Rose, Jr. v. Azar, II

Opinion

UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

MONTE A. ROSE, JR., et al.,

Plaintiffs, v. Civil Action No. 19-2848 (JEB)

XAVIER BECERRA, Secretary of Health and Human Services, et al.,

Defendants.

MEMORANDUM OPINION

During the Trump Administration, the U.S. Department of Health and Human Services encouraged states to seek waivers from HHS permitting them to tinker with their Medicaid programs. In particular, certain states sought to add work requirements and other obligations in order for their citizens to obtain Medicaid or remain on its rolls. When affected citizens of Kentucky, Arkansas, and New Hampshire challenged the Secretary’s issuance of those waivers, they wound up before this Court. In a series of Opinions, appeals from which were affirmed (or dismissed as moot) by the D.C. Circuit, this Court invalidated the waivers, finding them inconsistent with the Medicaid Act. See Stewart v. Azar (Stewart I), 313 F. Supp. 3d 237, 243 (D.D.C. 2018); Stewart v. Azar (Stewart II), 366 F. Supp. 3d 125, 138 (D.D.C. 2019); Gresham v. Azar, 363 F. Supp. 3d 165, 169 (D.D.C. 2019), aff’d, 950 F.3d 93 (D.C. Cir. 2020), vacated and remanded as moot sub nom., Becerra v. Gresham, 142 S. Ct. 1665 (2022); Philbrick v. Azar, 397 F. Supp. 3d 11, 16 (D.D.C. 2019), aff’d, 2020 WL 2621222 (D.C. Cir. May 20, 2020), vacated and remanded as moot sub nom., Becerra, 142 S. Ct. 1665. While the book on this saga has been closed for some time, the sequel now arrives.

2

Since 2007, Indiana has offered Hoosiers a health plan — aptly named the Healthy Indiana Plan or HIP — designed to resemble commercial, high-deductible insurance. While HIP initially targeted individuals who were ineligible for Medicaid, it evolved over time. Following the enactment of the Patient Protection and Affordable Care Act (ACA), which expanded Medicaid to cover “the entire nonelderly population with income below 133 percent of the poverty level,” Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 583 (2012), the State unveiled HIP 2.0. That version of the program offers a two-tiered plan to all non-disabled adults in Indiana with incomes at or below 138% of the federal poverty level (FPL). It restricts coverage, however, in several ways, including by charging monthly income-based premiums, offering no retroactive coverage, and providing no assurance of non-emergency medical transportation (NEMT). The Secretary of Health and Human Services approved HIP 2.0 in January 2015, waiving several core Medicaid requirements in the process.

HHS has permitted Indiana to extend the program several times since then. Most recently, in October 2020, the Secretary approved the current incarnation of HIP 2.0 for a ten- year period. After the Biden Administration took the reins, the agency informed the State that it would reassess that approval. In December 2023, however, it announced that although it had concerns with the State’s premium requirements, it would not take any action to withdraw the approval at that time.

In this suit, three Indiana residents challenge the Secretary’s 2020 approval and 2023 decision, arguing that both violate the Administrative Procedure Act. In now seeking summary judgment, the challengers contend that the Secretary’s 2020 approval of the State’s plan suffers from the same problems that befell his prior approvals of other states’ plans and therefore must meet the same fate. HHS and Indiana counter in a Cross-Motion and a Motion to Dismiss,

3

respectively, that HIP 2.0 is a different kettle of fish. That is slightly ironic given that the agency pointed to Indiana’s program repeatedly in its Stewart I briefing as being similar to Kentucky’s plan, see, e.g., Stewart v. Azar, No. 18-152, ECF Nos. 54 (HHS Opp.) at 15 n.6, 37 & n.13; 107 (HHS Opp. on Remand) at 30–31, 35 n.9, 38, but that is neither here nor there. At the end of the day, as this Court has explained before, the Secretary has significant discretion to approve demonstration projects that promote the objectives of the Medicaid Act, and it is not for the Court to second guess his policy decisions or substitute its judgment for his. “But courts retain a role, and an important one, in ensuring that agencies have engaged in reasoned decisionmaking.” Judulang v. Holder, 565 U.S. 42, 53 (2011). That includes evaluating “whether the decision was based on a consideration of the relevant factors.” Id. (citation omitted).

On that score, the Court concludes that HHS has once again failed to comply with this fundamental administrative-law requirement. Although the Secretary sprinkled the word “coverage” into his approval letter this time, he did not adequately consider whether the program “would in fact help the state furnish medical assistance to its citizens, a central objective of Medicaid.” Stewart I, 313 F. Supp. 3d at 243. Nor does Defendants’ all-too-familiar rejoinder that HIP 2.0 promotes other purposes of Medicaid — i.e., health outcomes and fiscal sustainability — remedy this deficiency. The Court will therefore grant summary judgment to Plaintiffs and vacate the 2020 approval. Because such vacatur affords Plaintiffs full relief, there is no need for the Court to grapple with whether the agency’s 2023 letter — which declined to withdraw that approval — was also arbitrary and capricious or to consider Plaintiffs’ myriad other reasons why the 2020 approval was unlawful.

4

I. Background The Court begins with an overview of the relevant history and provisions of the Medicaid Act. It then summarizes the Kentucky, Arkansas, and New Hampshire litigation — which, given its similarities to what is transpiring here, provides important context — before turning to Indiana’s challenged plan and, finally, this case’s procedural history.

A. Legal Background Since 1965, the federal government and the states have worked together to provide medical assistance to certain vulnerable populations under Title XIX of the Social Security Act, commonly known as Medicaid. See 42 U.S.C. § 1396-1. The Centers for Medicare and Medicaid Services (CMS), a federal agency within the Department of Health and Human Services, has primary responsibility for overseeing Medicaid programs. Under the cooperative federal-state arrangement, participating states submit their “plans for medical assistance” to the Secretary of HHS. Id. To receive federal funding, those plans — along with any material changes to them — must be “approved by the Secretary.” Id.; see also 42 C.F.R. § 430.12(c)(2). Currently, all states have chosen to participate in the program.

To be approved, state plans must comply with certain minimum parameters set out in the Medicaid Act. See 42 U.S.C. § 1396a(a) (listing over 80 separate requirements). One such provision requires state plans to “mak[e] medical assistance available” to certain low-income individuals. Id. § 1396a(a)(10)(A). That group originally comprised pregnant women, children, and their families; some foster children; the elderly; and people with certain disabilities. Id. In 2010, however, Congress enacted the ACA, colloquially known as Obamacare, “to increase the number of Americans covered by health insurance.” NFIB, 567 U.S. at 538. Of relevance here, that statute required participating states to expand Medicaid coverage to additional low-income

5

adults under 65 who did not previously qualify — the so-called expansion population. See 42 U.S.C. § 1396a(a)(10)(A)(i)(VIII).

Generally, a state must cover all qualified individuals or forfeit its federal Medicaid funding. Id. § 1396a(a)(10)(B); id. § 1396c. That was originally so for the ACA expansion population as well. Id. § 1396c. In NFIB, however, the Supreme Court held that Congress could not, consistent with the Spending Clause of the Constitution, condition a state’s entire Medicaid funds on its agreeing to the expansion. See 567 U.S. at 584–85. As a result, states could choose not to cover the new population and lose no more than the funds that would have been appropriated for that group. Id. at 587. If the state, conversely, does decide to provide coverage, those individuals would become part of its mandatory population. Id. at 585–87 (explaining that Congress may “offer[] funds under the Affordable Care Act to expand the availability of health care, and require[] that States accepting such funds comply with the conditions on their use”). In that instance, the state must afford the expansion group “full benefits” — i.e., it must provide “medical assistance for all services covered under the State plan” that are substantially equivalent “in amount, duration, or scope . . . to the medical assistance available for [other] individual[s]” covered under the Act. See 42 U.S.C. § 1396d(y)(2)(B); 42 C.F.R. § 433.204(a)(2).

The Medicaid Act, in addition to defining who is entitled to coverage, also ensures what coverage those enrolled individuals receive and at what cost. Under § 1396a, states must cover certain basic medical services, see 42 U.S.C. §§ 1396a(a)(10)(A), 1396d(a), and the statute limits the amount and type of premiums, deductions, or other cost-sharing charges that a state can impose on such care. Id. § 1396a(a)(14); see also id. § 1396o. Other provisions that are relevant here require states to provide up to three months of retroactive coverage once a beneficiary

6

enrolls, id. § 1396a(a)(34), and to ensure that recipients receive all “necessary transportation . . . to and from providers.” 42 C.F.R. § 431.53. Finally, states must “provide such safeguards as may be necessary to assure that eligibility” and services “will be provided, in a manner consistent with simplicity of administration and the best interests of the recipients.” 42 U.S.C. § 1396a(a)(19).

Both before and after the passage of the ACA, a state accepting federal Medicaid funds is not entirely locked in; instead, if it wishes to deviate from certain of the Act’s requirements, it can seek a waiver from the Secretary of HHS. See 42 U.S.C. § 1315. In particular, Section 1115 of the Social Security Act allows the Secretary to approve “experimental, pilot, or demonstration project[s] which, in [his] judgment . . . , [are] likely to assist in promoting the [Act’s] objectives.” 42 U.S.C. § 1315(a). As conceived, experimental projects were “expected to be selectively approved by the Department and to be those which are designed to improve the techniques of administering assistance.” S. Rep. No. 1589, 87th Cong., 2d Sess. 19, reprinted in 1962 U.S.C.C.A.N. 1943, 1962. Once the Secretary has greenlighted such a project, he can then waive compliance with the requirements of § 1396a “to the extent and for the period . . . necessary to enable [the] State . . . to carry out such project.” 42 U.S.C. § 1315(a)(1).

While the ultimate decision whether to grant Section 1115 approval rests with the Secretary, he must nonetheless jump through certain hoops. Before HHS can act on a waiver application, for example, the state “must provide at least a 30-day public notice[-]and[-]comment period” regarding the proposed program and hold at least two hearings at least 20 days before submitting the application. See 42 C.F.R. §§ 431.408(a)(1), (3). Once a state completes those prerequisites, it then sends an application to HHS. Id. § 431.412 (listing application requirements). After the agency notifies the state that it has received the waiver application, a

7

federal 30-day public-notice period commences, and the agency must wait at least 45 days before rendering a final decision. Id. §§ 431.416(b), (e)(1).

B. Factual Background 1. Prior Section 1115 Litigation Quite a few states have been interested in utilizing Section 1115 waivers to experiment with Medicaid reform in recent years, especially after the Trump Administration promised to “take all actions consistent with law to minimize” the ACA’s impact until it could be repealed — which, as we all now know, never occurred — including “us[ing] existing Section 1115 demonstration authority” to revamp Medicaid. See Exec. Order No. 13765, 82 Fed. Reg. 8351 (Jan. 20, 2017); Sec’y of Health & Hum. Servs., Dear Governor Letter (Mar. 14, 2017), https://perma.cc/E5ZA-WFGW. As the Court noted earlier, CMS approved demonstration projects submitted by Kentucky, Arkansas, and New Hampshire in 2018, and this Court struck down each. While it focused largely on the impermissibility of the projects’ work requirements, there were other infirmities, too.

Kentucky’s program — called Kentucky HEALTH — required non-exempt adults aged 19 to 64 who received Medicaid coverage through the expansion to complete and report 80 hours per month of qualifying activities, such as employment, education, or job training. Stewart I, 313 F. Supp. 3d at 246. The failure to do so or to report an exemption resulted in the termination of Medicaid coverage. Id. at 246–47. Beneficiaries were also required to pay monthly premiums, provide information for an annual redetermination, and report changes in income or circumstances that affected Medicaid eligibility promptly; failure to do so could result in a six- month “lockout” during which the State could deny Medicaid coverage for any beneficiary with an income above 100% of the FPL. Id. Like Indiana’s program, Kentucky HEALTH offered

8

sharply restricted retroactive coverage or assurance of NEMT. Id. at 246. Although the Commonwealth estimated that about 95,000 persons would lose Medicaid within five years of implementation given the program’s restrictions, the Secretary nevertheless approved it under Section 1115 and granted the requisite waivers to implement it because he found that it was likely to “improv[e] health outcomes” and “increas[e] individual engagement in health care decisions.” Id. at 258, 262 (cleaned up).

Several Kentuckians challenged that approval, contending that HHS had not satisfactorily explained why Kentucky HEALTH promoted Medicaid’s objectives and that its approval exceeded the agency’s statutory authority. After analyzing the “objectives” of Medicaid, this Court concluded that the plaintiffs were right that “the Secretary never adequately considered whether Kentucky HEALTH would in fact help the state furnish medical assistance to its citizens, a central objective of Medicaid.” Id. at 243; see also id. at 260–61 (“Congress evinced a clear interest in enabling each State, as far as practicable, to provide payment of part or all of the cost of medical care and services” for populations covered by the Act.) (cleaned up). The Court therefore vacated his approval and remanded the matter to the agency. Id. at 273–74.

HHS eventually reapproved Kentucky’s project, reasoning that it would further the Medicaid Act’s objectives by 1) promoting the health and financial independence of beneficiaries, a justification the Court had found wanting in the first round, 2) increasing coverage by allowing Kentucky to cover the expansion population when it would not do so otherwise, and 3) advancing the fiscal sustainability of the state’s Medicaid program. Stewart II, 366 F. Supp. 3d at 138, 140. Believing these justifications still unsatisfactory, the Bluegrass State plaintiffs returned to this Court, which again agreed. Concluding that the agency’s rationales fared no better this time around and that its new explanation still did not contend with

9

the possibility of coverage loss, the Court vacated the approval for the second time. Id. at 138– 39.

Arkansas’s project — the Arkansas Works Amendments — followed a similar, albeit more abbreviated, path. Like Kentucky, the Bear State proposed to require most able-bodied beneficiaries in the expansion population aged 19 to 49 to complete 80 hours per month of qualifying employment or other activities and to eliminate retroactive eligibility. Gresham, 363 F. Supp. 3d at 172. Arkansas did not estimate the effects those amendments would have on coverage, but the Secretary approved them because he believed that they were “likely to assist in improving health outcomes” and “incentivize beneficiaries to engage in their own health care.” Id. Several residents of Arkansas sued, adopting a position identical to that of the challengers in Stewart I. Finding its guiding principle in Yogi Berra’s aphorism, “It’s déjà vu all over again,” the Court concluded that the Secretary’s failures were nearly identical to those in Stewart I: he did not adequately assess the effects on Medicaid coverage. Despite conceding that providing medical care to the needy is “Medicaid’s core objective[,]” HHS “neither offered [its] own estimates of coverage loss nor grappled with comments in the administrative record projecting that the [proposal] would lead a substantial number of Arkansas residents to be disenrolled from Medicaid.” Id. at 175 (cleaned up). Its approval, accordingly, was vacated, too. Id. at 185.

New Hampshire’s so-called Granite Advantage likewise conditioned Medicaid coverage on compliance with a work requirement — in fact, an even stricter one than Kentucky’s and Arkansas’s — and eliminated retroactive coverage. Philbrick, 397 F. Supp. 3d at 18. A few New Hampshirites brought suit, arguing that the Secretary’s approval suffered from the same deficiency as the others. Id. at 15. The Court concurred: “[T]he agency has still not contended

10

with the possibility that the project would cause a substantial number of persons to lose their health-care coverage.” Id. And so, like the plans before it, Granite Advantage fell. Id. at 33.

The D.C. Circuit affirmed this Court’s decisions as to the Arkansas Works Amendments, Gresham, 950 F.3d 93, and Granite Advantage, Philbrick, 2020 WL 2621222. Kentucky voluntarily terminated Kentucky HEALTH while its appeal was pending, and so the Court of Appeals granted the Commonwealth’s motion to dismiss. Gresham, 950 F.3d at 95–96. The Supreme Court subsequently granted certiorari to review the decisions on both Arkansas’s and New Hampshire’s programs, Azar v. Gresham, 141 S. Ct. 890 (2020), but HHS (with the Biden Administration now in office) mooted the case when it withdrew approvals of those programs in 2021. The Supreme Court, accordingly, granted the Government’s motion to vacate the judgments in Gresham and Philbrick under United States v. Munsingwear, Inc., 340 U.S. 36 (1950). See Becerra, 142 S. Ct. 1665.

2. Healthy Indiana Plan

This is the context in which the program challenged here arrives, and its history is itself somewhat long and winding. The Healthy Indiana Plan’s original iteration dates back to 2007, long before the passage of the ACA. See 2007 Ind. Acts 3525. The first of its kind, HIP coupled the optional provision of healthcare coverage with several features that required waivers of Medicaid requirements, including monthly premiums, limits on retroactive eligibility, limits on NEMT, reporting requirements, and lockout penalties associated with nonpayment of premiums or failure to comply with reporting requirements. See ECF No. 50-1, Exh. B (2007 Special Terms and Conditions) at 49–97; AR 8264. Believing that it would “test a model of health coverage that emphasizes private health insurance, personal responsibility, and ‘ownership’ of

11

health care,” HHS in 2007 granted the requisite waivers under Section 1115 for Indiana to implement HIP for five years. See ECF No. 50-1 (Approval) at 44.

After the passage of the ACA and the Supreme Court’s decision in NFIB, Indiana was not keen on a traditional Medicaid expansion. Before his entrance on the national stage, then- Governor Mike Pence announced that the State would “predicate any expansion of Medicaid . . . on [its] ability to promote Hoosier innovation in the Healthy Indiana Plan to the expanded population.” Letter from Gov. Pence to Sec’y of Health & Hum. Servs. at 2 (Feb. 13, 2013), https://perma.cc/AN6X-AXAE. In other words, Indiana would expand through HIP or not at all. See Letter from Gov. Pence to President Obama at 2 (Oct. 2, 2014), https://perma.cc/27P6-QQHQ (“[F]rom the outset of my administration I have made it clear that Indiana will not expand traditional Medicaid . . . . Our administration will not support efforts to remove or water down the Healthy Indiana Plan’s core principles, essentially changing this proven program into an expansion of traditional Medicaid.”). HHS approved a series of short- term HIP extensions while it negotiated with the State. See AR 8240.

In 2014, Indiana proposed a new version of HIP, aptly rebranded HIP 2.0, which would replace traditional Medicaid for all non-disabled low-income adults ages 19 to 64, including the expansion population and certain parents and caretakers. See AR 8238–40. It would retain many features of HIP except that, unlike its predecessor, it would offer two tiers of benefits: HIP Plus, which would cover vision and dental services and require no co-payments, and HIP Basic, which would not cover those services and would require co-payments. See ECF No. 32-3 (2015 Approval) at 2–3. Enrollees who paid monthly premiums into their so-called “POWER Accounts” (designed to operate like health-savings accounts) would be covered by the former, while enrollees who failed to pay would either be defaulted to the latter or have their coverage

12

terminated, depending on their income level. Id. HHS approved the second-generation of HIP under Section 1115 for a three-year period in 2015. Id. at 1–5.

In 2017, Indiana applied to extend it for three more years. See 2019 AR 2577–670. The State explained that it hoped to “maintain and develop” HIP 2.0 with only minor modifications “as Congress and the Administration develop much needed plans for repealing and replacing ObamaCare.” Letter from Gov. Eric J. Holcomb to Norris Cochran, Acting Sec’y of Health & Hum. Servs. (Jan. 31, 2017). It subsequently amended that application to add a community- engagement requirement akin to Kentucky’s, Arkansas’s, and New Hampshire’s work requirements. See 2019 AR 3832–924. In a letter that mirrors the ones approving those states’ programs (at some points verbatim), HHS approved the Hoosier State’s extension application — including the community-engagement requirement — in 2018. See 2019 AR 1–80.

This requirement, however, was short lived. In 2019, after Plaintiffs filed their original Complaint in this case, Indiana paused its implementation. The State explained that, although it remained committed to the requirement, it had determined to suspend it “to allow time for [this] lawsuit to be resolved and so that the court can address the challenge to HIP after similar legal challenges to [work requirements] in other states have worked their way through the appeals process.” News Release, Ind. Fam. & Soc. Servs. Admin., Pending Resolution of Federal Lawsuit, FSSA Will Temporarily Suspend Gateway to Work Reporting Requirement (Oct. 31, 2019), https://perma.cc/5WHM-BZQS. This was certainly prudent since, as the chronologically savvy reader will note, the appeals in Stewart, Gresham, and Philbrick were pending before the D.C. Circuit at the time.

On January 31, 2020, with the end date of the 2018 approval nearing, Indiana requested that HHS extend HIP 2.0 again, this time “with no substantive changes,” citing the program’s

13

“long-tenure and demonstrated success.” AR 8238. Unlike in the past, the State sought a ten- year extension, believing that shorter approval periods “create[] unnecessary administrative burdens for the State and federal government, and do[] not meaningfully enhance the oversight or transparency of the demonstration.” AR 8239. As fate would have it, that same day the Secretary declared a public-health emergency in light of COVID-19. See U.S. Dep’t of Health & Hum. Servs., Determination that a Public Health Emergency Exists (Jan. 31, 2020), https://perma.cc/C89D-W7E4. All states received enhanced federal funding to prevent the loss of Medicaid coverage during the pandemic, see Families First Coronavirus Response Act, Pub. L. No. 116-127, 134 Stat. 178 (2020), and Indiana accordingly suspended its premium requirements and lockout penalties. See Gov. Eric Holcomb, Exec. Order No. 20-05 (Mar. 19, 2020), https://perma.cc/MH9D-7PCC.

Fast forward to October 26, 2020: HHS approved the State’s application to extend HIP 2.0 for ten years, in part conditionally. See AR 1554–1752. Specifically, he conditioned approval of the lockout penalties and community-engagement requirement (both of which the State had asked to retain, even though it was presently implementing neither) “on the Supreme Court issuing a decision in Azar v. Gresham[] that legally authorizes these elements of the proposed extension to the demonstration.” AR 1554–55. He otherwise granted the request to extend HIP 2.0 unconditionally, finding that it would, inter alia, “furnish medical assistance in a manner that improves the sustainability of the safety net,” “test[] reforms designed to promote financial independence, which [would] improve continuity of coverage and lead to better health outcomes,” and “remove potential obstacles to a successful beneficiary transition to commercial coverage.” AR 1561–66. This is the approval that Plaintiffs challenge here.

14

But that is not the end of the story. Shortly after President Biden took office in 2021, HHS sent letters to states that had previously received approvals (conditional or otherwise) to implement work or community-engagement requirements, explaining that it had “preliminarily determined” that such requirements “would not promote the objectives of the Medicaid program.” AR 1076. In its letter to Indiana, the agency announced that it was “commencing a process of determining whether to withdraw” its conditional approval of the community- engagement requirement, as well as its approval of the other HIP 2.0 waiver authorities. Id. Sure enough, five months later, the Secretary withdrew the conditional approval of Indiana’s community-engagement requirement and reiterated that he was in the process of reassessing the other waiver authorities that had been approved by his predecessor in 2020. See AR 838–61.

On December 22, 2023, HHS sent Indiana a letter — which Plaintiffs also challenge —

announcing that it had concluded that reassessment. See AR 1. Although it had “concerns with premium requirements in section 1115 demonstrations generally based on the large body of evidence suggesting that premiums beyond those authorized under the Medicaid statute reduce access to coverage and care among populations that Medicaid is designed to serve,” the agency explained that it would not be “taking any action now on the premium authority, or any other authority, in the approved HIP demonstration.” AR 1. That was because, “given the totality of the circumstances,” it had “concluded that withdrawing those authorities at [that] time [was] too disruptive, particularly in the context of the state needing to maintain focus on keeping people covered through Medicaid unwinding and the resumption of Medicaid renewals following the COVID-19 Public Health Emergency.” AR 1.

15

As a result, HIP 2.0 lives on to this day. The key features that the State has authority to implement at present — thanks to the waivers that the Secretary granted in 2020 and declined to withdraw in 2023 — are as follows:

1) Limits on retroactive eligibility, which excuse Indiana from providing three months of retroactive eligibility for HIP 2.0 beneficiaries, except pregnant women;

2) Monthly premiums, which vary based on income level and include a surcharge for tobacco users; the State can also change premium amounts annually without requesting an amendment, so long as the premiums do not exceed 3% of household income;

3) Limits on NEMT, which relieve the State of the requirement to assure NEMT to and from providers for members of the expansion population — i.e., adults without disabilities — except for those who are medically frail; and

4) Reporting requirements, which mandate that individuals provide information for an annual redetermination and report changes in income or circumstances that affect Medicaid eligibility within 10 days.

The program notably no longer includes any community-engagement requirements or lockout penalties. Authority for the former was withdrawn in 2021, as explained above, see AR 838–61, and authority for the latter was contingent on the Supreme Court’s issuing a decision in Gresham that it never did. See AR 1554–55; see also ECF No. 55 (HHS MSJ) at 9 n.2. Readers familiar with this Court’s decisions in Stewart I, Stewart II, Gresham, and Philbrick should thus keep in mind that the current challenge has nothing to do with work (or community-engagement) requirements.

A final note on premiums: although the COVID-19 public-health emergency ended in May 2023, see U.S. Dep’t of Health & Hum. Servs., HHS Secretary Xavier Becerra Statement on End of the COVID-19 Public Health Emergency (May 11, 2023), https://perma.cc/XK25-GWFB, Indiana has not yet resumed implementation of HIP 2.0’s premium requirements but apparently

16

plans to do so on July 1. See ECF No. 49 (Joint Status Report), ¶ 10; see Family & Soc. Servs. Admin, How a Return to Normal Will Impact Some Indiana Medicaid Members (last visited June 19, 2024), https://perma.cc/BE5P-VSJ7.

3. Indiana’s SUD/SMI Programs Back when Indiana applied to renew HIP 2.0 in 2017, it also sought approval for a substance-use-disorder (SUD) program. See 2019 AR 2609–13. Traditionally, Medicaid bars states from receiving “payments with respect to care or services for any individual who has not attained 65 years of age and who is a patient in an institution for mental diseases [IMD].” 42 U.S.C. § 1396d(a)(32)(B); see also id. § 1396d(a)(14), (16)(A) (separately allowing payments for individuals under age 21). An IMD is a “hospital, nursing facility, or other institution . . . that is primarily engaged in providing diagnosis, treatment, or care of persons with mental diseases.” Id. § 1396d(i). In other words, the statute prohibits the federal government from reimbursing treatment in mental-health facilities (at least for beneficiaries between 21–64).

Increasingly, this provision has posed problems for states. An estimated 21% of Medicaid-eligible adults suffer from a substance-use disorder, and Indiana’s citizens are no exception. See 2019 AR 2600. In 2017, the State explained that “[d]rug addiction is a widespread problem in Indiana that affects the lives of far too many Hoosiers,” and it noted that since 2009, “more Hoosiers have lost their lives due to a drug overdose than in automobile accidents on state highways.” 2019 AR 2610. Medicaid’s IMD exclusion, it explained, had “resulted not only in a lack of access to appropriate mental health services for certain Medicaid beneficiaries, but also an increase in the amount of uncompensated care IMDs provide to adult Medicaid beneficiaries and the indigent.” 2019 AR 2612. The waiver of that exclusion, it asserted, “would allow Medicaid patients to access at least 15 new facilities across the state, and

17

potentially increase capacity at 12 other facilities.” 2019 AR 2613. Concurring, HHS approved the State’s SUD Program in 2018, noting that it would “ensure that a broad continuum of care is available to Indiana Medicaid beneficiaries with a SUD.” 2019 AR 3.

When Indiana sought a ten-year extension of HIP 2.0 in 2020, it separately requested a five-year extension of the SUD program, which it had recently amended to include a serious mental illness (SMI) component. See AR 8288–315. The Secretary approved the five-year extension of the SUD/SMI programs, finding that “[t]he SUD program directly supports Medicaid’s objectives by improving access to high-quality services, and it is critical to addressing Indiana’s substance use epidemic,” and the SMI program “assists in delivering high- quality, clinically appropriate treatment to beneficiaries diagnosed with SMI and receiving treatment while they are short-term residents in free-standing psychiatric hospitals that qualify as IMDs.” AR 1561. Plaintiffs have not challenged the approval of the SUD/SMI programs here, but they are relevant inasmuch as they play a role in the Court’s forthcoming explanation of the scope of its analysis.

C. Procedural History Several residents of Indiana filed this lawsuit on September 23, 2019, challenging the Secretary’s 2018 approval of HIP 2.0. See ECF No. 1 (Compl.). Because it was designated as related to Stewart, Gresham, and Philbrick, the case was assigned to this Court. See ECF No. 2 (Notice of Related Case). The State of Indiana then intervened as a defendant. See ECF No. 12 (Mot. to Intervene); Minute Order of Oct. 15, 2019 (granting Mot. to Intervene). From November 2019 until March 2024, proceedings were stayed. See Minute Orders of Nov. 21, 2019 (staying case pending resolution of appeals in Stewart and Gresham); Apr. 6, 2020

18

(continuing stay during pendency of COVID public-health emergency); Feb. 9, 2024 (continuing stay while HHS completed review of 2020 approval); Mar. 28, 2024 (lifting stay).

In January 2024, three Hoosiers filed a supplemental Complaint, which is the operative one here and no longer challenges the Secretary’s 2018 approval of HIP 2.0. See ECF No. 50-1 (Supp. Compl.). Instead, it takes aim at the Secretary’s 2020 approval and 2023 letter declining to withdraw that approval. Id., ¶¶ 179–210. Plaintiffs assert that both were arbitrary and capricious in violation of the APA. Id. at 39. Dueling Cross-Motions for Summary Judgment, as well as a Motion to Dismiss, are now ripe. II. Legal Standard A. Motions for Summary Judgment The challengers and HHS have cross-moved for summary judgment on the administrative record. The summary-judgment standard set forth in Federal Rule of Civil Procedure 56(c), therefore, “does not apply because of the limited role of a court in reviewing the administrative record.” Sierra Club v. Mainella, 459 F. Supp. 2d 76, 89 (D.D.C. 2006); see also Bloch v. Powell, 227 F. Supp. 2d 25, 30 (D.D.C. 2002), aff’d, 348 F.3d 1060 (D.C. Cir. 2003). “[T]he function of the district court is to determine whether or not as a matter of law the evidence in the administrative record permitted the agency to make the decision it did.” Sierra Club, 459 F. Supp. 2d. at 90 (cleaned up). “Summary judgment is the proper mechanism for deciding, as a matter of law, whether an agency action is supported by the administrative record and consistent with the [APA] standard of review.” Loma Linda Univ. Med. Ctr. v. Sebelius, 684 F. Supp. 2d 42, 52 (D.D.C. 2010) (cleaned up).

The APA “sets forth the full extent of judicial authority to review executive agency action for procedural correctness.” FCC v. Fox Television Stations, Inc., 556 U.S. 502, 513

19

(2009). It requires courts to “hold unlawful and set aside agency action, findings, and conclusions” that are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2). Agency action is arbitrary and capricious if, for example, the agency “entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Motor Vehicle Mfrs. Ass’n of U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).

In other words, an agency is required to “examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.” Id. at 43 (quoting Burlington Truck Lines v. United States, 371 U.S. 156, 168 (1962)) (cleaned up). Courts, accordingly, “do not defer to the agency’s conclusory or unsupported suppositions,” United Techs. Corp. v. Dep’t of Def., 601 F.3d 557, 562 (D.C. Cir. 2010) (quoting McDonnell Douglas Corp. v. Dep’t of the Air Force, 375 F.3d 1182, 1187 (D.C. Cir. 2004)), and “agency ‘litigating positions’ are not entitled to deference when they are merely [agency] counsel’s ‘post hoc rationalizations’ for agency action, advanced for the first time in the reviewing court.” Martin v. Occupational Safety & Health Review Comm’n, 499 U.S. 144, 156 (1991) (citation omitted). Although a reviewing court “may not supply a reasoned basis for the agency’s action that the agency itself has not given,” a decision that is not fully explained may, nevertheless, be upheld “if the agency’s path may reasonably be discerned.” Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285–86 (1974) (citation omitted).

20

B. Motion to Dismiss Indiana’s Motion to Dismiss invokes the legal standards for dismissal under Federal Rules of Civil Procedure 12(b)(1) and 12(b)(6). When a defendant brings a Rule 12(b)(1) motion, the plaintiff must demonstrate that the court indeed has subject-matter jurisdiction to hear his claims. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 561 (1992); U.S. Ecology, Inc. v. U.S. Dep’t of Interior, 231 F.3d 20, 24 (D.C. Cir. 2000). “Because subject-matter jurisdiction focuses on the court’s power to hear the plaintiff’s claim, a Rule 12(b)(1) motion [also] imposes on the court an affirmative obligation to ensure that it is acting within the scope of its jurisdictional authority.” Grand Lodge of Fraternal Order of Police v. Ashcroft, 185 F. Supp. 2d 9, 13 (D.D.C. 2001). For this reason, “the [p]laintiff’s factual allegations in the complaint . . . will bear closer scrutiny in resolving a 12(b)(1) motion’ than in resolving a 12(b)(6) motion for failure to state a claim.” Id. at 13–14 (citation omitted).

Rule 12(b)(6), conversely, provides for the dismissal of an action where a complaint fails to “state a claim upon which relief can be granted.” In evaluating a motion to dismiss under that Rule, the court must “treat the complaint’s factual allegations as true and must grant plaintiff the benefit of all inferences that can be derived from the facts alleged.” Sparrow v. United Air Lines, Inc., 216 F.3d 1111, 1113 (D.C. Cir. 2000) (internal quotation marks and citation omitted). A court need not accept as true, however, “a legal conclusion couched as a factual allegation.” Trudeau v. FTC, 456 F.3d 178, 193 (D.C. Cir. 2006) (citation omitted). Although “detailed factual allegations” are not necessary to withstand a Rule 12(b)(6) motion, Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555 (2007), “a complaint must contain sufficient factual matter, [if] accepted as true, to state a claim to relief that is plausible on its face.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (internal quotation marks omitted). A plaintiff may survive a

21

Rule 12(b)(6) motion even if “recovery is very remote and unlikely,” but the allegations “must be enough to raise a right to relief above the speculative level.” Twombly, 550 U.S. at 555–56. III. Analysis Like their predecessors from the other states, Plaintiffs target nearly every aspect of the demonstration project they ask the Court to strike down. First, they challenge HIP 2.0 “as a whole,” maintaining that the Secretary erred in his 2020 approval by finding that the project was likely to further objectives of Medicaid, failing to adequately consider whether the project was “experimental,” and lacking statutory authority to grant a ten-year extension. See Supp. Compl. at 34–35 (Count I) (capitalization altered). Second, they attack individual features of the program — i.e., the premiums, the lack of retroactive coverage, and the limits on NEMT — on the ground that each is unlikely to promote Medicaid’s goals. Id., ¶¶ 192–210 (Counts III–V). In Count III, Plaintiffs further allege that the Secretary cannot approve premium requirements through his Section 1115 authority. In addition, Plaintiffs challenge his issuance of the 2023 letter declining to withdraw the 2020 approval. Id., ¶¶ 188–91 (Count II).

As in Stewart I, and for reasons discussed in more detail below, the Court need adjudicate only one count of Plaintiffs’ Complaint to grant them full relief: Count I, which challenges the Secretary’s 2020 approval of HIP 2.0 as a whole. Before it can reach that dispute, however, the Court must first take up several threshold issues.

A. Threshold Issues Because they’ve seen this film before (and they didn’t like the ending), Defendants seek to leave out the side door. Cf. Taylor Swift, Exile, on Folklore (Republic Records 2020). Specifically, they argue that this Court has no power to review the Secretary’s 2020 approval because (1) Plaintiffs lack standing, and (2) the decision is “committed to agency discretion by

22

law,” 5 U.S.C. § 701(a)(2), thus barring any judicial oversight under the APA. The Court rejected similar arguments in Stewart I, and their sequel yields the same result.

1. Standing

Article III of the Constitution limits the jurisdiction of federal courts to actual “Cases”

and “Controversies.” U.S. Const., art. III, § 2. But not just any dispute will do. See Lujan, 504 U.S. at 559–61. A plaintiff must demonstrate that she suffers: 1) an injury-in-fact that is 2) caused by the conduct complained of and 3) is “likely” to be “redressed by a favorable decision.” Id. at 560–61 (cleaned up). The Court confines its standing analysis to Count I.

a. Injury/Causation

In a suit for injunctive relief, “past harm is not sufficient to establish an injury in fact.”

Nat’l Whistleblower Ctr. v. HHS, 839 F. Supp. 2d 40, 45–46 (D.D.C. 2012). The plaintiff, rather, must show “a real and immediate — as opposed to merely conjectural or hypothetical — threat of future injury.” Nat. Res. Def. Council v. Pena, 147 F.3d 1012, 1022 (D.C. Cir. 1998) (citation omitted). “[I]n assessing plaintiffs’ standing, [the Court] must assume they will prevail on the merits of their . . . claims.” LaRoque v. Holder, 650 F.3d 777, 785 (D.C. Cir. 2011).

Here, Plaintiffs assert, inter alia, economic injuries stemming from the Secretary’s 2020 approval of the premium requirements in HIP 2.0. See ECF Nos. 54-2 (Declaration of Monte A. Rose, Jr.), ¶ 16; 54-3 (Declaration of Chelsey Lang), ¶¶ 14–15; 54-4 (Declaration of Emily Rames), ¶¶ 13–14; 64 (Supplemental Declaration of Emily Rames). Without that approval, Indiana could not charge Medicaid beneficiaries such as Plaintiffs more than “nominal” premiums. See 42 U.S.C. § 1396o. With it, the State can require enrollees to pay income-based premiums each month, which can be up to 3% of household income (with punishment for non- payment, including in some circumstances termination of coverage). See AR 1557, 1559, 1599.

23

As the Court explained in Stewart I, “This sort of financial loss falls in the heartland of Article III standing.” 313 F. Supp. 3d at 251; see Carpenters Indus. Cncl v. Zinke, 854 F.3d 1, 5 (D.C. Cir. 2017) (“Economic harm . . . clearly constitutes an injury-in-fact.”). For such economic harm, “amount is irrelevant. A dollar of economic harm is still an injury-in-fact for standing purposes.” Zinke, 854 F.3d at 5; see Czyzewski v. Jevic Holding Corp., 580 U.S. 451, 464 (2017) (“[A] loss of even a small amount of money is ordinarily an ‘injury.’”).

While HHS concedes that Plaintiffs have standing, see HHS MSJ at 14 (initially arguing that none of Plaintiffs’ declarations establishes that they will be required to pay premiums to obtain coverage under HIP 2.0); ECF No. 66 (HHS Resp. to Supp. Decl.) at 1 (withdrawing that argument following lodging of Supplemental Declaration), Indiana insists that Plaintiffs do not adequately “allege that they are at actual or imminent risk of disenrollment for failure to pay.” ECF No. 58 (Indiana MTD) at 20 (cleaned up).

To understand the State’s argument, recall how HIP 2.0’s two-tiered benefits system operates. Enrollees who pay their premiums receive HIP Plus (the more desirable plan). Enrollees with incomes at or below 100% of the FPL who neglect to pay their premiums are defaulted to HIP Basic (the less desirable one). Only enrollees with incomes above 100% of the FPL who do not pay their premiums are disenrolled from HIP 2.0 entirely. As Indiana apparently sees it, the only individuals who have standing to challenge the Secretary’s approval of the program on the basis of increased premiums are those who would lose coverage entirely for failure to pay — viz., those with incomes above 100% of the FPL.

To this, Plaintiffs offer two responses. First, they submit that even individuals who “will not suffer the most extreme consequence” — i.e., total loss of coverage — for failure to pay premiums have an injury-in-fact insofar as they “stand to lose coverage for [dental and vision]

24

services” if they are defaulted to HIP Basic. See ECF No. 59 (Pl. Reply) at 4. Second, even if it were true that at least one Plaintiff must establish that she is likely to lose her Medicaid coverage entirely if she fails to pay premiums, Plaintiffs argue that Rames has done so. Id. at 4–5. The Court need not evaluate the first response because it agrees with the second.

Rames is a 29-year-old Hoosier who had a gross income in 2023 of approximately $17,790. See Rames Decl., ¶¶ 1–2. Her Supplemental Declaration confirms that she is unmarried and has no children. See Supp. Rames Decl., ¶¶ 2–3. Her income is, accordingly, above 100% of the FPL. See Annual Update of the HHS Poverty Guidelines, 88 Fed. Reg. 3,424 (Jan. 19, 2023) (setting FPL at $14,580 for single-member households in 2023); see also HHS Resp. to Supp. Decl. at 1 (agreeing that Rames’s “income exceeds the federal poverty level”). That means she is subject to termination from HIP 2.0 for nonpayment of premiums and therefore suffers a concrete injury traceable to the program. See Rames Decl., ¶ 13 (“Given my income, if I do not pay the monthly premium, I will lose my Medicaid coverage.”); see also HHS Resp. to Supp. Decl. at 1. That is all that is needed for these Plaintiffs to challenge the program’s approval. Stewart I, 313 F. Supp. 3d at 252; see Animal Legal Def. Fund, Inc. v. Glickman, 154 F.3d 426, 429 (D.C. Cir. 1998) (in suit brought by multiple plaintiffs, only one must possess standing for case to proceed).

b. Redressability

Having established an injury traceable to HHS’s conduct, Plaintiffs must also show “a likelihood that the requested relief will redress the alleged [harm].” Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 103 (1998) (emphasis added). Generally, courts will find “standing exists where the challenged government action authorized conduct that would otherwise have been illegal.” Renal Physicians Ass’n v. HHS, 489 F.3d 1267, 1275 (D.C. Cir. 2007). “In such

25

cases, if the authorization is removed, the conduct will become illegal and therefore very likely cease.” Id.

There is no doubt that the challenged government conduct here — viz., the Secretary’s 2020 approval — provided the necessary authorization for HIP 2.0. Because that program would otherwise run afoul of Section 1396a’s coverage requirements, Indiana required the Secretary’s approval and waiver authority to continue implementing it past the three-year timeframe approved in 2018. See 42 U.S.C. § 1315; see also 42 C.F.R. § 430.12(c). Should this Court decide that CMS unlawfully approved it in 2020, the program (including each component challenged by Plaintiffs) could no longer be implemented. That is all they need for redressability purposes. See Renal Physicians, 489 F.3d at 1275.

The Government disputes none of this. Indiana, however, retorts that Plaintiffs cannot satisfy the redressability prong because if they prevail in this suit, the State will not continue covering the Medicaid expansion population at all. See Indiana MTD at 20. “Whether Indiana continues to provide Medicaid coverage to the [expansion] population” if HIP 2.0 is invalidated “is its choice to make,” the State insists, and “it has already made that choice.” Id.

The State seems to forget that this Court rejected an analogous argument in Stewart I.

There, Kentucky similarly “argu[ed] that Plaintiffs cannot satisfy the redressability prong because if they prevail in this action, the Commonwealth will not continue participating in expanded Medicaid.” 313 F. Supp. 3d at 252. For support, it pointed to an Executive Order directing the Commonwealth to “unexpand” Medicaid if any aspect of Kentucky HEALTH was invalidated. Id. The Court explained that the Governor’s Executive Order had “no bearing on the standing analysis,” however, because it could not take effect before this Court’s decision. Id. So, “[e]ven if Kentucky were able to ‘unexpand’ Medicaid,” the Court reasoned, the plaintiffs

26

“would have, at minimum, momentary relief.” Id.; see also Stewart II, 366 F. Supp. 3d at 137 (similarly concluding that “[v]acating the approval” on remand “would afford Plaintiffs complete relief”). It is just as true in this case, where there is no equivalent Executive Order, that Plaintiffs would receive at least “momentary relief” from vacatur of the challenged approval.

To be sure, Indiana sees things differently. It believes that under its state law, Hoosiers in the expansion population cannot retain the benefits of HIP 2.0 without the premium requirements. See ECF No. 61 (Indiana Reply) at 3. Specifically, the State cites two provisions: (1) Indiana Code § 12-15-44.5-10(a), which provides that the secretary of the Indiana Family and Social Services Administration (FSSA) can “provide benefits to [the expansion population] only in accordance with this chapter”; and (2) Indiana Code § 12-15-44.5-10(b), which authorizes the secretary of FSSA to “negotiate and make changes to the [Indiana Medicaid] plan,” except that he may not “negotiate or change the plan” to, among other things, reduce premiums. See Indiana Reply at 3. Indiana believes that, as a result of the combination of those provisions, vacatur would necessarily “prevent Indiana from providing coverage to [Plaintiffs] and others in the Medicaid expansion population” full stop, thereby affording Plaintiffs no relief whatsoever. Id.

There are a few problems with the State’s position. To start, vacatur would not automatically terminate all of HIP 2.0 or Plaintiffs’ eligibility for Medicaid, as the State fears. In fact, the Indiana Code explicitly indicates under what circumstances HIP 2.0 “shall” be terminated, and those circumstances do not include the lack of Section 1115 approval or invalidation of such approval by a court. See Ind. Code § 12-15-44.5-4(b). Indeed, as Plaintiffs point out, at least one of the HIP components that state law designates as non-negotiable — the six-month lockout penalties for failure to pay premiums — is no longer approved, yet the State has not terminated HIP 2.0 or otherwise stopped covering the expansion population. See Pl.

27

Reply at 7–8; see also Ind. Code § 12-15-44.5-10(b)(2) (FSSA “may not negotiate or change the plan” to “[r]emove or reduce the penalties for nonpayment set forth in section 4.7”); id. § 12-15- 44.5-4.7(e)(2) (section 4.7 includes six-month lockout penalties); HHS MSJ at 9 n.2 (acknowledging such penalties not approved because Supreme Court never issued decision authorizing them). Because Plaintiffs — like all members of the expansion population — derive their eligibility for Medicaid through Indiana’s state plan, see AR 1588, they would remain Medicaid eligible even if the Secretary’s 2020 approval is vacated, unless and until the State takes additional action to terminate their coverage. See Pl. Reply at 7.

Even if state law did mandate the unexpansion of Medicaid in the face of a court’s invalidation of a Section 1115 approval, moreover, it is far from a foregone conclusion that such a mandate would be lawful. See 42 U.S.C. § 1396a(a)(10)(A)(i) (requiring state to “provide . . . for making medical assistance available” to mandatory populations); id. § 1396a(a)(10)(A)(i)(VIII) (categorizing expansion population as mandatory); Stewart I, 313 F. Supp. 3d at 261 (“Should a state choose to [expand Medicaid], those individuals [in the expansion group] become part of its mandatory population.”); see also NFIB, 567 U.S. at 580–86 (holding that federal government may not compel states to take up Medicaid expansion in first place by threatening to withhold all Medicaid funding, but leaving intact statutory scheme otherwise treating expansion population as mandatory once covered).

What is more, Indiana’s argument loses sight of what Plaintiffs’ lawsuit targets and where expansion-population members derive their Medicaid eligibility from. The challengers take aim not at the premium requirements themselves, but at the Secretary’s 2020 approval of HIP 2.0, which they believe failed to adequately consider the relevant factors. “[T]hose adversely affected by a discretionary agency decision generally have standing to complain that

28

the agency based its decision upon an improper legal ground.” FEC v. Akins, 524 U.S. 11, 25 (1998). Redress, therefore, is not eliminating the premium requirement but vacating the Secretary’s approval so that he may consider the factors he arbitrarily and capriciously ignored. The Court is, accordingly, unpersuaded by the State’s concerns about redress.

c. Relief Sought

Finally, Indiana argues that Plaintiffs “must demonstrate an injury from each aspect” of HIP 2.0 that their Complaint challenges — that is, from not only its premium requirements, but also its lack of retroactive coverage and NEMT. See Indiana Reply at 2. Once again, the Court has already entertained and rebuffed an equivalent argument.

As it explained in Stewart I, “It is true that ‘a plaintiff must demonstrate standing separately for each form of relief sought,’ and ‘for each claim he seeks to press.’” 313 F. Supp. 3d at 253 (first quoting Friends of the Earth, Inc. v. Laidlaw Environ. Servs., Inc., 528 U.S. 167, 185 (2000); and then quoting DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 352 (2006)); see also, e.g., City of Los Angeles v. Lyons, 461 U.S. 95, 109 (1983) (plaintiff must have standing to pursue both damages and injunctive relief). The relevant “claim” pressed here, however, is only Count I, in which Plaintiffs challenge the Secretary’s 2020 approval of HIP 2.0 as a whole. For standing purposes, moreover, the Court must assume that Plaintiffs would prevail on that claim. See Est. of Boyland v. U.S. Dep’t of Agric., 913 F.3d 117, 119 (D.C. Cir. 2019). In other words, it must assume that the Secretary’s approval of the entire program was arbitrary and capricious because he failed to adequately consider an important objective of Medicaid.

The “relief sought” in Count I — viz., vacatur of the Secretary’s approval of the program — is tied to that claim. “Unlike individual sections of a statute or provisions in a regulation, the Court cannot parse the Secretary’s approval of a program” under Section 1115. Stewart I, 313 F.

29

Supp. 3d at 253 (cleaned up). The agency considered HIP 2.0 as a whole before deciding whether to approve it, rather than separately analyzing each challenged component. See HHS MSJ at 20–21; see also Indiana MTD at 28 (acknowledging that “Section 1115 requires CMS to ask whether the ‘project’ as a whole is likely to advance Medicaid’s objectives, not whether individual components are likely to do so”). The Court, accordingly, examines the approval of the project as a whole, too. State Farm, 463 U.S. at 50 (“[A]n agency’s action must be upheld, if at all, on the basis articulated by the agency itself.”). Were the Secretary arbitrary and capricious in approving HIP 2.0, the Court would strike down that approval in toto, as it did with his approvals of Kentucky HEALTH, the Arkansas Works Amendments, and Granite Advantage.

The Court therefore “need ask only whether Plaintiffs have an interest in some portion”

of the benefits affected by [Indiana]’s program.” Stewart I, 313 F. Supp. 3d at 254 (cleaned up). Here the answer is yes, and the premiums are the most concrete interest. For the reasons provided above, vacating the Secretary’s 2020 approval of HIP 2.0 would sufficiently redress the injuries caused by those premiums. Hence Plaintiffs have standing for Count I.

2. Justiciability

HHS next asserts that even if Plaintiffs have standing, this Court cannot review the Secretary’s actions under Section 1115 because they are “committed to agency discretion,” rendering them unreviewable under Section 701(a)(2) of the APA. See HHS MSJ at 18–19. The Court in Stewart I analyzed this issue at length in response to an indistinguishable argument. See 313 F. Supp. 3d at 254–57. Following every other court to consider the question, it concluded that it could review the Secretary’s decisions because “there is some ‘law to apply.’” Id. at 254– 56. On appeal, the D.C. Circuit agreed. Gresham, 950 F.3d at 98 (“Although the Secretary has

30

considerable discretion to grant a waiver, we reject the government’s contention that such discretion renders his waiver decisions unreviewable.”).

Because the Government acknowledges that it is renewing this already vanquished argument merely to “preserve [it] for appellate review,” HHS MSJ at 19, and this Court could not depart from binding Circuit law even if it wanted to (an unlikely circumstance here), it will not spill any more ink on the matter. The approval of HIP 2.0 is judicially reviewable.

B. Merits Issues Preliminaries dispatched, we march on to the merits. The Court begins by outlining the scope of the challenge before it and then reaches the key question: whether the Secretary’s 2020 approval of HIP 2.0 was “arbitrary, capricious . . . , or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). Finding that it was, the Court concludes with the question of the proper remedy.

1. Scope

In addition to limiting its analysis to Count I, as discussed above, the Court, over the State’s protestations, treats the Secretary’s approval of HIP 2.0 as distinct from his approval of Indiana’s SUD/SMI programs. See Stewart I, 313 F. Supp. 3d at 257–58 (similarly treating Kentucky HEALTH as separate from state’s SUD program); contra Indiana MTD at 27.

Even though Indiana submitted the application for the SUD/SMI programs inside the same package as the application for HIP 2.0 in 2020, they are “wholly distinct.” Stewart I, 313 F. Supp. 3d at 258. They cater to different Hoosiers: the former is available to all Medicaid beneficiaries, while the latter serves only adults without disabilities. See AR 1587–90. They have different approval periods: at Indiana’s request, the Secretary approved the SUD/SMI program through 2025, while he approved HIP 2.0 through 2030. See AR 1555. And they have

31

different purposes: HIP 2.0 is “designed to prepare beneficiaries for the personal responsibility required to maintain coverage and continuity of care they will experience when they seek commercial insurance coverage.” AR 1557. The SUD/SMI programs, by contrast, are designed to “ensure that a broad continuum of care” is available to Indiana residents with substance-use disorders, and they “increase identification, initiation, and engagement of Medicaid beneficiaries with SMI.” AR 1560–61; cf. Stewart I, 313 F. Supp. 3d at 258 (noting that Kentucky HEALTH and SUD program had different audiences, start dates, and purposes). In recognition of their distinct identities, the Secretary has solicited and regularly approved stand-alone SUD demonstrations in other states, without packaging in elements similar to HIP 2.0. See, e.g., Letter from Seema Verma, Adm’r, CMS, to Jen Steele, Medicaid Dir., La. Dep’t of Health & Hospitals (Feb 1., 2018), https://perma.cc/886Q-3MX2; Letter from Seema Verma to Cynthia Beane, Comm’r, W. Va. Bureau for Med. Servs. (Oct. 6, 2017), https://perma.cc/V7Y7-JNZ6.

Both Indiana and HHS effectively treated the SUD/SMI programs and HIP 2.0 as separate demonstration projects. See, e.g., AR 8235–87 (State’s 2020 application for HIP 2.0); 8288–315 (State’s separate 2020 application for SUD/SMI programs); AR 8234 (letter from Governor Eric Holcomb to HHS noting that “[i]n addition to authorizing HIP, the existing waiver includes authority to operate a Substance Use Disorder (SUD) demonstration”); AR 1613 –16, 1618–24, 1629–31 (Secretary requiring Indiana to submit separate implementation plans, monitoring protocols, mid-point assessments, and evaluations). Although the Secretary nominally referred to the SUD/SMI programs as “include[d]” in “[t]he ‘HIP demonstration,’” Indiana MTD at 27 (quoting AR 1560), that label does not control. See Stewart I, 313 F. Supp. 3d at 258. Notably, even HHS does not contest before this Court that the projects are separate.

32

This makes sense because the Secretary “can issue only those waivers ‘necessary’ to support the project,” which means he cannot “piggyback other unrelated waivers onto [an] approval” of a SUD or SMI program. Id. (quoting 42 U.S.C. § 1315(a)(1)). In this case, the Secretary determined that no waivers were in fact necessary for Indiana to implement the SUD/SMI programs in 2020. Simply by approving them, he ensured that SUD/SMI costs were treated as reimbursable under Medicaid. See 42 U.S.C. § 1315(a)(2)(A). He did not, moreover, consider whether the waivers that he granted for the State to implement HIP 2.0’s various components — e.g., premiums, lack of retroactive eligibility, and lack of NEMT — were “necessary” to carry out the SUD/SMI programs. See AR 1575–77. Instead, he authorized only those waivers that were “necessary” to carry out HIP 2.0’s features. See AR 1575–77. It is therefore HIP 2.0, not the SUD/SMI programs, that, “in the judgment of the Secretary,” needed to be “likely to assist in promoting the objectives of [the Medicaid statute].” 42 U.S.C. § 1315(a). Otherwise, none of those waivers would be “necessary . . . to carry out such [a] project.” Id. § 1315(a)(1). This Court will, accordingly, consider HIP 2.0 standing alone, as it did with Kentucky HEALTH in Stewart I.

2. Arbitrary & Capricious Review To reward those readers still awake, the Court finally turns to the nub of the parties’

dispute: whether the Secretary acted arbitrarily or capriciously in concluding in 2020 that HIP 2.0 was “likely to assist in promoting the objectives” of the Medicaid Act. See 42 U.S.C. § 1315(a).

In assessing that question, the Court “is not empowered to substitute its judgment for that of the agency.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416 (1971). Nor can it “presume even to comment upon the wisdom of [Indiana’s] effort at [Medicaid] reform.”

33

C.K. v. N.J. Dep’t of Health & Hum. Servs., 92 F.3d 171, 181 (3d Cir. 1996). Still, it is a fundamental principle of administrative law that “agencies are required to engage in reasoned decisionmaking.” Michigan v. EPA, 576 U.S. 743, 750 (2015) (cleaned up); see also Pub. Citizen, Inc. v. FAA, 988 F.2d 186, 197 (D.C. Cir. 1993) (“The requirement that agency action not be arbitrary or capricious includes a requirement that the agency adequately explain its result[.]”). This means that an agency must “examine all relevant factors and record evidence.” Am. Wild Horse Pres. Campaign v. Perdue, 873 F.3d 914, 923 (D.C. Cir. 2017). At minimum, the Secretary cannot “entirely fail[] to consider an important aspect of the problem.” State Farm, 463 U.S. at 43. Rather, he must “adequately analyze the . . . consequences” of his actions. Am. Wild Horse, 873 F.3d at 932. In doing so, “[s]tating that a factor was considered . . . is not a substitute for considering it.” Getty v. Fed. Savs. & Loan Ins. Corp., 805 F.2d 1050, 1055 (D.C. Cir. 1986). The agency must instead provide more than “conclusory statements” to prove it “consider[ed] [the relevant] priorities.” Id. at 1057.

Plaintiffs believe that the Secretary failed to engage in this type of reasoned decisionmaking even under that deferential standard. Their position is simple: as is now very well established, “a central objective of the Medicaid Act [is] to furnish medical assistance to the populations covered by the Act.” ECF No. 54-1 (Pl. MSJ) at 16 (quoting Stewart II, 366 F. Supp. 3d at 139). Yet the Secretary “failed to adequately consider whether HIP [2.0] ‘would in fact help the state furnish medical assistance to its citizens.’” Id. at 20 (quoting Stewart I, 313 F. Supp. 3d at 243). He neither sufficiently assessed how the project would reduce or promote coverage nor contended with comments in the administrative record projecting that HIP 2.0 elements diminish coverage. Id. As such, Plaintiffs say, the Secretary’s decision was arbitrary and capricious.

34

Although his approval is less obviously infirm in this case than in those that preceded it, the Court ultimately sides with Plaintiffs. Its analysis unfolds in two parts. First, it summarizes the now-familiar view that the core objective of the Medicaid Act is to furnish healthcare coverage to the needy and explains why the Secretary neglected to adequately consider that objective here. Second, turning to Defendants’ counterarguments — nearly all of which were addressed at length in Stewart II and again in Philbrick — the Court explains why the agency’s consideration of other asserted Medicaid goals does not alter the outcome.

a. Coverage as Objective of Medicaid Act The Secretary, as outlined above, can only approve demonstration projects that are “likely to assist in promoting the objectives” of the Medicaid Act. See 42 U.S.C. § 1315(a). Before greenlighting a project, he must therefore identify the objectives of the Act and explain why the demonstration is likely to promote them.

In its prior Medicaid cases, the Court held that the provision of medical assistance to beneficiaries — both recipients of traditional Medicaid and members of the expansion population — is the central purpose of the Act, see, e.g., Stewart I, 313 F. Supp. 3d at 260–62, and the Circuit confirmed as much. Gresham, 950 F.3d at 99 (“The district court is indisputably correct that the principal objective of Medicaid is providing health care coverage.”). Because all parties accept that coverage is a core purpose, see Pl. MSJ at 17; HHS MSJ at 21; Indiana MTD at 34, the Court will not repeat its discussion of why that is so but instead directs curious readers to its prior Opinions. See Stewart I, 313 F. Supp. 3d at 260–62; Stewart II, 366 F. Supp. 3d at 138; Gresham, 363 F. Supp. 3d at 176; see also Gresham, 950 F.3d at 99.

Having rightly identified the provision of Medicaid coverage as a key objective, see AR 1555 (“[A]n important objective of the Medicaid program is to furnish medical assistance and

35

other services.”), the agency was required to reasonably explain whether HIP 2.0 would advance or impede that goal. In other words, HHS “needed to consider whether [extending] the demonstration project would be likely to cause recipients to lose coverage and whether it would cause others to gain coverage.” Gresham, 363 F. Supp. 3d at 177. As the Court explains below, this it did not do.

i. Risk to Coverage

Before approving a proposed demonstration or extending an extant one, the Secretary must address whether it negatively affects Medicaid coverage. Here, commenters sounded the alarm bells: HIP 2.0’s components were threatening coverage for low-income individuals, and extending the demonstration for another ten years would have a further coverage-shrinking effect. Citing extensive research from studies conducted over the past two decades, they explained how each provision of the program — namely, the (1) increased premiums, see AR 7194–95, 8096–97, 8196–97, 8201, 8206; see also AR 4764, 5959, 5970; (2) the waiver of retroactive eligibility, see AR 7203–06, 8207, 8213; and (3) the waiver of NEMT coverage, see AR 7162, 7213–16, 8112–13, 8223–29 — reduces coverage, see Appendix A (reproducing comments), especially among Black recipients. See, e.g., AR 8098, 8111–12, 8220 (citing AR 4773–74). Indiana, for its part, had received similar comments from Hoosiers concerned about coverage and noted as much in its application. See, e.g., AR 8273–74.

A plethora of individuals also offered the Secretary their firsthand accounts, some in rather earthy language. See, e.g., AR 8135 (“I was one of the many Hoosiers who lost their [HIP] coverage due to being unable to afford my power account payment. I’m writing today to ask that you eliminate . . . power accounts[] and any other means of confusing red tape that might prevent everyday Hoosiers from needed healthcare.”); AR 8138 (“[T]he POWER

36

accounts . . . form barriers to assuring that Hoosiers have health insurance. The complex paperwork associated with these aspects of HIP have cause[d] people that I know to be confused, to loose [sic] their health insurance, or to give-up on applying for coverage under HIP.”); AR 8142 (“[I’ve] seen 1st hand the many harms the power account has done to those who are trying to both link to and retain health care as well ensure medication compliance and hold a job. Get rid of the fucking power account[.]”).

Those concerns were further “factually substantiated” by the data from various evaluations of HIP 2.0. Humane Soc’y of United States v. Zinke, 865 F.3d 585, 606 (D.C. Cir. 2017); see AR 4602–930 (2020 interim report); AR 5955–6026 (2017 premiums assessment); AR 6027–93 (2016 NEMT waiver evaluation). For example, between February 2015 and November 2016, almost 10,000 individuals with incomes above 100% of the FPL were disenrolled from HIP Plus for nonpayment of their premiums, and another almost 4,000 were disenrolled from HIP Basic for the same reason. See AR 5959. During that same period, an additional nearly 48,000 Indianans were found eligible for HIP 2.0 but never enrolled because they did not make their first premium payment. Id. In other words, nearly 60,000 Hoosiers — a whopping 29% of all Hoosiers subject to premiums — were disenrolled from or never enrolled in HIP 2.0 because of non-payment. Id. And over half of all beneficiaries, or 324,840 Indiana residents, who were required to pay premiums missed at least one payment in 2015–16. See AR 5970. Loss of coverage persisted in 2017–18, the evaluation shows: 26,037 HIP Plus enrollees were disenrolled for failure to pay. See AR 4764.

While the Secretary was not required to address each comment (or each piece of evidence) in writing, he concedes that he needed to at least “review and consider all” of them.

37

See HHS MSJ at 18 (quoting 42 C.F.R. § 431.416(d)(2)). Instead of doing so, he acknowledged the existence of the concerns as to each of the components but largely looked the other way.

Beginning with premiums, the Secretary remarked that “commenters indicated that premium amounts were burdensome to beneficiaries and would prevent individuals from maintaining coverage.” AR 1570. He declined, however, to explain why he disagreed with that assessment or why, if he agreed with it, he nonetheless thought HIP 2.0 was coverage promoting. Instead, he reiterated that the only beneficiaries who could be disenrolled from HIP 2.0 for nonpayment of premiums are those with incomes over 100% of the FPL, and he observed that “the number and proportion of individuals disenrolled due to non-payment [has] decreased over time.” AR 1570. True as those points may be, the former is of little consolation to the tens of thousands of Hoosiers with incomes between 100 and 133% of the FPL — who are part of Medicaid’s mandatory population yet subject to disenrollment for failure to pay premiums — and the latter is of no comfort at all to the Indiana residents who have lost their coverage to date.

As Plaintiffs observe, moreover, the Secretary did not engage with the many comments asserting that the decline in disenrollment could be explained by confounding factors, including the increase in the number of beneficiaries determined medically frail, and flagging that the 2020 evaluation did not include the number of individuals who were found eligible but never enrolled in HIP given their failure to pay premiums. See AR 7196, 8219–20. What is more, because the core objective of Medicaid is coverage promotion, the agency does not get very far when it concedes that a Section 1115 project causes coverage losses — even when it insists that only a subset of beneficiaries suffers those losses and that the number of losses has been decreasing over time.

38

The Secretary also noted in his approval that he can suspend implementation of the premium requirements if “monitoring or evaluation findings indicate substantial and sustained directional change inconsistent with state targets.” AR 1564. That, as the challengers point out, makes “one wonder[] what findings would prompt the Secretary to order a suspension” in light of “the substantial and sustained coverage loss that has already occurred.” Pl. MSJ at 22. All the more so given that, several days before the approval, the State told him explicitly that it “anticipat[ed] a substantial increase in disenrollment” for failure to pay premiums after the COVID emergency ended, and the Secretary nonetheless issued the approval. See AR 3302.

As a final attempt to address the issues with premiums, the Secretary made a move reminiscent of one in Stewart I, Stewart II, and Philbrick: he asserted that Indiana has taken steps to minimize coverage loss stemming from premiums — namely, exempting pregnant women and medically frail individuals from disenrollment for lack of payment and placing an aggregate cap on premiums. See AR 1565–66, 1570–71. In each of the prior cases where the Secretary pointed to similar guardrails, the Court found his response to be “no answer at all” since the guardrails were part of the original Section 1115 applications and thus baked into the commenters’ concerns. Stewart I, 313 F. Supp. 3d at 263; see Stewart II, 366 F. Supp. 3d at 142–43; Philbrick, 397 F. Supp. 3d at 25. The response he offers here is even less of an answer because the relevant guardrails — i.e., exemptions and an aggregate cap — were not only part of Indiana’s application but have also long been part of HIP, including when an astounding number of individuals faced disenrollment. See Pl. MSJ at 21–22.

As to HIP 2.0’s restrictions on retroactive coverage, the agency’s analysis fares no better.

This Court has twice explained that “restricting retroactive eligibility will, by definition, reduce coverage.” Stewart II, 366 F. Supp. 3d at 143 (quoting Stewart I, 313 F. Supp. 3d at 265); see

39

HHS MSJ at 25 (conceding that “waiving retroactive coverage by necessity eliminates some coverage”). The Secretary recognized that “[a] few commenters expressed concerns that the waiver of retroactive eligibility will . . . reduce coverage.” AR 1571. But rather than tackling those concerns, he simply stated that the demonstration was “designed to prepare beneficiaries for commercial insurance and to improve the uptake of preventative services, thus improving beneficiary health.” AR 1571. Fair enough, but that has nothing to do with coverage. He also noted that Indiana would “continue to provide outreach and education about how to apply for and receive HIP coverage,” presumptive eligibility, and a “Fast Track” enrollment option. See AR 1564–65. Once again, those longstanding features of HIP are beside the point. Cf. AR 7204 (describing why presumptive eligibility is not a substitute for retroactive coverage); AR 8094–95 (noting widespread confusion about eligibility and the Fast Track option).

The story is essentially the same for HIP 2.0’s restrictions on NEMT. All the Secretary offered by way of response to the comments highlighting that NEMT elimination reduces healthcare access and utilization was that the waiver applies only to the expansion population and that the State exempts pregnant women, the medically frail, and certain parents, and caretakers. See AR 1572. He expressed his belief that “this approach adequately addresses commenters’ concern, as it minimizes the impact on vulnerable beneficiaries while also achieving the state’s goal of recreating the experience of commercial insurance market, which does not offer the NEMT benefit.” AR 1572. But those exemptions are nothing new, and the Secretary “cannot limit his review to only ‘vulnerable individuals’”; he must also consider how the project affects all enrolled groups. Stewart I, 313 F. Supp. 3d at 263–64.

Indiana advances two arguments as to why the Secretary adequately considered the risk to coverage, but both are more properly attended to in subsequent sections. In a nutshell, it

40

contends that HIP 2.0 does not decrease coverage because it allows the State to cover a population it would not otherwise cover — namely, the expansion group. The Court will analyze this argument in the ensuing section on whether the Secretary adequately considered if the project would promote coverage. See infra Section III.B.2.a.ii. The State also maintains that any coverage loss is outweighed by the project’s promotion of other purposes of Medicaid, including health outcomes and fiscal sustainability. This point will be discussed below in Section III.B.2.b.

ii. Promote Coverage

Unlike in the prior Medicaid cases, the Secretary devoted a subsection of his approval to explaining why “[t]he demonstration will expand coverage.” AR 1560–61. Sounds promising. The problem? That subsection speaks only of the SUD/SMI programs, which are separate from HIP 2.0. It says nothing whatsoever about how HIP 2.0 generally — or its premium requirements, lack of retroactive eligibility, or lack of NEMT specifically — promote coverage. Even if the SUD/SMI programs were part of HIP 2.0, moreover, that subsection did not discuss whether Indiana’s program would “on balance” promote coverage and, if so, why or how. Stewart I, 313 F. Supp. 3d at 265. The Secretary did not, for instance, suggest that providing SUD/SMI treatment would offset (much less justify) the loss of coverage resulting from HIP 2.0’s other features. Nor did he explain how the SUD/SMI programs furnish medical assistance to the tens of thousands of Hoosiers who, given the premium requirements, cannot afford to enroll in HIP 2.0 in the first place or will enroll and later be disenrolled for nonpayment.

In addition to that subsection, the agency sprinkled references to “coverage” throughout the approval letter. The Government tries to pluck those references out of context to show that “CMS’s coverage determinations were not limited to [the SUD/SMI programs].” HHS MSJ at 21. To start, it points to the Secretary’s statement that HIP 2.0 created “incentives for individuals

41

to enroll as soon as possible and to obtain preventive services.” Id. (quoting AR 1561). The thinking seems to be that eliminating retroactive eligibility would encourage Hoosiers to obtain and maintain coverage, even when they are healthy, thereby increasing coverage. If that rings a bell, it is because the Secretary said the same thing about the waiver of retroactive eligibility in his approvals of Kentucky’s, Arkansas’s, and New Hampshire’s programs. See Stewart I, 313 F. Supp. 3d at 265; Gresham, 363 F. Supp. 3d at 179; Philbrick, 397 F. Supp. 3d at 25. As the Court explained in those cases, such a “conclusory” reference cannot suffice, “especially when viewed in light of” an obvious counterargument: as is conveyed in the comments, restricting retroactive coverage quite obviously — tautologically, even — reduces coverage. Stewart I, 313 F. Supp. 3d at 265 (quoting Getty, 805 F.2d at 1057).

The possibility that eliminating retroactive coverage will galvanize Indianans to seek coverage earlier, moreover, has nothing to do with the coverage loss that the evidence shows that the premium requirements cause. Insofar as the Secretary believes that eliminating retroactive coverage might (counterintuitively) increase coverage, he needed to weigh the promotion side of the scale against the risk-of-loss side when approving the project. See Philbrick, 397 F. Supp. 3d at 25 (same where incentive possibility “ha[d] nothing to do with” loss that community- engagement requirement would cause). That he did not do. Indeed, he did not even purport to offer this potential incentive as a reason why HIP 2.0 would increase coverage; he offered it as a reason why the “demonstration will . . . improve[] the sustainability of the safety net,” AR 1561, a separate rationale discussed in Section III.B.2.b.ii, infra.

Still seeking to paint a coherent picture of “coverage-promotion rationales” in the approval letter, the Government next claims that “CMS . . . determined that the HIP demonstration . . . promoted ‘continuity of coverage and care.’” HHS MSJ at 21 (quoting AR

42

1564). That is a creative reading of the approval at best and a misrepresentation at worst. Cf. Pl. Reply at 17–18. In fact, the approval stated that “[t]o increase awareness of the waiver of retroactive eligibility and promote the objectives of the Medicaid program (e.g., continuity of coverage and care), Indiana will continue to provide outreach and education about how to apply for and receive HIP coverage to the public and to Medicaid providers.” AR 1564 (emphasis added). At the risk of stating the obvious: noting that the State will continue undertaking efforts to promote coverage to counterbalance an inherently coverage-reducing feature of HIP 2.0 — i.e., the waiver of retroactive eligibility — is not the same as determining that the project itself promotes coverage. This snippet, moreover, comes from the Secretary’s explanation of why the demonstration would “test[] reforms designed to promote financial independence,” not why it would increase coverage. See AR 1563.

Beyond the stray references to “coverage” in the approval, Defendants offer three substantive arguments about coverage promotion. First, they submit that HIP 2.0 promotes coverage because HIP Plus covers optional services not required under the Medicaid Act — viz., vision, dental, and chiropractic services. See Indiana MTD at 26–28; HHS MSJ at 22–24. More services = more coverage, or so the thinking seems to go. HHS points out that its “argument on this front is distinct from the one this Court rejected previously.” ECF No. 62 (HHS Reply) at 10. That may be so, but little needs to be said to dispense with it.

The Secretary’s obligation was to determine whether HIP 2.0, as a whole, would aid the State in furnishing medical assistance to the ACA’s mandatory populations, which includes the expansion population. Even if he could have reasonably concluded that HIP promotes that objective based on balancing the fact that HIP 2.0 furnishes more medical assistance than is required to some members of the expansion population with the fact that it furnishes no medical

43

assistance whatsoever to other members of that population (i.e., those who cannot afford to enroll or are disenrolled for failure to pay their premiums), he did not engage in the requisite balancing here. See, e.g., AR 1562 (noting that program’s challenged features would “help Indiana to continue to cover non-mandatory benefits,” without further analysis); contra HHS Reply at 10. And, as Plaintiffs point out, the fact that a demonstration offers optional services to some subset of the mandatory populations obviously does not automatically render it coverage promoting. Were it otherwise, the Secretary could waive the majority of a given state’s Medicaid Act’s requirements so long as that state offered, say, chiropractic services to some subset of beneficiaries. See Pl. Reply at 17 n.3. That is not how Section 1115 is supposed to work.

Second, Indiana advances an argument that the Court addressed in depth in Stewart II and again in Philbrick: because the State would “simply de-expand Medicaid” if HIP 2.0 was not approved, any coverage provided to the expansion population through the demonstration is properly understood as increasing Medicaid coverage. Stewart II, 366 F. Supp. 3d at 153; Philbrick, 397 F. Supp. 3d at 25; see Indiana MTD at 29 (“Clearly, providing expanded Medicaid coverage to the more than 570,000 Hoosiers covered in 2020 . . . is superior to not expanding coverage at all[.]”); see also Indiana MTD at 24–26, 28. “A demonstration that shrinks coverage may thus be coverage promoting for the purposes of § 1115,” Defendants’ argument goes, “as long as the state threatens that if the demonstration is not approved, it will discontinue coverage entirely.” Stewart II, 366 F. Supp. 3d at 153. As the Court explained in both prior cases, there are three interrelated reasons why this position is unconvincing.

For starters, it incorrectly assumes that “a state has additional discretion to diminish or condition eligibility for the expansion — as opposed to the traditional — population.” Id. Citing NFIB, Indiana urges that it has the prerogative to de-expand. See, e.g., Indiana MTD at 26. That

44

may (or may not) be true, see Section III.A.1.b, supra, but the privilege the State seeks to exercise here is not to de-expand, but rather to implement the ACA expansion as an à la carte exercise, picking and choosing which of Congress’s mandates it wishes to implement. Nothing in NFIB sanctions that. Rather, the Supreme Court was crystal clear: “Nothing in [NFIB] precludes Congress from offering funds under the Affordable Care Act to expand the availability of health care[] and requiring that States accepting such funds comply with the conditions on their use.” 567 U.S. at 585. The case held only that Congress was “not free . . . to penalize States that choose not to participate in [the expansion] by taking away their existing Medicaid funding.” Id. In other words, the Court held that, as with traditional Medicaid, Congress may impose requirements on the states for the use of expansion funds. Nothing in that analysis allows for “additional discretion” in how the states comply with Medicaid requirements for the expansion population as compared to the traditional one.

While Indiana is wrong to posit that its treatment of the expansion population is undergirded with any greater discretion than its administration of any other part of the Medicaid program, its arguments about flexibility vis-à-vis the expansion population are ultimately a red herring. That is because the entire Medicaid program is optional for states. If the State is correct that threats to terminate the expansion program can supply the baseline for the Secretary’s Section 1115 review, the Court does not see why that argument would not be equally good as applied to traditional Medicaid. The argument must thus posit that any Section 1115 program that maintains any coverage for any set of individuals promotes the objectives of the Medicaid Act as long as the state threatens to terminate all of Medicaid in the absence of waiver approval. The second defect in Indiana’s position, then, is that it has no limiting principle.

45

Indeed, it would also boundlessly expand HHS’s approval authority, as the Court noted in previous Medicaid cases raising variants of Indiana’s argument. See Stewart II, 366 F. Supp. 3d at 153–54; Philbrick, 397 F. Supp. 3d at 26–27. To explain more fully: under the State’s reasoning, states may threaten to de-expand, or indeed do away with all of, Medicaid if HHS does not approve whatever waiver of whatever Medicaid requirements they wish to obtain. The Secretary could then always approve those waivers, no matter how arcane the eligibility conditions or how few individuals in the expansion population it would cover, because “any waiver would be coverage promoting compared to a world in which the state offers no coverage at all.” Stewart II, 366 F. Supp. 3d at 154. This reading of the Act would give HHS practically unbridled discretion to implement the Medicaid Act as “an à la carte exercise, picking and choosing which of Congress’s mandates it wishes to implement.” Id. at 153–54. In addition to the constitutional concerns such an interpretation may raise, cf. Clinton v. City of New York, 524 U.S. 417, 440–47 (1998), it constitutes “an impermissible construction of the statute . . . because [it] is utterly unreasonable in [its] breadth.” Philbrick, 397 F. Supp. 3d at 27 (quoting Aid Ass’n for Lutherans v. U.S. Postal Serv., 321 F.3d 1166, 1178 (D.C. Cir. 2003)).

The last, and perhaps most important, reason to reject Indiana’s position is that it is inconsistent with the text of the statute. Section 1115 requires the Secretary to evaluate whether the project “is likely to assist in promoting the objectives” of the Act. See 42 U.S.C. § 1315(a). Against what baseline is he supposed to evaluate the project? The structure of the waiver provision assumes the implementation of the Act. The overarching provision authorizing these waivers stipulates that, if the Secretary makes a judgment that a demonstration promotes the objectives of the Act, he may then waive compliance with certain provisions “to the extent and for the period . . . necessary” to carry out the project. See 42 U.S.C. § 1315(a), (a)(1). It

46

confirms that the relevant baseline is whether the waiver will promote the objectives of the Act as compared to compliance with the statute’s requirements, “not as compared with a hypothetical future universe” where the Act has no force. Stewart II, 366 F. Supp. 3d at 154; Philbrick, 397 F. Supp. 3d at 27. That is, the provision contemplates a limited waiver from compliance with the Act’s provisions.

The State has a couple of baseline-related ripostes. For one, it maintains that it is uniquely situated because its state law “prohibit[s] the continuation of Medicaid expansion . . . except through HIP,” as it reminded the Secretary in its renewal request. See Indiana MTD at 26 (quoting AR 8240). That argument fails twice over: once for the reasons offered above in the context of the redressability analysis, see Section III.A.1.b, supra, and again because the Secretary did not approve the project on the basis that, without it, state law would mandate de- expansion. See Michigan, 576 U.S. at 758 (“[A] court may uphold agency action only on the grounds that the agency invoked when it took the action.”); see also Pl. Reply at 16.

For another, Indiana points to Georgia v. Brooks-LaSure, 2022 WL 3581859 (S.D. Ga.

Aug. 19, 2022), in which the Southern District of Georgia found that HHS had erred in measuring Georgia’s demonstration project “against a baseline of full Medicaid expansion, rather than taking the demonstration on its own terms.” Id. at *13–15; see Indiana Reply at 10. That court accepted Georgia’s argument that, when a state has not already expanded Medicaid but instead proposes a Section 1115 demonstration project as part of an expansion in the first instance, the “appropriate point of comparison to determine if it further[s] the objectives of Medicaid [i]s a world without [the demonstration] (i.e., no expansion at all), not a hypothetical world of condition[-]free expansion.” Brooks-LaSure, 2022 WL 3581859, at *14 (cleaned up). While appreciating Brooks-LaSure’s support for the State’s position, this Court is not persuaded

47

to alter its analysis of the issue initially articulated in Stewart II, 366 F. Supp. 3d at 154. Particularly in light of what the Court has discussed above, understanding the baseline as full implementation of the Act is the only way that Section 1115 makes sense.

That brings us to Defendants’ third and final coverage-promotion argument. Relying on a federal evaluation of HIP 2.0’s effectiveness between 2016 and 2018, which HHS cited nowhere in the approval, Indiana asserts that “[t]he evidence is unequivocal that HIP serves an important objective of the Medicaid program by increasing the persons covered.” Indiana MTD at 25 (cleaned up). For support, it regurgitates a string of favorable quotes from the evaluation that seem to imply as much, including: AR 4308 (the “first and most important lesson learned from Indiana’s Section 1115 demonstration” is that it “resulted in a program that achieved a key goal of both the ACA and the state — a significant expansion in health insurance coverage”); AR 4281 (“[T]he demonstration increased coverage to more low-income individuals and increased access to health care services.”); AR 4288 (“HIP 2.0 led to a significant increase in health insurance coverage in Indiana.”); AR 4296 (“Childless adults in Indiana experienced significant gains in coverage.”); AR 4308 (“[H]ealth insurance coverage in Indiana was significantly higher[.]”).

Those are great soundbites when taken out of context, but they do not help the State’s cause. Why? Because they almost all come from portions of the evaluation assessing coverage levels under HIP 2.0 relative to what coverage would have been if the State had not expanded Medicaid at all. See, e.g., AR 4288 (“[H]ealth insurance coverage for childless adults increased significantly more in Indiana than what would have been expected if Indiana had not expanded Medicaid.”); AR 4296 (“Childless adults in Indiana experienced significant gains in health insurance coverage . . . relative to the changes for childless adults in similar comparison states

48

that did not expand Medicaid[.]”); AR 4308 (“[H]ealth insurance coverage in Indiana was significantly higher than what would have been expected if Indiana had not expanded Medicaid.”) (emphases added). But, as explained above, the relevant question under Section 1115 is not whether a demonstration project will increase coverage as compared to unexpanded Medicaid (the answer to that would, it seems, always be yes), but whether the project will increase coverage as compared to expanded Medicaid with no waivers. All that the State’s preferred evaluation had to say on that score is that “under HIP 2.0 Indiana achieved comparable gains in coverage in 2017–18 as would have been expected if Indiana had expanded Medicaid without a demonstration.” AR 4288. That is not exactly persuasive evidence that HIP 2.0 is coverage promoting.

* * *

At bottom, the Secretary’s coverage analysis was a step up from the previous ones involving other states, but he still wins no cigar. Although he acknowledged coverage as an objective and paid lip service to worries about the demonstration negatively affecting it, “[s]tating that a factor was considered . . . is not a substitute for considering it,” Getty, 805 F.2d at 1055, and “[n]odding to concerns raised by commenters only to dismiss them in a conclusory manner is not a hallmark of reasoned decisionmaking.” Gresham, 950 F.3d at 103. The Court cannot sustain his approval on a coverage-based rationale.

b. Other Objectives of Medicaid Act Shifting gears, Defendants offer another pair of rationales. They contend that the Secretary reasonably approved HIP 2.0 on the ground that it would likely advance two other Medicaid objectives: the health of beneficiaries and the fiscal sustainability of the safety net. See Indiana MTD at 34 (proffering both); see also HHS MSJ at 21–22 (proffering fiscal

49

sustainability). Neither saves the day: the former is not a stand-alone goal of Medicaid, and the Secretary’s consideration of the latter was deficient.

i. Beneficiary Health

According to Indiana, HIP 2.0 is independently justified because it improves the health of Medicaid beneficiaries by “incentivizing beneficiaries to take personal responsibility for their health and seek preventative care.” Indiana MTD at 34. Although HHS does not advance such an argument before this Court, the Secretary expressed a similar view in his approval. See AR 1557 (“Our demonstration authority under section 1115 of the Act should be interpreted to allow us to offer states more flexibility to experiment with different ways of improving health outcomes.”); see also AR 1556 (“[S]ection 1115 demonstration projects should present an opportunity for states to experiment with reforms that . . . focus on interventions that drive better health outcomes and quality of life improvements.”).

Unfortunately for Defendants, this Court and — more importantly — the D.C. Circuit have already concluded that health is not a freestanding objective of the Medicaid Act. Stewart I, 313 F. Supp. 3d at 266–267; Stewart II, 366 F. Supp. 3d at 144; Philbrick, 397 F. Supp. 3d at 28–29; Gresham, 950 F.3d at 1001. In the Circuit’s words, “While furnishing health care coverage and better health outcomes may be connected goals, the text specifically addresses only coverage.” Gresham, 950 F.3d at 1001. And agencies are “bound, not only by the ultimate purposes Congress has selected, but by the means it has deemed appropriate, and prescribed, for the pursuit of those purposes.” Id. (quoting MCI Telecomms. Corp. v. Am. Tel. & Tel. Co., 512 U.S. 218, 231 n.4 (1994)); see also Waterkeeper All. v. EPA, 853 F.3d 527, 535 (D.C. Cir. 2017); Colo. River Indian Tribes v. Nat’l Indian Gaming Comm’n, 466 F.3d 134, 139–40 (D.C. Cir. 2006). “The means that Congress selected to achieve the objectives of Medicaid was to

50

provide health care coverage to populations that otherwise could not afford it.” Gresham, 950 F.3d at 101.

The State nonetheless insists that health must be a stand-alone objective because “[t]he federal government and States do not spend money on Medicaid simply for the sake of spending money,” but for the sake of improving health. See Indiana MTD at 34. While this position offers some surface appeal, it is the same argument that the Secretary proposed and this Court found unpersuasive in Philbrick. See 397 F. Supp. 3d at 28 (“[T]he agency persists in the contrary view on the ground that there is little intrinsic value in paying for services if those services are not advancing the health and wellness of individuals receiving them. This position does not change the Court’s mind.”) (cleaned up). As explained there, if that position were correct, the Secretary could “justify actions as consistent with the law even if they are clearly at odds with it” simply “[b]y defining a statute’s purposes up a level of generality.” Id. That is not what Congress intended. And ignoring the fact that “a statute’s objectives are often bound up with the way Congress sought to solve a particular problem . . . grants largely unbounded discretion to agencies, whose exercise of that discretion can veer far afield from anything resembling the statute Congress wrote.” Id. Because improving health outcomes is not a freestanding purpose of the statute, the Secretary’s consideration of it cannot support his Section 1115 analysis.

Even if it were, the agency’s analysis of it fell flat. As with the second Kentucky approval and the (only) New Hampshire approval, commenters emphasized significant negative effects that they believed HIP 2.0 would have on health outcomes. See, e.g., AR 7210–13, 7199– 02, 8228–29; see also AR 7216–18; AR 7197–98; AR 8212; AR 8156–57. Without contending with those comments, HHS simply asserted that HIP 2.0 would improve beneficiary health by

51

promoting financial independence. See AR 1563–65. It neither “consider[ed] the health benefits of the project relative to its harms to the health of those who might lose their coverage,” Stewart II, 366 F. Supp. 3d at 145, nor weighed the net effects on health against the effects on coverage more generally. Indiana’s argument that HIP 2.0 is supported by its promotion of health alone is therefore without merit.

ii. Fiscal Sustainability With the cards in their hands dwindling, both sets of Defendants think that they have a final ace up their sleeves: fiscal sustainability. They argue that HHS reasonably concluded that HIP 2.0 furthers the safety net’s sustainability, which they believe to be an independent Medicaid objective. As HHS put it in the approval, “[S]ection 1115 demonstration projects should . . . provide an opportunity for states to test policies that ensure the fiscal sustainability of the Medicaid program, better ‘enabling each [s]tate, as far as practicable under the conditions in such [s]tate’ to furnish medical assistance, . . . while making it more viable for states to furnish medical assistance to a broader range of persons in need.” AR 1556 (quoting 42 U.S.C. § 1396- 1) (emphasis added); see HHS MSJ at 21–22. To that end, it posited that HIP 2.0 would “furnish medical assistance in a manner that improves the sustainability of the safety net.” AR 1561. Indiana concurs. See Indiana MTD at 34–35.

The Government similarly proposed fiscal sustainability as an objective in Stewart II and Philbrick. As the Court explained in the former, “Given that the Act stipulates mandatory floors for benefits and coverage populations,” see 42 U.S.C. §§ 1396a, 1396a(a)(10)(A), the practicability qualifier on which they rely “is as easily understood as establishing a ceiling as it is lowering the floor.” Stewart II, 366 F. Supp. 3d at 149. Applying the Chevron framework, however, it found that the word “objectives” as it appears in Section 1115 is ambiguous. Id.

52

(citing Chevron U.S.A., Inc. v. Nat’l Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984)). At step two, it could not find the Secretary’s interpretation unreasonable, leading it to conclude that the Secretary may “take into account fiscal sustainability in determining under § 1115 whether a demonstration project promotes the objectives of the Act.” Id. Although Chevron’s future is uncertain, no party asks the Court to revisit that conclusion. Instead, Plaintiffs assert that even assuming arguendo that fiscal sustainability is a valid consideration in a Section 1115 project, the Secretary’s explanation for why HIP 2.0 promoted it was arbitrary and capricious. See Pl. MSJ at 20. The Court agrees.

To begin, although the Secretary need not “quantify some exact amount” of savings that he expects, “he must make some finding that supports his conclusion that the project” supports fiscal sustainability. Stewart II, 366 F. Supp. 3d at 149–50; see also Indiana Reply at 13–14 (CMS need not “provide precise ‘per person cost[]’ estimates” or display “mathematical precision”). The Secretary’s first error here is that he “made no finding that [HIP] would save [Indiana] any amount of money or otherwise make the program more sustainable in some way.” Stewart II, 366 F. Supp. 3d at 149; see Philbrick, 397 F. Supp. 3d at 29–30. Instead, he conclusorily asserted that HIP 2.0’s lack of retroactive coverage would “reduce the cost of providing Medicaid coverage to beneficiaries . . . by reducing the incidence of chronic or preventable conditions, and by helping to ensure chronic conditions are well managed,” and its waiver of NEMT would “enabl[e] the state to better contain Medicaid costs and more efficiently focus resources on providing accessible and high-quality health coverage.” AR 1561–62; see also AR 1572 (stating that waiver of NEMT “is expected to improve the fiscal sustainability of the state’s safety net”).

53

Never mind that the evidence in the record suggested that the cost of charging enrollees premiums often exceeds the amount of the premiums collected, rendering a program like HIP 2.0 expensive to administer. See AR 7201. Or that the State nowhere justified the extension of its program on cost-saving grounds or offered any financial information whatsoever, meaning that it would have been near impossible for HHS to perform any calculations. See AR 8235–87. Or that five days before the approval, the State explicitly told the Government that it was “not waiving NEMT for cost savings,” but rather “the purpose [was] to align with commercial market plan designs.” AR 3300; see also AR 7215, 8113, 8224–25 (NEMT is cost-effective); cf. Philbrick, 397 F. Supp. 3d at 29 (“The glaring disconnect between the Secretary’s position and [the state]’s raises substantial questions about how the agency came to believe the program would improve the State’s fiscal circumstances, underscoring the need for reasoned analysis of this issue.”). The issue is why, in the face of such evidence, the Secretary thinks HIP 2.0 is cost saving. The answer is anyone’s guess, for the approval letter is devoid of clues.

The agency rejoins that it made a finding that “HIP’s cost savings meant that Indiana could provide ‘additional benefits’ through HIP Plus that were ‘non-mandatory,’ such as ‘dental’ and ‘vision benefits,’ and that Indiana could provide those benefits to more people.” HHS MSJ at 22 (quoting AR 1561–62); see also AR 1556 (stating HIP was likely to make “it more viable for states to furnish medical assistance to a broader range of persons in need” or provide “additional benefits to existing beneficiaries”). That rejoinder misses the mark: it assumes that HIP has “cost savings” of which to speak, but one of the issues is that the Secretary never explained his determination that such savings exist in the first place. “[W]ithout a finding about the savings that [HIP 2.0] could be expected to yield[,] the Secretary could not make a reasoned

54

decision” that it would enable the State to offer benefits to more people or otherwise “promote fiscal sustainability.” Stewart II, 366. F. Supp. 3d at 150. The Court sees no such finding here.

The point, of course, is not that it was impossible on this record for the Secretary to reasonably conclude that HIP 2.0 would prove cost efficient. Indeed, the Court does not pass on the persuasiveness of the evidence. The point is that, while the Secretary “may well have arrived at a different conclusion” than the commenters and even the State (with respect to NEMT), “he needed to explain how he got there.” Philbrick, 397 F. Supp. 3d at 30. Merely “stat[ing] that the waiver would improve HIP’s sustainability,” even “repeatedly,” will not do. See Indiana Reply at 12 (citing AR 1562, 1564, 1571–72).

The second shortfall in the Secretary’s analysis was his failure to “compare the benefit of savings to the consequences for coverage.” Stewart II, 366 F. Supp. 3d at 150; see Philbrick, 397 F. Supp. 3d at 30. He did not, for instance, consider the burden that HIP 2.0 would impose on Medicaid recipients and explain why such imposition was necessary under the State’s financial circumstances. Maybe realizing as much, the State asks the Court to connect the dots for the agency. “Any Medicaid administrator knows services cost money,” it assures the Court. See Indiana Reply at 13. “If it were costless to eliminate contributions from POWER Accounts, provide retroactive coverage, or assure nonemergency medical transportation, one wonders why anyone — plaintiffs included — would care about this case.” Id. That will not suffice.

Both Supreme Court and D.C. Circuit caselaw “make eminently clear that a project that enhances financial sustainability may not advance the objectives of Medicaid if it significantly impedes or curtails Medicaid services or coverage.” Stewart II, 366 F. Supp. 3d at 152; see Pharm. Rsch. & Mfrs. of Am. v. Walsh, 538 U.S. 644, 664–65 (2003) (“The fact that the [challenged program] may serve Medicaid-related purposes, both by providing benefits to needy

55

persons and by curtailing the State’s Medicaid costs, would not provide a sufficient basis for upholding the program if it severely curtailed Medicaid recipients’ access to prescription drugs.”) (emphasis added); Pharm. Rsch. & Mfrs. of Am. v. Thompson, 362 F.3d 817, 824–26 (D.C. Cir. 2004) (upholding HHS’s decision to impose minor restriction on access to prescription drugs because agency adequately considered and explained necessity of restriction and pointing out “absence of any demonstrable significant impediment to Medicaid services from [challenged] requirement”). As explained in Stewart II, “That there are limits on the extent to which fiscal sustainability can justify cuts like those outlined in these cases makes sense. Most cuts to Medicaid services would reduce the cost of Medicaid and thus advance the sustainability of the program to some extent. But it would be nonsensical to conclude that any cut therefore always promotes the Act’s objectives.” 366 F. Supp. 3d at 152; see Pl. MSJ at 18–19.

In sum, the Court concludes that the Secretary’s determination that HIP 2.0 advances the purposes of Medicaid because it would improve the fiscal sustainability of Indiana’s Medicaid program is arbitrary and capricious. In so concluding, the Court does not question the agency’s predictive judgments or evaluate the evidence before it on this issue. It has simply looked for — and been unable to find — what the APA requires: a reasoned explanation that considers the factors relevant to the agency’s decision.

For the avoidance of doubt, the Court reiterates what it said in Stewart II: in finding the Secretary’s position unreasonable, it does not suggest that the agency may never consider the fiscal sustainability of the Medicaid program. States do not have limitless finances to cover healthcare, and the Secretary very well might properly assess whether a more efficient way of administering a state’s Medicaid program would save resources or whether, as in Thompson, a state might save money by continuing to deliver mandatory care to mandatory populations while

56

restricting precisely which kinds of tests or medications are available, for example. Those considerations are not incompatible with the prime objective of the Act. But that is, once again, not the exercise the Secretary engaged in here.

3. Remedy

Unable to conclude that the Secretary adequately evaluated whether HIP 2.0 was likely to promote the objectives of the Medicaid Act in issuing his 2020 approval, the Court holds that such approval was invalid in toto. Because that approval was necessary for the State to implement each of the challenged components, the Court need not tackle Plaintiffs’ alternative bases for vacating some or all components — e.g., that the Secretary did not adequately examine if the extension was “experimental”; that the length of the extension violates Section 1115(e) and (f); that he did not reasonably conclude that the length of the extension was necessary; and that he lacked the statutory authority to approve the requested premiums because Sections 1396o and 1396o-1 mandate them. Nor does it need to consider Plaintiffs’ challenge to the Secretary’s 2023 letter declining to rescind the now-invalid 2020 approval. While those questions could perhaps resurface on remand, the Court sees no need to enter that thicket now.

All that is left, then, is the issue of remedy. This issue is vigorously contested, with Plaintiffs urging vacatur upon remand and HHS and Indiana pushing for remand without vacatur.

“Vacatur is the normal remedy under the APA, which provides that a reviewing court ‘shall . . . set aside’ unlawful agency action.” Long Island Power Auth. v. FERC, 27 F.4th 705, 717 (D.C. Cir. 2022) (quoting 5 U.S.C. § 706(2)); see Bhd. of Locomotive Engineers & Trainmen v. Fed. R.R. Admin., 972 F.3d 83, 117 (D.C. Cir. 2020) (“Vacatur is the normal remedy when we are faced with unsustainable agency action.”) (cleaned up); Sierra Club v. Van Antwerp, 719 F. Supp. 2d 77, 78 (D.D.C. 2010) (“[B]oth the Supreme Court and the D.C. Circuit

57

Court have held that remand, along with vacatur, is the presumptively appropriate remedy for a violation of the APA.”). That said, courts “sometimes decline to vacate an agency’s action.” Allina Health Servs. v. Sebelius, 746 F.3d 1102, 1110 (D.C. Cir. 2014). That decision depends on both the “seriousness of the order’s deficiencies (and thus the extent of doubt whether the agency chose correctly) and the disruptive consequences of an interim change.” Allied-Signal, Inc. v. U.S. Nuclear Reg. Comm’n, 988 F.2d 146, 150-51 (D.C. Cir. 1993) (citation omitted); see also Standing Rock Sioux Tribe v. U.S. Army Corps of Eng’rs, 282 F. Supp. 3d 91, 103 (D.D.C. 2017) (declining to vacate when agency “largely complied” with statute and could likely substantiate prior conclusions on remand).

In each of its prior Medicaid cases, the Court concluded that both factors supported vacatur. The Secretary’s failure to adequately consider an objective of Medicaid is a serious deficiency. Stewart I, 313 F. Supp. 3d at 273. As to Kentucky HEALTH, vacatur was not overly disruptive because the project had “yet to take effect” and the plaintiffs would suffer “serious harm[s]” were Kentucky HEALTH allowed to be implemented pending further proceedings. Id.; see also Stewart II, 366 F. Supp. 3d at 156. The same was true in the Granite State because it “ha[d] not fully implemented the community-engagement requirements” of Granite Advantage. Philbrick, 397 F. Supp. 3d at 32. While vacatur would concededly have been disruptive in Arkansas given that the Work Amendments had already begun in part, the seriousness of this disruption — which was largely administrative in nature — had to be “balanced against the harms that Plaintiffs and persons like them will experience if the program remains in effect.” Gresham, 363 F. Supp. 3d at 184. The Court thus vacated the Bear State’s program. Id. Although the factual circumstances are different in this case, the same outcome is warranted.

58

a. Seriousness of Deficiencies As HHS seems to realize, the first factor does not favor the Government. See HHS MSJ at 41–43 (declining to argue that factor cuts against vacatur). The Court has repeatedly concluded that the Secretary’s failure to adequately consider an objective of Medicaid is a “major shortcoming” going “to the heart” of his decisions. Stewart I, 313 F. Supp. 3d at 273; see also Stewart II, 366 F. Supp. 3d at 155–56; Gresham, 363 F. Supp. 3d at 182–83; Philbrick, 397 F. Supp. 3d at 32; cf. Humane Soc’y, 865 F.3d at 614 (“fail[ure] to address” important aspect of problem is “major shortcoming[]”); SecurityPoint Holdings, Inc. v. TSA, 867 F.3d 180, 185 (D.C. Cir. 2017) (“[T]he court must vacate a decision that ‘entirely failed to consider an important aspect of the problem.’”) (quoting State Farm, 463 U.S. at 43); Wedgewood Village Pharmacy v. DEA, 509 F.3d 541, 552-53 (D.C. Cir. 2007) (vacating after failure to consider important aspect of problem).

Indiana nonetheless protests. It insists that Plaintiffs’ “criticisms” of the 2020 approval “largely go to the adequacy of CMS’s explanation regarding whether HIP was likely to promote Medicaid’s goals despite its impacts on coverage,” and “[t]hat could be addressed on remand.” Indiana MTD at 40. The Court does not rule out the possibility that HHS could indeed rehabilitate the approval on remand. Cf. Gresham, 363 F. Supp. 3d at 182 (finding that first factor disfavors HHS but declining to conclude that “it will be impossible for the agency to justify its approval of a demonstration project like this one”). That said, given that the Secretary issued the deficient approval in this case notwithstanding the benefit of the Court’s Opinions in Stewart I, Stewart II, Gresham, and Philbrick and the Circuit’s opinion in Gresham, all of which offered crystal-clear guidance that Section 1115 mandated that coverage considerations be a central part of the analysis, the Court is less confident in HHS’s curative ability. At the very

59

least, it is once again fair to say that “the road to cur[ing] the deficiency . . . is, at best, a rocky one,” which “strongly weigh[s] in favor of vacatur.” Id. at 183.

b. Seriousness of Disruption The second factor is closer. Because Indiana has been implementing HIP 2.0 for almost a decade, HHS contends that vacating the Section 1115 approval “at this late stage would be immense[ly]” disruptive. See HHS MSJ at 42; see also Gresham, 363 F. Supp. 3d at 183 (factor was “closer call” than in Stewart I in part because Arkansas had already been implementing demonstration). Indiana likewise supposes that vacatur would “prove enormously disruptive to the many members who have ‘relied on [HIP] in good faith’ for years,’ as well as others.” Indiana MTD at 40 (quoting A.L. Pharma, Inc. v. Shalala, 62 F.3d 1484, 1492 (D.C. Cir. 1995)).

Begin with the Government’s concerns, which are twofold. First, it posits that “vacating the entire HIP approval” while the Hoosier State is focused on its unwinding efforts following the end of the COVID public-health emergency would cause “chaos.” HHS MSJ at 42. For support, the agency points to its December 2023 letter in which, as the reader will recall, it announced its belief “that withdrawing [Indiana’s] authorities at this time [was] too disruptive, particularly in the context of the state needing to maintain focus on keeping people covered through Medicaid unwinding and the resumption of Medicaid renewals following the COVID-19 Public Health Emergency.” AR 1. The letter predicted that, among other things, “the added complexity from any action to the demonstration’s overall operational system may lead to inaccuracies in beneficiary eligibility determinations during unwinding and result in beneficiaries being inadvertently disenrolled and delays to new beneficiary enrollment.” AR 2.

The Court appreciates the agency’s concerns about the inconvenience of vacating the Section 1115 approval while the State is engaged in its unwinding process, but such

60

inconvenience does not tip the scales against vacatur. As Plaintiffs point out, the argument that vacating the approval would be unduly disruptive is seriously undermined by the reality that the premium requirements and associated penalties have not been in effect since March 2020. See Pl. MSJ at 35 (“Common sense dictates that the reintroduction of a requirement for Medicaid beneficiaries, with which they have not had to comply for nearly four years, will be considerably more disruptive for them and for the State.”). It is far from clear why retaining the status quo — viz., declining to charge beneficiaries premiums and impose associated penalties — would “result in beneficiaries being inadvertently disenrolled.” AR 2. Of course, the reason that the State has been able to sustain HIP 2.0 without charging premiums for four years is no mystery: the Families First Coronavirus Response Act provided enhanced federal funding for that purpose. But the State makes no argument that if it cannot charge premiums, all of HIP 2.0 will fall because of lack of funding. As to the features of the program that have been in effect since 2020, neither Defendant proposes any arguments specific to the waivers of retroactive coverage or NEMT.

What is more, the Government evidently does not believe that it is impossible for states to adjust to changes in their waivers and authorities in the midst of COVID unwinding. In fact, just one month before issuing the 2023 letter here, the Secretary withdrew Wisconsin’s authority to impose premiums in the midst of the very same unwinding process. See Letter from Daniel Tsai, Deputy Admin. and Dir. CMCS, to Jamie Kuhn, State Medicaid Dir., Wis. Dep’t of Health Servs. (Nov. 17, 2023), https://perma.cc/7TZN-YDLX; see also Pl. MSJ at 35. Not only that: it told Indiana in its 2023 letter that if the State determined that it could terminate (read: decline to restart) premiums without affecting the unwinding efforts, then HHS “encourage[d] the state to do so” and “would be available to provide technical assistance on mitigating any operational

61

challenges from such a termination, including charting a path to do so without causing undue beneficiary loss of coverage from errors in eligibility determinations processes.” AR 2. The Court sees no reason why HHS cannot provide similar assistance in the face of vacatur, especially since Indiana is nearly (if not entirely) finished with its unwinding process.

Second, HHS maintains that because vacatur would “change the path by which [HIP]

beneficiaries have been receiving medical coverage,” the “resulting confusion would be great.” HHS MSJ at 43. Specifically, it anticipates Indiana facing “significant operational challenges . . . in making changes to its operations and messaging.” HHS Reply at 24. The Court rejected a similar argument in Gresham, explaining that although vacating “requirements that have already been implemented may send mixed messages, . . . any disruption in this respect is not sufficiently significant to avoid vacatur.” 363 F. Supp. 3d at 184. The same is true here. In fact, as in that case, the structure of HIP 2.0, considered with the timing of this Opinion, renders vacatur less disruptive that might be expected. Because Indiana has not implemented the premium requirements since 2020 and thus has not disenrolled (or defaulted from HIP Plus to HIP Basic) any beneficiaries for failure to pay premiums since the 2020 approval, vacatur of that approval will not require it to re-enroll persons who have lost their coverage, with the administrative and communication-related headaches that might entail. Id. The bottom line is the same as in Gresham. Although changes in operations and messaging may prove complicated, “[t]his is not a case in which the egg has been scrambled, and it is too late to reverse course.” Id. (quoting Allina Health, 746 F.3d at 1110–11) (cleaned up). And Indiana is not the first state to have to change course in the midst of active waivers of retroactive coverage and NEMT. See Gresham, 363 F. Supp. 3d at 183–84; Philbrick, 397 F. Supp. 3d at 32–33.

62

The State’s worries are slightly different: it principally fears that vacatur would undermine its ability to provide coverage to the expansion population at all. See Indiana MTD at 41–42. Its argument on this score is much the same as the one it offered in the redressability context: because Indiana law provides for coverage of the expansion population “only” through HIP, see Ind. Code § 12-15-44.5-10(a), and the FSSA is not permitted to modify HIP’s premium requirements, id. § 12-15-44.5-10(b)(1)–10(b)(8), “vacatur would immediately eliminate the very existence of Indiana’s HIP program.” Indiana MTD at 41–42. As the Court explained above, see Section III.A.1.b, supra, the expansion population’s eligibility for Medicaid does not depend on the Secretary’s approval of the Section 1115 application. Disapproval of the lockout- penalties component of the project, moreover, did not eliminate the existence of HIP 2.0. And it is not even clear that Indiana could de-expand Medicaid if it wanted to. The Court is therefore not persuaded by Indiana’s arguments on this front.

Finally, it bears emphasis that the disruptions to Indiana’s administration of its Medicaid program “must be balanced against the harms that Plaintiffs and persons like them will experience if the program remains in effect.” Gresham, 363 F. Supp. 3d at 184; cf. A.L. Pharma, Inc., 62 F.3d at 1492 (explaining vacatur inappropriate because “nothing in the record suggests that significant harm would result from allowing the approval to remain in effect pending the agency’s further explanation”). The 2020 approval imposes barriers to health-insurance coverage for tens of thousands of Hoosiers. Indeed, HHS itself acknowledged as much — both in its 2023 letter and in an issue brief it published in 2021 reviewing the evidence on the impact of demonstration policies, including HIP 2.0 in Indiana. See, e.g., AR 4–10 (walking through evidence that premiums reduce coverage); U.S. Dep’t of Health & Hum. Servs., Medicaid Demonstrations and Impacts on Health Coverage: A Review of the Evidence (Mar. 11, 2021),

63

https://perma.cc/5KY2-K94S (citing research that premiums and waiver of retroactive coverage have reduced coverage). Weighing the harms these persons will suffer from leaving in place a legally deficient approval against the disruptions due to vacatur leads the Court to conclude that the 2020 approval of HIP 2.0 cannot stand.

* * *

As a last resort, HHS argues that, rather than vacating the 2020 approval as a whole, this Court should tailor any relief solely to Plaintiffs and the aspects of the program that they have successfully challenged. See HHS MSJ at 41. The Court is no more moved by this appeal than it was in Stewart II, 366 F. Supp. 3d at 155, or Philbrick, 397 F. Supp. 3d at 32. As it explained in the latter, in an APA case, the “ordinary result” of the Court’s finding an agency action unlawful is to vacate that action — not to judicially re-write what the agency did so that it somehow does not apply to a narrow group of people or so that it persists piecemeal. Philbrick, 397 F. Supp. 3d at 32; see Harmon v. Thornburgh, 878 F.2d 484, 495 n.21 (D.C. Cir. 1989). The D.C. Circuit, accordingly, has explained that “if the plaintiff prevails” in challenging an agency action under the APA, “the result is that the rule is invalidated, not simply that the court forbids its application to a particular individual.” Nat’l Min. Ass’n v. U.S. Army Corps of Engineers, 145 F.3d 1399, 1409 (D.C. Cir. 1998) (quoting Lujan v. Nat’l Wildlife Fed., 497 U.S. 871, 913 (1990) (Blackmun, J., dissenting)). The Court sees no reason to depart from the ordinary practice of vacating the agency action found unlawful under the APA. IV. Conclusion For these reasons, the Court will grant Plaintiffs’ Motion for Summary Judgment on Count I. It will also deny HHS’s Cross-Motion and Indiana’s Motion to Dismiss, vacate the

64

Secretary’s 2020 approval of HIP 2.0, and remand to the agency. A separate Order so stating will issue this day.

/s/ James E. Boasberg

JAMES E. BOASBERG

Chief Judge

Date: June 27, 2024

65

Appendix A

Premiums AR 5959 (Lewin Group) (“Between February 1, 2015 and November 30, 2016, 9,636 unique individuals with incomes above 100 percent of the FPL were disenrolled from HIP Plus coverage due to non-payment of [the POWER Account Contribution] and subject to a six-month disenrollment period. . . . An additional 3,914 individuals with incomes above 100 percent of the FPL were disenrolled from HIP Basic coverage due to non-payment of PAC . . . . Together, these counts of Leavers and Never Members sum to 57,189 unique members disenrolled or not enrolled due to non-payment, which represents 29 percent of individuals who could be disenrolled or not enrolled due to non-payment during the timeframe.”); AR 7194 (National Health Law Program) (“Redundant research proves that premiums deter and reduce enrollment among low-income individuals. Numerous studies, conducted over the course of almost two decades, have examined the effects of imposing premiums in Medicaid . . . [and] show the same patterns — people facing premiums are less likely to enroll, more likely to drop coverage, and more likely to become uninsured.”); AR 8097 (Indiana Legal Services, Inc.) (“Studies show that premium requirements lead to coverage loss and disincentivize enrollment.”); AR 8111 (Federal AIDS Policy Partnership) (“Under HIP Plus, people are losing coverage; and under HIP Basic, people have the pretext of coverage but are discouraged from using it. . . . HIP has served to demonstrate that premiums and co-pays reduce coverage and access to care.”); AR 8196–97 (American College of Obstetricians & Gynecologists)

(“Evidence suggests that the HIP premium requirements are keeping people from getting covered.”) (citing studies); AR 8206 (Cystic Fibrosis Foundation) (“Experience with this program shows that the POWER Account contributions result in fewer people enrolled in coverage.”); AR 8219–20 (Center on Budget & Policy Priorities, Georgetown University Center for Children and Families, 15 other organizations, and a professor of public policy) (“Premiums are decreasing participation in coverage.”) (capitalization altered).

Retroactive AR 7203 (National Health Law Program) (“When Indiana received Eligibility permission to waive retroactive coverage in 2015, CMS required the State to continue to provide some retroactive coverage to § 1931 parents and caretaker relatives. The State reported to CMS that 13.9 percent of people in that eligibility category who enrolled in Medicaid needed retroactive coverage, with their costs incurred averaging $1,561 per person. Iowa estimated that waiving retroactive coverage in its Medicaid program would decrease coverage by 3,344 people every month and over 40,000 people every year.”); AR 7204 (National Health Law Program) (“Ultimately, many providers likely stop providing care to individuals who are eligible for Medicaid but have not enrolled, meaning that low-income individuals experience a

66

substantial delay in receiving medically necessary care. Notably, Congress passed the retroactive coverage requirement in part to avoid this very problem.”); AR 8207 (Cystic Fibrosis Foundation)

(“Limiting retroactive coverage and delaying eligibility for up to 60 days can result in gaps in coverage or costly medical bills for individuals in need of care.”); AR 8213 (American Cancer Society Cancer Action Network, et al.) (“Retroactive eligibility in Medicaid prevents gaps in coverage, by covering individuals for up to 90 days (or quarter of the year) prior to the month of application, assuming the individual is eligible for Medicaid coverage during that time frame. It is common that individuals are unaware they are eligible for Medicaid until a medical event or diagnosis occurs.”).

NEMT AR 7162 (Leukemia & Lymphoma Society) (“Low-income patients may not own a car and may lack access to reliable public transportation, especially in rural areas. Additionally, some patients may not be able to drive themselves due to their condition or sideeffects of their treatment. Removing this benefit will therefore harm patients who need to attend regular visits with their providers to manage their medications and treatments.”); AR 7213–14 (National Health Law Program) (“[E]liminating NEMT runs counter to the objectives of the Medicaid Act, as it will reduce access to medically necessary services for HIP enrollees. . . . Many people who live in poverty simply do not have the means to access medically necessary services on their own. Access to private vehicles is lower and transportation barriers are higher among lower-income populations, and Medicaid beneficiaries in particular. . . . [D]ata from Indiana and Iowa — the two states that have received permission to eliminate NEMT for the expansion population — demonstrate that many enrollees cannot access care without NEMT.”); AR 8112 (Federal AIDS Policy Partnership) (“What Indiana’s experiment did show is that transportation is still a barrier to health care for HIP enrollees.

Transportation contributed to missing an appointment for 25% of enrollees with the NEMT benefit and 32% of enrollees without NEMT.”); AR 8223 (Medical Transportation Access Coalition)

(“[W]e specifically ask CMS to work with Indiana to reinstate comprehensive access to NEMT services in light of the following considerations: notably, the importance of NEMT in accessing care, the fact that the NEMT waiver fails to demonstrate it furthers a purpose of the Medicaid statute, the increasing prominence of NEMT as a feature of health insurance benefits designed to improve health and well-being, and concerns regarding the NEMT waiver’s negative impact on access to care and adherence to clinically recommended care.”).