiny-is-too-indeterminate argument, not the anti-retroactivity argument before the panel in Cutter. Cutter, 423 F.3d at 586 (citing Charles v. Verhagen, 348 F.3d 601, 608 (7th Cir. 2003)).

24 Treasury doubled down on this representation at oral argument, stating, “[o]n the question of how much has to be spelled out in the statute itself, Cutter involved a challenge to RLUIPA. RLUIPA said if you restrict the religious exercise of people in covered institutions, you have to satisfy strict scrutiny. It’s of course highly unclear how strict scrutiny is going to apply to a particular kind of policy. But the court said, and every other court to consider the issue said, you don’t have to have spelled that out in the statute.” Recording of Oral Arg. at 16:46– 17:09. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 40

Set aside Treasury’s misreading of Cutter, however, and pretend that Cutter had actually adjudicated the strict-scrutiny-is-too-indeterminate argument, so that we were bound today by a holding that RLUIPA’s least-restrictive-means test satisfied the Spending Clause. What is that supposed to tell us about the Offset Provision? There is no comparable quarter-century of history in which the Supreme Court decided dozens of cases, and the lower courts decided hundreds of cases, construing what it means to “directly or indirectly offset a reduction in the net tax revenue . . . resulting from” a tax cut. 42 U.S.C. § 802(c)(2)(A). Indeed, Treasury conceded at oral argument that it is aware of no example in which the phrase “directly or indirectly offset” has ever even been used in a Spending Clause statute, much less been given an authoritative construction by the Supreme Court in the context of tax cuts. Recording of Oral Arg. at 43:00– 43:50. So Treasury’s invocation of RLUIPA, it turns out, underscores a reason the Offset Provision is impermissibly vague: given its terms’ apparent novelty, there is, unlike in the context of religious free exercise, no expansive and authoritative corpus of federal-court precedents which the states might have consulted in attempting to discern the nature of their obligations.

b. “Directly or Indirectly” in the U.S. Code

Perhaps implicitly recognizing the dearth of relevant caselaw, Treasury last suggests that Tennessee should have been able to ascertain its obligations under the Offset Provision because various other federal statutes employ the phrase “directly or indirectly.” Appellants’ Supp. Br. at 3. It “is commonly used simply to underscore that a restriction cannot be circumvented through formalities,” Treasury says, and “appears more than a thousand times in the U.S. Code.” Id. That is perhaps true, but this factoid seems of no consequence to us for at least three reasons.

First, as we noted above, Treasury conceded that it has no example of such a phrase in a Spending Clause statute, much less one in the particular context of taxation, and, less still, one that survived ambiguity challenges in federal court. Recording of Oral Arg. at 43:00–43:50.

Second, most of these other uses appear to have no conceivable relevance to the Offset Provision. And, for that matter, they may make the vagueness issues even worse. For instance, consider 22 U.S.C. § 9214(a)(3), which provides that Treasury may freeze the assets of anyone No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 41

who “knowingly, directly or indirectly, imports, exports, or reexports luxury goods into North Korea.” The phrase in that context seemingly bars the use of third-party intermediaries to circumvent a trade restriction. So Company A, wishing to execute proscribed shipments to North Korea, still violates the statute by shipping the goods to Company B in Shanghai for reexport into Pyongyang. Or take 29 U.S.C. § 432(a)(2), which requires officers of labor organizations to report on stock they hold “directly or indirectly” in the business they seek to unionize. Once again, the statute seems to bar circumvention through the use of a third-party intermediary, such as holding the stock indirectly in an index fund. Or last, consider 15 U.S.C. § 7410(b), which prohibits the award of a “grant or fellowship . . . directly or indirectly, to any alien from a country that is a state sponsor of international terrorism.” Again, and for the third time, the phrase seems to bar the use of a third-party intermediary to circumvent the restriction: the grant cannot be distributed to an “institution of higher education” for redistribution to the alien. § 7410(c). So at best, these other uses have nothing to do with the Offset Provision, and, at worst, might have misleadingly suggested that it imposed some particularized bar on the use of third-party intermediaries to launder ARPA funds.

Last, as the above examples suggest, the relevant phrase is not just “directly or indirectly,” but “directly or indirectly offset.” 42 U.S.C. § 802(c)(2)(A). The issue here is not establishing that the Offset Provision bars “circumvent[ion] through formalities” in some broad, general sense, but in determining whatever conduct the Offset Provision might treat as having “directly or indirectly offset” a tax cut. For only then could the states have “ascertain[ed]” their obligations under the Act. Arlington Cent. Sch. Dist. Bd. of Educ., 548 U.S. at 296 (quoting Pennhurst, 451 U.S. at 17)). But that phrase, as far as we can tell from our research, occurs exactly once in the entire U.S. Code—in the Offset Provision.

IV.

In closing, we reiterate the central conclusions we have reached today. Treasury’s credible disavowal of the money-is-fungible interpretation mooted Kentucky’s challenge to the Offset Provision, and so the district court erred when it enjoined Treasury from enforcing the Offset Provision against Kentucky. We thus REVERSE the district court’s justiciability holding as to Kentucky and VACATE the permanent injunction to the extent it bars enforcement of the No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 42

Offset Provision against Kentucky. By contrast, we AFFIRM the district court’s injunction as to Tennessee. We do so because “[c]larity is demanded whenever Congress legislates through the spending power[.]” Haight, 763 F.3d at 568. Yet clarity is just what the Offset Provision lacks. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 43

______________________________________________________

CONCURRING IN PART AND DISSENTING IN PART ______________________________________________________

NALBANDIAN, Circuit Judge, concurring in part and dissenting in part. I concur with nearly all of the majority’s well-reasoned opinion. Importantly, I agree with the majority that the vagueness of the American Rescue Plan Act of 2021 (“ARPA”) violates the Spending Clause. And I agree that Tennessee has standing for the reasons that the majority gives. My only disagreement is about whether Kentucky can press its claim. In short, I believe that both Tennessee and Kentucky (“States”) have standing for reasons related to the federal government’s intrusion on their sovereign-taxing authority. And I don’t believe that the Department of Treasury’s (“Treasury”) Rules on ARPA’s enforcement (“Rules”) affect justiciability.1 So I concur with respect to Tennessee’s participation in the case, but respectfully dissent over Kentucky’s.

Standing arises here in three possible ways: through what the majority calls the “imminent-recoupment,” “compliance-cost,” and “sovereign-authority” theories. My concern is

1 My analysis assumes that the majority is correct that the one-party rule doesn’t apply. (See Majority Opinion, at 20 n.12.) But I’m not sure that’s the case. The rule allows courts to review claims so long as one plaintiff has standing. Massachusetts v. EPA, 549 U.S. 497, 505 (2007). And courts have applied the rule to other Article III requirements like mootness. Nat’l Rifle Ass’n of Am. v. Magaw, 132 F.3d 272, 278 n.4 (6th Cir. 1997); see Nat’l Rifle Ass’n of Am., Inc. v. McCraw, 719 F.3d 338, 344 n.3 (5th Cir. 2013). But this doesn’t mean that a party can obtain relief to which it is not entitled. So we can eventually address standing when a plaintiff would obtain “attorney’s fees” or other “relief different from that sought by plaintiffs whose standing has not been questioned.” Gen. Bldg. Contractors Ass’n v. Pennsylvania, 458 U.S. 375, 402 n.22 (1982); see 13B Charles A. Wright & Arthur R. Miller, Federal Practice & Procedure, § 3531.15, at *4 (3d ed. 2022). But a different situation arises when each plaintiff would obtain the same relief regardless. For cases involving injunctive or declaratory remedies, the practical effects of granting relief may apply to each plaintiff even if we dismiss one for lack of standing. For example, in Bowsher v. Synar, the Supreme Court applied the one-party rule to avoid analyzing other plaintiffs’ standing. 478 U.S. 714, 721 (1986). Without returning to the standing questions, the Court affirmed relief that declared a statute unconstitutional. See id. at 736. Because the standing determinations wouldn’t affect how the Court distributed relief, the Court didn’t need to revisit its use of the rule. See id.; accord McCraw, 719 F.3d at 344 n.3 (recognizing that courts “do not need to verify the independent standing of the other co-plaintiffs” when one party with standing “rais[es] the same claims and issues” (quotation omitted)). Here, Tennessee meets Article III’s requirements, and the relief granted to Tennessee applies to Kentucky regardless: Treasury cannot enforce ARPA’s unconstitutional conditions. So we did not need to resolve Kentucky’s mootness. That aside, I analyze why both States meet Article III’s requirements. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 44

with the last theory, under which I believe both States have standing. To establish standing under any theory, of course, the States must assert an injury that is (1) actual or imminent and concrete and particularized, (2) traceable to Treasury, and (3) likely to be redressed by a favorable decision. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61 (1992). Looking at the sovereign-authority theory, I believe that the only question here is whether the States assert an imminent injury under the first inquiry. And I find that both Kentucky and Tennessee meet this requirement.

With respect to “injury,” courts have recognized for over a century that states “are not normal litigants for the purposes of invoking federal jurisdiction.” Massachusetts v. EPA, 549 U.S. 497, 518 (2007). This regard stems from each state’s “well-founded desire to preserve its sovereign territory.” Id. at 519. For that reason, we don’t treat states as “mere provinces or political corporations.” Alden v. Maine, 527 U.S. 706, 715 (1999). Instead, we recognize their “residuary and inviolable sovereignty.” The Federalist No. 39, at 245 (James Madison) (Clinton Rossiter ed., 1961). And we give “special” recognition to a case when a state sues the federal government. Saginaw County v. STAT Emergency Med. Servs., Inc., 946 F.3d 951, 957 (6th Cir. 2020).

Although states hold a unique status in federal court, they cannot avoid “the constitutional baseline” of Article III. Id. States must still prove a cognizable case or controversy. Arizona v. Biden, 40 F.4th 375, 385 (6th Cir. 2022) (“Article III’s foundational standing requirements remain for private and public litigants alike.”).

Still, states have “special solicitude” when they incur “quasi-sovereign” injuries. Id. They cannot “bypass proof of injury in particular or Article III in general,” but they may incur injuries that private parties cannot. Id. at 385–86. Among other things, states can allege sovereign-related injuries like federal regulation over local-lawmaking authorities, threatened intrusions on state territory, or public nuisances, in which a state seeks “to safeguard its domain and its health, comfort and welfare.” Id at 386. (quoting Kentucky v. Biden, 23 F.4th 585, 596 (6th Cir. 2022)); see Massachusetts, 549 U.S. at 517, 521–23 (recognizing the sovereign and quasi-sovereign interests in protecting coastal lands from rising sea levels). No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 45

And the list doesn’t end there. This Court has acknowledged other ways that sovereign or quasi-sovereign interests can support state standing. States can “plausibly allege[] that the federal government has intruded upon an area traditionally left to the states.” Kentucky, 23 F.4th at 599. They can allege that federal enforcement threatens current or future state policies. See id. And they can allege federal threats to their economies. See id. at 599–601 (“[States] . . . have a quasi-sovereign interest in defending their economies from the alleged negative ramifications of [federal law].”). Indeed, as federal regulation has increased over the states, the list of sovereign injuries has grown.

Finally, a quick note about the “imminence” part of the injury-in-fact inquiry. The majority analyzes imminence using the “pre-enforcement” test from Susan B. Anthony List v. Driehaus, 573 U.S. 149 (2014).2 And that is the way that we typically assess imminence when a case concerns a pre-enforcement challenge. But because the States here allege sovereign and quasi-sovereign injuries, we can assess imminence under a slightly different analysis.

A state can establish an imminent injury by showing a “risk of harm” to their sovereign or quasi-sovereign interests. Massachusetts, 549 U.S. at 521 (quoting Lujan, 504 U.S. at 560). In Massachusetts v. EPA, EPA’s refusal to regulate greenhouse gas emissions presented an imminent injury to state interests related to climate change. Id. The “risk of harm” of sea levels rising and damaging state-coastal property created the imminent injury. Id. at 521–23; see Saginaw County, 946 F.3d at 957 (recognizing that the imminence in Massachusetts came from this “risk”). And the Supreme Court reasoned that a state that has a procedural right to protect its sovereign interests can satisfy Article III’s requirements “without meeting all the normal standards for redressability and immediacy.” Massachusetts, 549 U.S. at 517–18.

We have analyzed imminent harms to sovereign interests in pre-enforcement challenges too. In our recent decision in Kentucky v. Biden, which concerned a pre-enforcement challenge to COVID-19 mandates, we found standing under the sovereign-authority theory. See 23 F.4th

2 Under the Susan B. Anthony test, the States needed to show (1) injuries in the original complaint that establish an intention to engage in conduct arguably affected with a constitutional interest, (2) that ARPA arguably proscribes this conduct, and (3) that if the States should pursue such conduct, a credible threat of recoupment action exists. See 573 U.S. at 161–64. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 46

at 598–601. We held that the sovereign injuries alleged met Article III’s imminence requirements because the states there “show[ed the] negative effects” of federal policy on their sovereign interests. Id. at 602. And sister circuits applying the sovereign-authority theory have also analyzed imminence similarly.3 Here, the States allege the same risk of harm to their sovereign interests.

Applying the sovereign-authority framework, the States have standing in this case. See Massachusetts, 549 U.S. at 520. They allege that ARPA “unconstitutionally intrud[es] on their sovereign authority, by interfering with the orderly management of their fiscal affairs, and by requiring them to forgo their constitutional taxing powers or face an action to return much- needed federal funds after they have already been spent.” (R. 1, Original Complaint, PageID 5 ¶ 12.) Each threat poses an imminent “risk of harm to” their sovereign interests. Massachusetts, 549 U.S. at 521. As described below, they have alleged with particularity ARPA’s “negative effects” on their taxing powers, citizens, and economy. Kentucky, 23 F.4th at 602.

This Court’s growing list of sovereign and quasi-sovereign injuries reinforces the States’ standing in three ways. First, they have “plausibly alleged that the federal government has intruded upon an area traditionally left to the states”—state taxes on state citizens. Kentucky, 23 F.4th at 599; see (R. 1, Original Complaint, PageID 2–3, 5–6 ¶¶ 1–3, 12 (discussing the federalism implications of ARPA); R. 1, Original Complaint, PageID 15–16 ¶ 40 (“[T]he power to tax and spend is a sovereign function that lies at the core of State power.”)); see generally Dep’t of Revenue of Or. v. ACF Indus., Inc., 510 U.S. 332, 345 (1994) (describing the tax power as “central to state sovereignty”).

3 See, e.g., Texas v. Biden, 20 F.4th 928, 970 (5th Cir. 2021) (reasoning that a state’s “special solicitude in the standing inquiry . . . means imminence and redressability are easier to establish [] than usual”), rev’d and remanded on other grounds, 142 S. Ct. 2528 (2022); Texas v. United States, 809 F.3d 134, 154–55 (5th Cir. 2015) (concluding that states had sovereign-authority standing in a pre-enforcement challenge to federal immigration law without using the pre-enforcement framework); Sturgeon v. Masica, 768 F.3d 1066, 1073–74 (9th Cir. 2014) (reasoning that a state failed to prove an “actual or imminent” injury by not identifying “any actual conflict” between federal and state law that showed how a federal requirement would “interfere[] with the state’s control over and management of” state affairs), vacated and remanded sub nom. on other grounds, Sturgeon v. Frost, 577 U.S. 424 (2016); Se. Fed. Power Customers, Inc. v. Geren, 514 F.3d 1316, 1322–23 (D.C. Cir. 2008) (concluding that states had an injury-in-fact because they “credibly claim[ed] to fear” that proposed changes to water-storage uses would result in diminished water flows). No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 47

Second, after discussing their past tax cuts, they allege that ARPA’s vague restrictions “chill[] legislative action” to enact similar policies. (R. 1, Original Complaint, PageID 17 ¶ 43; see id. at PageID 16–17 ¶¶ 41–42.) Financial hardship in part led the States to cut taxes for homeowners and businesses alike. (Id.) Kentucky, for example, enacted a bill intended to “invest in and revitalize a predominantly minority community in Kentucky’s largest city.” (Id. at PageID 16 ¶ 41 (citing 2021 Ky. H.B. No. 321 (NS)).) The tax cuts there fall within a “core part of its sovereign duty” and will lead to a “decrease in net revenue.” (Id.) Likewise, Tennessee seeks to cut its “professional privilege tax” to “attract new businesses and residents, continuing Tennessee’s proven record in promoting economic growth that benefits the entire [s]tate.” (Id. at PageID 16 ¶ 42 (citing H.B. 0987, S.B. 0184, 112th General Assembly (2021)).) Tennessee also expects the bill to reduce state revenue. (Id.)

All this to say, ARPA’s vague conditions chill the States from enacting similar tax cuts. The States allege they cannot confidently enact tax policy because they fear imminent recoupment action for exercising their sovereign-taxing authority. (See id. at PageID 17 ¶¶ 43– 44.) And because of the “fungible” quality of money, the States construe ARPA to “potentially affect[] all State legislative and executive actions that reduce net tax revenues,” even if they have nothing to do with ARPA’s COVID-19 relief. (Id. at PageID 16, 17 ¶¶ 41, 44 (quotation omitted).) ARPA then “likely implicates” the powers to “make and enforce policies and regulations” and the “traditional prerogative to superintend” local taxes on state citizens. Kentucky, 23 F.4th at 599.

To top it all off, the States have standing because ARPA’s chilling effect “threatens to damage each of the [S]tates’ economies.” Id. at 599. The chilling effect to enact tax policies that “invest in and revitalize” communities, (R. 1, Original Complaint, PageID 16 ¶ 41), or “attract new businesses and residents” to promote “economic growth,” (id. at PageID 16 ¶ 42), stagnates the States’ fiscal health and interest in helping their citizens. Kentucky and Tennessee plausibly allege that resistance to ARPA will result in fewer tax cuts for homeowners and businesses, “all to the detriment of their state economies.” Kentucky, 23 F.4th at 601. Because the States have sovereign and quasi-sovereign interests in “defending their economies from the alleged negative ramifications” of ARPA, they also “have standing to contest it.” Id. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 48

The States’ “sovereign prerogatives are now lodged in the Federal Government, and Congress has ordered” Treasury to enforce the Offset Provision and initiate recoupment actions to recover any misused funds. Massachusetts, 549 U.S. at 519; see 42 U.S.C. § 802(e). That the States face threats from exercising their traditional-taxing authority only reinforces that they have a sufficient stake in the case so as “to warrant the exercise of federal judicial power.” Massachusetts, 549 U.S. at 519. So along with giving this case “special” recognition because the States are suing the federal government, Saginaw County, 946 F.3d at 957, we should afford the States “special solicitude in our standing analysis.” Massachusetts, 549 U.S. at 520.

Thus, I agree with the majority that the States allege imminent injuries that grant them standing.4 Recognizing that the States suffer from sovereign injuries raises this question: Did Treasury’s Rules moot the States’ legal interests in obtaining relief?

This is where I part ways with the majority; I believe the answer is no. Treasury’s Rules do not moot Kentucky or Tennessee’s controversy over their sovereign interests. Even with the guidance, the States still suffer from the same harm explained above. And both meet Article III’s requirements as a result.

A case only becomes moot “when the issues presented are no longer live or parties lack a legally cognizable interest in the outcome.” Thomas v. City of Memphis, 996 F.3d 318, 323 (6th Cir. 2021). Said differently, parties do not have a mootness problem if “the relief sought would, if granted, make a difference to the legal interests of the parties.” Ford v. Wilder, 469 F.3d 500, 504 (6th Cir. 2006) (internal quotation omitted). Even with the Rules, the States still need injunctive and declaratory relief to avoid Treasury’s enforcement of ARPA’s unconstitutionally vague conditions.

4 One oddity about this case, it’s not crystal clear to me what the States’ cause of action is here—ARPA, for example, doesn’t appear to recognize their right to sue the federal government. And this Court has held that a cause of action must exist apart from standing, even in a sovereign-authority case. Kentucky, 23 F.4th at 602. It appears to me, however, that the Supreme Court has recognized an action in cases that allege Spending Clause violations under the Constitution itself. See Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 540, 575 (2012); New York v. United States, 505 U.S. 144, 154, 172 (1992); South Dakota v. Dole, 483 U.S. 203, 205, 210–12 (1987). And, tellingly, the government here does not contest the States’ ability to sue apart from their justiciability arguments. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 49

The States’ claims withstand mootness because their issues are still “live.” Thomas, 996 F.3d at 323. As they originally alleged, ARPA still fails to provide clear notice of a funding condition. (R. 1, Original Complaint, PageID 17–19 ¶¶ 46–54.) And even later, the States alleged that Treasury “cannot cure the inherent ambiguity in [ARPA] because Congress must provide the recipients of federal funds with clear notice of any conditions on the use of the funds.”5 (R. 23, Amended Complaint, PageID 148 ¶ 51.) (emphasis in original). With that in mind, they allege the Rules “cannot save” ARPA from its unconstitutional conditions. (Id. at PageID 150 ¶ 61.) This all rings true.

Although the Rules tried to fill in the blanks of ARPA, Treasury cannot resolve ARPA’s open-endedness. The Rules tried to construct ARPA’s guidelines by providing what ARPA didn’t: a revenue baseline and guidance on when Treasury would enforce the Offset Provision. See 31 C.F.R. §§ 35.3; 35.8(b); see also Coronavirus State and Local Fiscal Recovery Funds, 87 Fed. Reg. 4338, 4423, 4428 (Jan. 27, 2022) (to be codified at 31 C.F.R. § 35). But Treasury couldn’t provide the clear notice of the conditions ARPA entailed; Congress, through ARPA alone, had that responsibility. And nothing in ARPA alludes to the Rules’ selected baseline or construction of the Offset Provision. ARPA’s text does not clearly determine why a 2019 baseline applies or why another baseline (like a different year) does not. Nor does ARPA address whether it prohibits a reduction in expected tax revenues or whether it prohibits a reduction in actual tax revenues. Only the Rules purport to provide the answer (that states can choose either), but that answer does not stem from ARPA’s text.

And, perhaps most importantly, the regulation does not fix the Offset Provision in three other ways: Vagueness still exists from ARPA’s lack of explanation on how to (1) calculate a “reduction” in net tax revenue, (2) determine whether such a reduction resulted from a tax cut, and (3) tell what particular conduct constitutes an “indirect” offset.

5 Indeed, one scholar contends that only federal statutes―not federal agency conditions―can “defeat state law.” Philip Hamburger, Purchasing Submission: Conditions, Power, and Freedom 130 (2021). As traditionally recognized, statutes “enjoy the obligation of law” because “a legislative body representative of the people” adopts them. Id. Hamburger notes that this logic does not legitimize a federal agency’s elaboration of a statute. Id. at 131. Agencies, of course, do not represent the people in the same way Congress does, so their rules may not render state law void. Id. No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 50

Indeed, Kentucky and Tennessee allege that the Rules made their concerns over their tax authorities even worse. (See R. 23, Amended Complaint, at PageID 147 ¶ 50.) If ARPA is impermissibly vague, as the States allege and the majority concludes, then Kentucky and Tennessee still face “an unlawfully-imposed quandary in determining how to exercise its sovereign taxing power.” Ohio v. Yellen, 547 F. Supp. 3d 713, 725 (S.D. Ohio 2021). The States’ legislators considering tax changes may delay, second guess, or abandon parts of tax policies because ARPA does not explain the impact that such changes will have on their ability to retain ARPA funds. See id.

And even if the Rules could clarify ARPA’s open-endedness, the States still face a live threat to their sovereign authority. The Rules seek to allow states to enact tax cuts resulting in revenue reductions so long as they identify permissible sources of offsetting funds. See, e.g., 87 Fed. Reg. at 4428. Although allowing for some tax cuts, the Rules still narrow the range of permissible tax policies the States may enact, which in turn takes a toll on the States’ citizens and economies.

What’s more, the Rules do not ease the States’ concern regarding Treasury’s enforcement against them. Secretary Yellen’s threat to enforce the Offset Provision in her earlier letter to the States still stands. (See R. 1-2, Yellen Letter, PageID 33–34.) Indeed, the Rules make matters worse. They expressly reserve the Secretary’s broad authority to enforce ARPA’s requirements and provide that “[n]othing” in the Rules “shall limit the authority of the Secretary to take action to enforce conditions or violations of law . . . .” 31 C.F.R. § 35.4(a) (2022). So the Rules do not limit ARPA’s enforcement; they instead provide the Secretary broad enforcement discretion. If a state accepts the funds and the Secretary believes that the state violates ARPA, that state must “repay to the Secretary an amount equal to the amount of funds used in violation [thereof].” 42 U.S.C. § 802(e). All this considered, a credible threat of enforcement still stands.

Next, while facing “live” issues, the States continue to have “a legally cognizable interest in the outcome” of this case. Thomas, 996 F.3d at 323. The Rules do not “mak[e] it ‘impossible for the [C]ourt to grant any effectual relief[.]’” Id. at 330 (quotations omitted). The Court can still grant injunctive and declaratory relief given the States’ continued stake in the case without creating an “advisory opinion that Article III prohibits.” Id. That such relief “would have [a] No. 21-6108 Commonwealth of Ky., et al. v. Yellen, et al. Page 51

practical effect” on the States reaffirms that mootness does not pose a problem here. Id.; see Ford, 469 F.3d at 504. Because ARPA’s vague conditions still restrict States from enacting tax cuts in their sovereign capacities and Treasury-recoupment actions remain a credible threat, the relief sought affects the parties’ current legal interests. See Ford, 469 F.3d at 504–05. If granted, the States’ relief would help them avoid enforcement of ARPA’s unconstitutionally vague conditions.

For these reasons, I do not find the claims moot like the majority. I acknowledge that the States did not allege they would fail to offset their funds with a permissible revenue source. See 87 Fed. Reg. at 4428. But I don’t believe that moots Treasury’s threat of recoupment action. Again, the Rules still limit the States from enacting tax policies if they do not offset a net reduction with permissible revenue sources. This restraint makes the States fear that they “cannot lower their citizens’ tax burdens without suffering a penalty.” (R. 23, Amended Complaint, PageID 131 ¶ 1.) And seeing that the Rules do not “limit the authority of the Secretary” to enforce ARPA, a credible threat of recoupment action still stands. 31 C.F.R. § 35.4(a). Or if the States wish to comply with the Rules, they must do something—either raise other taxes or lower expenditures elsewhere in the budget to offset a revenue reduction. That something creates an ongoing injury. On that note, the Rules also leave the States’ claims of vagueness intact, which leads ARPA to intrude upon the States’ tax powers—“an area traditionally left to the states.” Kentucky, 23 F.4th at 599. And the chilling effects and the threats to the States’ economies remain. See id. at 599–601. Because the States still face the effects of ARPA’s vagueness, the Rules do not moot this case.

I concur with almost all of the majority opinion. But I respectfully dissent in part only to explain why I believe the Rules do not moot the States’ controversy over their sovereign-tax interests. Thus, I would grant both Kentucky and Tennessee their sought-after relief.