Christopher BERTRAND, by his parents Inez BERTRAND and Daniel Bertrand, and Frank Patterson, Jr., by his parents Janice Patterson and Frank Patterson, Sr., Plaintiffs-Appellants, v. Barry S. MARAM, Director, Illinois Department of Healthcare and Family Services, et al., Defendants-Appellees.
Christopher Bertrand and Frank Patterson are developmentally disabled adults who want residential habilitation services under the Medicaid program in Illinois. “Residential habilitation” is a set of services provided in the applicant's home (or another residence) by nurses and other professionals. The goal is to see that the person remains safe and healthy; it is an alternative to institutionalization for those unable to care for themselves. Illinois covers the costs of residential habilitation and similar services through its Community Integrated Living Arrangement (CILA) program, part of the state's Home and Community-Based Services (HCBS) program.
The federal government does not require states to provide CILA services as a condition of participation. Indeed, federal law does not even allow states to provide such services as part of the Medicaid program (though states may do so separately, at their own expense) unless they apply for and receive a waiver of Medicaid's normal rules. The possibility of waiver, see 42 U.S.C. § 1396n(b), was extended to home and community-based services by 42 U.S.C. § 1396n(c)(1):
The Secretary may by waiver provide that a State plan approved under this subchapter may include as “medical assistance” under such plan payment for part or all of the cost of home or community-based services (other than room and board) approved by the Secretary which are provided pursuant to a written plan of care to individuals with respect to whom there has been a determination that but for the provision of such services the individuals would require the level of care provided in a hospital or a nursing facility or intermediate care facility for the mentally retarded the cost of which could be reimbursed under the State plan.
Illinois applied for and received a waiver entitling it to federal reimbursement for such services provided to 10,000 people.
When Bertrand and Patterson applied for residential habilitation services, however, they were turned down. Although both Bertrand and Patterson “require the level of care provided in a hospital or a nursing facility or intermediate care facility for the mentally retarded”, and each already received some services under the HCBS program, each was told that he did not satisfy the state's “priority population criteria” for residential habilitation under the CILA sub-program. These are the criteria:
(1) individuals who are in crisis situations (e.g., including but not limited to, persons who have lost their caregivers, persons who are in abusive or neglectful situations); (2) individuals who are wards of the Illinois Department of Children and Family Services and are approaching the age of 22 and individuals who are aging out of children's residential services funded by the Office of Developmental Disabilities; (3) individuals who reside in State-Operated Developmental Centers; (4) Bogard class members, i.e., certain individuals with developmental disabilities who currently reside in a nursing facility; (5) individuals with mental retardation who reside in State-Operated Mental Health Hospitals; (6) individuals with aging caregivers; and (7) individuals who reside in private ICFs/MR or ICFs/DD.
Those not on the list cannot be reimbursed for residential habilitation even if medical providers are willing to offer that service at a price Illinois is willing to pay. Bertrand applied for reconsideration, arguing (via his parents as next friends) that he comes within category (6). He lives with his parents, both of whom are nearing retirement. The state reversed its decision; Bertrand has been receiving residential habilitation services at state expense since May 24, 2005. But Patterson remains outside the CILA sub-program.
Plaintiffs maintain that the state's administration of its HCBS program violates 42 U.S.C. § 1396a(a)(8), which says that every state plan must “provide that all individuals wishing to make application for medical assistance under the plan shall have opportunity to do so, and that such assistance shall be furnished with reasonable promptness to all eligible individuals”.
Illinois argued that Bertrand's claim should be dismissed as moot. The district judge disagreed, remarking that Bertrand had moved for class certification before he was accepted into the CILA program. Paradoxically, however, the judge then refused to act on Bertrand's motion, ruling that class treatment is itself moot because Illinois is entitled to prevail-not, as the state principally argued, because there is no private right of action under § 1396a(a)(8), but because the Secretary of Health and Human Services approved the state's “priority population criteria,” and anyone not on the list is not “eligible” for services as § 1396a(a)(8) uses that term. Bertrand v. Maram, 2006 WL 2735494, 2006 U.S. Dist. LEXIS 68935 (N.D.Ill. Sept. 25, 2006).
The district court mishandled the issues related to class certification. Bertrand and Patterson filed this suit seeking to represent a class. Fed.R.Civ.P. 23(c)(1) directs district courts to grant or deny class certification “early” in the litigation. Yet the district court bypassed that subject, ruled on the merits almost two years after the suit had been filed, and then insisted that the class does not matter. “Early” is a plastic term that affords latitude to district judges in case management, but “never” is not within any plausible understanding of “early.”
Judge Kendall, who finally resolved plaintiffs' claim, was assigned to the case about a year after its commencement and is not responsible for her predecessor's failure to make an “early” decision about the class. That the subject may have fallen between stools is unfortunate, however. Prompt decision is essential, as Bertrand's situation shows. Board of School Commissioners of Indianapolis v. Jacobs, 420 U.S. 128, 95 S.Ct. 848, 43 L.Ed.2d 74 (1975), holds that, if a class representative's personal claim becomes moot after certification, then the suit may continue-for the class as a whole retains a live claim. But if the would-be representative's claim becomes moot before certification, then the case must be dismissed, see Sosna v. Iowa, 419 U.S. 393, 95 S.Ct. 553, 42 L.Ed.2d 532 (1975), unless someone else intervenes to carry on as the representative. Parole Commission v. Geraghty, 445 U.S. 388, 100 S.Ct. 1202, 63 L.Ed.2d 479 (1980).
Apparently the district court saw class suits as opportunities for one-way intervention: if the representative plaintiff wins, then class certification extends the victory to a larger group. That was a common view before the 1966 amendments to Rule 23, which were designed to divorce class certification from the merits. See the Committee Note to the 1966 amendment. After the 1966 amendments, treatment of plaintiffs and defendants is supposed to be symmetric, which is possible only if a class is certified (or not) before decision on the merits. Class-action status must be granted (or denied) early not only to avoid problems with mootness, and provide an opportunity for interlocutory review, see Fed.R.Civ.P. 23(f), but also to clarify who will be bound by the decision. The larger the class, the more the litigants may invest in discovery and briefing to ensure that the case is decided correctly. Until everyone knows who will, and who will not, be bound by the outcome, it is difficult to make informed decisions about how the case should proceed.
The district judge may have equated the precedential effect of a decision with its preclusive effect, but the two differ. Without a certified class, any other Medicaid applicant is free to file another suit and present the same arguments; decisions of district courts do not block successive litigation by similarly situated persons. Although decisions of appellate courts have broader authority, in the absence of class certification any other applicant may start over and try to distinguish the adverse precedent. Likewise, if the first plaintiff wins and the court of appeals affirms, the agency may try to distinguish the adverse precedent, or deny its authoritative status, when denying relief to a similarly situated applicant. It takes a class certification to produce a conclusive resolution in one proceeding. Compare United States v. Mendoza, 464 U.S. 154, 104 S.Ct. 568, 78 L.Ed.2d 379 (1984), with Califano v. Yamasaki, 442 U.S. 682, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979).
In the event, however, no class was certified, and neither side seeks review of that decision. Bertrand's claim therefore must be dismissed as moot. The district court noted the motion for class certification but missed the vital qualification that the suit never became a class action. In a handful of situations, exemplified by Deposit Guaranty National Bank v. Roper, 445 U.S. 326, 100 S.Ct. 1166, 63 L.Ed.2d 427 (1980), and Primax Recoveries, Inc. v. Sevilla, 324 F.3d 544, 546-47 (7th Cir.2003), class certification may follow the defendant's actual or attempted satisfaction of the would-be representative's demand; the Court explained in Deposit Guaranty National Bank that this proviso is essential to prevent defendants from buying off all potential class representatives by meeting their demands, one at a time, and thus preventing effectual relief to a larger class of victims. Nothing of the sort occurred here-and, to repeat, no class has been certified, so even if Bertrand had been furnished CILA services for strategic reasons this would not justify allowing him to continue litigating in his own name. He lacks a stake in the outcome, and his claim must be dismissed as moot.
Section 1396a(a)(8), on which Patterson (the remaining plaintiff) relies, does not provide a private right of action. Neither does any other arguably relevant provision in the Medicaid Act. This leads Patterson to rely on 42 U.S.C. § 1983 and the approach of Maine v. Thiboutot, 448 U.S. 1, 100 S.Ct. 2502, 65 L.Ed.2d 555 (1980): § 1983 allows the enforcement of federal law (such as the Medicaid statute) against state actors (such as the Illinois Department of Healthcare and Family Services). To this Illinois responds that Medicaid is a funding statute that gives states an option rather than placing them under an obligation. If the state has not kept its end of the bargain, the argument goes, then the remedy is to cut off the funds rather than to order specific performance. The Medicaid Act provides some express private remedies; using § 1983 to augment them would upset the bargain that the state struck when it joined the program, the state insists. Illinois relies particularly on Gonzaga University v. Doe, 536 U.S. 273, 122 S.Ct. 2268, 153 L.Ed.2d 309 (2002), which reached just such a conclusion with respect to the Family Education Rights and Privacy Act, another federal statute that uses the lure of funds to achieve a national objective.
At least two courts of appeals have held since Gonzaga University, however, that § 1396a(a)(8) creates personal rights that are enforceable as long as the state continues to accept federal money. Sabree v. Richman, 367 F.3d 180 (3d Cir.2004); Bryson v. Shumway, 308 F.3d 79, 88-89 (1st Cir.2002). Another circuit reached the same conclusion with respect to § 1396a(a)(10), which is materially identical to § 1396a(a)(8). See S.D. v. Hood, 391 F.3d 581, 603 (5th Cir.2004). Sabree, the most thorough of these decisions, observes that before Gonzaga University the Court had held that one portion of the Medicaid Act may be enforced via § 1983, see Wilder v. Virginia Hospital Ass'n, 496 U.S. 498, 110 S.Ct. 2510, 110 L.Ed.2d 455 (1990), and that Gonzaga University did not overrule Wilder. Although Gonzaga University may have taken a new analytical approach, courts of appeals must follow the Supreme Court's earlier holdings until the Court itself overrules them. This circuit has itself assumed after Gonzaga University that § 1396a(a)(8) may be enforced via § 1983. See Bruggeman v. Blagojevich, 324 F.3d 906, 910-11 (7th Cir.2003) (reaching the merits of a claim under § 1396a(a)(8) without discussing the availability of a private right of action, although after mentioning Gonzaga University in another connection).
What one could say in response is that none of these decisions dealt with the application of § 1396a(a)(8) to a supplemental state program approved under a waiver. The idea behind § 1396a(a)(8) is that states must comply with all Medicaid obligations: to enter the program at all is to agree to supply medical services for every eligible person. Once Congress created the waiver program in 1981, however, that situation changed. Today a state may propose to cover applicants in some parts of its territory but not others, or to place a limit on the number of persons who receive treatment. “The Secretary, to the extent he finds it to be cost-effective and efficient and not inconsistent with the purposes of this subchapter, may waive such requirements of section 1396a of this title (other than subsection (s)) (other than sections 1396a(a)(15), 1396a(bb), and 1396a(a)(10)(A) of this title insofar as it requires provision of the care and services described in section 1396d(a)(2)(C) of this title) as may be necessary for a State” to accomplish certain objectives. 42 U.S.C. § 1396n(b). When Illinois adopted its HCBS program, it particularly asked the Secretary to waive the requirement that all comers be accepted, and the Secretary agreed. The HCBS program in Illinois will pay for no more than 10,000 persons.
There are three ways to keep within that limit. One is the price system, which is not possible under the Medicaid Act. The second is a queue: Everyone who wants to participate joins a line and is admitted as current participants move to some other state, are institutionalized, or die. The third is a triage device. Illinois chose the third way, through its “priority population criteria.” Patterson contends that § 1396a(a)(8) obliges the state to use the second device-and when there are unfilled slots available to put the next medically eligible person into them. But it is far from clear to us that dictating a means to implement a limited-enrollment program is a function of § 1396a(a)(8).
A cap on enrollment serves fiscal rather than medical objectives, saving money for both state and federal governments; one reason why the Secretary's approval is necessary for these optional programs is to ensure that states don't make commitments that cost the national government more than it is willing to spend. When legislation such as § 1396n(b) and (c)(1) is designed to save money rather than deliver subsidized care to everyone, § 1396a(a)(8) is a poor fit-and it is correspondingly more attractive to structure litigation with the Secretary as the defendant, so that the agency that made the waiver decision may be asked whether the state's program accurately carries out the conditions negotiated with the federal government. A state is entitled to use “priority population criteria” as an entry-control device if the Secretary has approved that use. But it is difficult to determine the Secretary's views in a proceeding against the state. A request that the Secretary terminate funding would avoid the empty-chair problem.
Because the parties have not briefed the question whether § 1983 supplies a private right of action to enforce claims under § 1396a(a)(8) in the context of waiver, we think it best to proceed as in Bruggeman: to assume that there is such an entitlement, while leaving resolution to the future. A private right of action is not a component of subject-matter jurisdiction, see Grable & Sons Metal Products, Inc. v. Darue Engineering & Manufacturing, 545 U.S. 308, 125 S.Ct. 2363, 162 L.Ed.2d 257 (2005), so this is a permissible approach.
May Illinois use the priority population criteria as a triage device? Patterson insists that the answer does not matter, because when he applied there were openings among the allocated 10,000 slots in the HCBS program-and, what is more, Patterson himself occupied one of them. An earlier decision had admitted Patterson to the program, though not to the set of CILA services that he most desired. In fiscal 2004, for example, there were 138 open slots in HCBS, and in fiscal 2003 there had been 230. It does not follow from this, however, that everyone with a developmental disability should be admitted immediately. Abolish the priority population criteria and a queue would develop. People with relatively weak needs for this service would receive it just because they applied first, while others with grave needs would be put off pending an opening. The only way to ensure that slots are available for those highest on the priority list is to hold some of them open at all times. Keeping 1% or 2% (100 to 200) of the slots available seems a prudent precaution on behalf of those with the greatest need.
When asked at oral argument whether Patterson could be removed from the CILA or HCBS program, after admission, in order to free up a slot for someone with a better claim to the resources, his lawyer gave a negative answer. According to counsel, anyone provided a particular service is in for life; any comparison between the needs of those already in and those making new applications is forbidden. That would lead some people to demand entry to the HCBS program, even if they did not require its services, to ensure the availability of CILA services later should a need (or a desire) develop. If that is so, then the only sensible approach is the one Illinois has chosen. Priorities must be established, and some slots must be kept open at all times to avoid turning away people “in crisis situations” (priority 1) and other high-need applicants. This is true not only for the home and community-based services as a whole but also for each component of that umbrella category. Medicaid makes each separate component of the umbrella program the subject of a “medical need” requirement, see 42 C.F.R. § 440.230(d), which makes a great deal of sense. A conclusion that a person could benefit from one aspect of HCBS does not mean that the person is a good candidate for every kind of related service.
Patterson does not contest the state's assignment of priorities or contend that his needs are equivalent to those of people who meet the “priority population criteria.” His argument, as we have said, is that the state must use an unsorted queue and provide services to everyone who could get some benefit from them. Yet a queue would be problematic under § 1396a(a)(8)-for recall that this statute demands provision of services “with reasonable promptness”. Taking everyone with a developmental disability, in order of application, would defeat prompt admission for those who would receive the greatest benefit.
The record establishes that the Centers for Medicare and Medicaid Services (CMS), the bureau within the Department of Health and Human Services that decides whether to grant states' applications for waiver of the Medicaid rules, knew about the criteria that Illinois proposed to use. The state's application says point blank that “[f]or residential services, the State gives service priority to eligible persons according to the following priority population criteria ․”. The application does not reveal whether CMS appreciated that Illinois would use these criteria to exclude some applicants from CILA services, and CMS did not write an opinion explaining its understanding of the state's program. An affidavit from one of the state's program administrators says that “[t]he Priority Population Criteria were discussed with the CMS review team during the review process”. What was said, concretely? The record does not reveal the answer-but then plaintiffs' counsel did not follow up with a deposition that might have produced the information. Plaintiffs have not offered any evidence tending to establish that CMS granted the state's request in ignorance of how Illinois employs the priority population criteria. Because plaintiffs bear the burden of persuasion, we must take it that the state's approach is agreeable to CMS.
And on that understanding the case is almost over. Another statute on which Patterson relies, 42 U.S.C. § 1396n(c)(2)(C), offers him no assistance. This subsection says that persons entitled to care must be “informed of the feasible alternatives, if available under the waiver, at the choice of the individuals”. Patterson does not say that he has been kept ignorant of options open to him. His argument is that CILA services should be “available,” but this subsection does not make any particular option “available” to anyone. It just requires the provision of information about options that are available.
Each side invokes Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984); United States v. Mead Corp., 533 U.S. 218, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001); and Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944). Illinois maintains that the agency's approval is entitled to Chevron deference. Plaintiffs insist, to the contrary, that a letter that CMS circulated in 2001 should receive Chevron deference (or at least Mead-Skidmore respectful consideration). To this the state replies that letters, not being regulations, should not play any role at all. (Just in case we disagree, the state insists that its use of the “priority population criteria” is consistent with the letter.) All of this is byplay. It would matter if the “priority population criteria” were something that the state had invented after receiving the waiver from CMS, and we had to decide whether the state's new approach was consistent with the statute and regulations. But that's not what happened; CMS considered the “priority population criteria” before granting the waiver. A state does not violate § 1396a(a)(8) by using the criteria that formed (part of) the basis for requesting a waiver under § 1396n(c)(1).
EASTERBROOK, Chief Judge.