July 18, 1986.
Willie M.J. Curtis, Ralph M. Johnson, San Francisco, Cal., Evelyn R. Frank, Legal Aid Soc. of Alameda Cty., Oakland, Cal., for petitioner.
Brenda F. Kohn, Cade L. Morrow, San Francisco, Cal., for respondent.
[EDITORS' NOTE: FOR MAJORITY OPINION SEE 781 F.2d 1421.]
The panel's opinion in this case is an exercise in judicial legislation. Disregarding principles of administrative law long established by the Supreme Court and this court, the opinion gives birth to a sweeping new doctrine: that a federal court may endow an agency's internal operating manual with the force of law. In so doing, the panel fails to acknowledge that the manual in question was prepared by government employees lacking authority to issue regulations; was expressly limited to use within the agency; and contravened properly issued regulations, that the panel overrules sub silentio. Equally unprecedented is the panel's alternative ruling, reversing the Secretary for the curious reason that his decision cannot be sustained under an inapplicable statute on which he did not rely. It is difficult to reconcile the opinion's tortuous path with the Supreme Court's admonition that administrative decisions not be set aside "simply because the court is unhappy with the result reached." Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 558, 98 S.Ct. 1197, 1219, 55 L.Ed.2d 460 (1978).
Aside from the severe financial impact this case will have by its own force — the government estimates $50 million annually, Pet.Reh. at 1 n. 1 — its precedential effect as the law of the nation's largest circuit is difficult to overestimate. The opinion seriously blurs lines dividing regulations from other agency issuances, and separating those employees who may lawfully bind the agency from those who may not. By making every federal agency's operating manuals and internal guidelines fair game to a claim that they carry the force of law, the opinion will wreak havoc with administrative decision-making, and shift significant power from the Executive Branch to the courts.
These arguments, and others discussed below, are advanced crisply and persuasively in Respondent's Petition for Rehearing with Suggestion for Rehearing En Banc. They remain unanswered and, to my mind, unanswerable. Yet answers they deserve. If the panel is unwilling or unable to address these contentions, the full court should. Because I believe the court did not live up to its grave responsibility for maintaining the fabric of the law intact, I respectfully dissent.
This case concerns the extent of federal participation in Medi-Cal, California's version of Medicaid. Title XIX of the Social Security Act, 42 U.S.C. § 1396 et seq. (1982 Supp. 1985), sets standards for such participation. Specifically, a state must provide medical assistance to recipients of welfare benefits. 42 U.S.C. § 1396a(a)(10)(A) (1982). In addition, a state may provide Medicaid coverage to various groups (e.g. the aged, blind or disabled) who are ineligible for welfare because their income is somewhat too high, but who are nevertheless poor. These people are deemed "medically needy." Id. §§ 1396a(a)(10)(C), 1396d. Each state must define who is medically needy in accordance with federal standards, id. § 1396a(a)(10(C)(i); the income standard for such recipients is known as the "medically needy income level" or MNIL.
At issue in this case is California's attempt to loosen the standards for Medi-Cal eligibility by increasing the MNIL for a family of two medically needy adults to 133-1/3% of the highest amount that would be paid to a family of three under AFDC. The Health Care Financing Administration (HCFA) disapproved the proposed revision because it violated 42 U.S.C. § 1396b(f) (1982). The Secretary affirmed.
A panel of this court reversed, basing its decision on alternative rulings. First, the panel concluded that the Secretary's interpretation of section 1396b(f) was inconsistent with a section of HCFA's Regional Office Manual (ROM). The panel held that this provision was a legislative regulation and interpreted it as allowing California's increase in the MNIL. The panel also held that the Secretary's action was barred by section 2373(c)(1) of the Deficit Reduction Act (DEFRA), Pub.L. 98-369, which imposes a moratorium on certain regulatory actions against state Medicaid plans.
1. The Statute, the Regulation and the ROM
The statute provides that a family's MNIL is "the amount determined, in accordance with standards prescribed by the Secretary, to be equivalent to 133-1/3 percent of the highest amount which would ordinarily be paid [under AFDC] to a family of the same size without any income or resources[.]" 42 U.S.C. § 1396b(f)(1)(B)(i) (1982) (emphasis added). The panel rejected the Secretary's argument that this section limited the MNIL to 133-1/3% of the maximum AFDC payment families of the same size could receive, pointing to the phrase "in accordance with standards prescribed by the Secretary." The panel reasoned that this language allows the Secretary to devise a formula permitting states to set the MNIL at more than 133-1/3% of the maximum amount payable under AFDC to families of the same size. The panel then held that HCFA's Regional Office Manual § 2572-D is that formula. It interpreted this ROM section as allowing the State to set the MNIL in terms of the income of an AFDC family of a different size. The panel rejected the Secretary's attempt to repudiate this ROM section. It held that the provision is a legislative regulation, having the force and effect of law, binding on the Secretary. 781 F.2d 1421 at 1424-29 (9th Cir. 1986). The panel errs in almost every respect.
Most fundamentally, the Regional Office Manual is not a body of regulations. As its name suggests, the ROM is an office manual, an internal directive from the agency's headquarters to its field offices. The manual's Foreword emphasizes this point:
The Health Insurance Regional Office Manual (ROM) is the principal channel for transmitting program instructions of ongoing reference value to regional offices. . . .
The ROM is not intended to stand alone but rather to serve as the regional supplement to other health insurance manuals; e.g., provider manuals, intermediary manuals. Therefore, substance in other issuances will not generally be repeated. Appropriate cross-references will be used instead.
. . . .
This manual is for the use of regional office employees only and should not be cited when addressing contractors, State agencies, providers, etc.
HCFA Regional Office Manual Part 6 (Medicaid) iii.
As the panel recognizes, numerous cases in the Supreme Court and this court stand for the proposition that the contents of such manuals are not regulations, have no legal force and do not bind the agency. See Schweiker v. Hansen, 450 U.S. 785, 789, 101 S.Ct. 1468, 1471, 67 L.Ed.2d 685 (1981) (SSA Claims Manual); Powderly v. Schweiker, 704 F.2d 1092, 1096, 1098 (9th Cir. 1983) (same); Whaley v. Schweiker, 663 F.2d 871, 873 (9th Cir. 1981) (same); see also Evelyn v. Schweiker, 685 F.2d 351, 352 n. 5 (9th Cir. 1982) (Program Operations Manual System). The panel cites no case to the contrary and careful research reveals none.
The panel's attempt to distinguish these cases is far off the mark. For example, it states:
Unlike the office manual provision in the above cases, ROM § 2572-D is not a field-level instructional guide to the application of the statute with respect to the inquiries and circumstances of particular individuals. It is directed not at office-level agency employees and specific recipients but at states and a major category of recipients under state plans.
781 F.2d at 1426-27. The panel gives no citation for this counterintuitive assertion. Nor do I believe it could, since it flies in the face of the ROM's own description of itself: "This manual is for the use of regional office employees only and should not be cited when addressing contractors, State agencies, providers, etc." ROM Pt. 6, at iii (emphasis added).
One searches the opinion in vain for an explanation of how the panel's ruling can be squared with this disclaimer, or even an acknowledgment that the disclaimer exists. Nor does the panel explain why something called an office manual would contain procedures "directed not at office-level agency employees . . . but at states and a major category of recipients under state plans." 781 F.2d at 1426-27.
The panel goes on to assert that the ROM "establishes the right of states to increase their adult couple MNIL" and "thereby `effect[s] a change in existing law or policy[.]'" Id. at 1427 (quoting Powderly, 704 F.2d at 1098). But this merely assumes the conclusion. A provision can effect a change in existing law or policy only if it has the force of law; if it does not, it cannot.
The panel further claims that the ROM has the force of law because it has a "statutory basis:"
Unlike the statutes interpreted by the agency manuals in the cases cited by the Secretary, Section 1903(f) expressly authorized the Secretary to prescribe "standards." Congress must have intended that standards thus prescribed by HHS would have the force and effect of law. In promulgating the ROM provision at issue, the Secretary followed the mandate of Section 1903(f).
Id. at 1427. But words on paper do not become legislative regulations merely because the Secretary could have promulgated them as such; he must actually have done so. The Secretary, however, denies doing any such thing, and all available evidence supports his claim.
Congress granted the authority to promulgate rules and regulations to the Secretary alone, see 42 U.S.C. § 1302 (1982), authority he has jealously retained. There is absolutely no indication that the Secretary had anything to do with writing or approving HCFA's Regional Office Manual, or any part of it. The Regional Office Manual is a publication of HCFA, an operating division of HHS. Like thousands of other internal guidelines issued every year by scores of federal agencies, ROM updates are issued anonymously; the Foreword alone is signed and that by the Director of the Bureau of Health Insurance, an official far subordinate to the Secretary. ROM Pt. 6, at iv. Guidelines issued by someone without authority to promulgate regulations are simply not regulations. See General Electric Co. v. Gilbert, 429 U.S. 125, 141, 97 S.Ct. 401, 410, 50 L.Ed.2d 343 (1976). This alone conclusively refutes the panel's assertion that ROM § 2572-D was promulgated pursuant to the authority of Section 1396b(f) and is binding on the Secretary. 781 F.2d at 1427.
While the Secretary had made various delegations to the Administrator of HCFA by the time ROM section 2572-D was issued, he specifically retained the authority to issue regulations, 44 Fed.Reg. 31045 (May 30, 1979) (Statement of Organization, Functions, and Delegations of Authority).
But there is much more. The Secretary in fact has promulgated regulations on this very subject, regulations that plainly conflict with ROM § 2572-D and the panel's decision. See 42 C.F.R. §§ 435.811, .1007 (1985). These regulations provide that federal participation
is available in expenditures for services provided to medically needy recipients whose annual income . . . does not exceed the following amounts . . .: (a) For couples and families of two or more, 133-1/3 percent of the highest money payment that would ordinarily be made under the State's AFDC plan to a family of the same size without income and resources.
Id. § 435.1007 (emphasis added). The ambiguity the panel detected in the statutory language is neatly excised by these regulations. The regulations are a valid exercise of the Secretary's authority and therefore have the force of law. To conclude, as the panel in effect does, that the Regional Office Manual takes precedence, violates the axiom that federal regulations may not be contravened by subordinate employees who lack authority to bind the agency. Federal Crop Insurance Corp. v. Merrill, 332 U.S. 380, 384-85, 68 S.Ct. 1, 3-4, 92 L.Ed. 10 (1947). In any collision between regulations and instructions in an office manual, the regulations must prevail. Moreover, the very existence of conflicting regulations issued by the Secretary contradicts the panel's assertion that the anonymous ROM update represented the Secretary's exercise of authority pursuant to 42 U.S.C. § 1396b(f).
The panel's opinion quotes the regulations, 781 F.2d at 1424, but ignores them; it does not attempt to reconcile the obvious conflict between the regulations and ROM § 2572-D, the linchpin of its decision; it does not even acknowledge the tension between the two. The panel effectively reads the regulations out of existence without any explanation or admission of the fact.
Even if ROM § 2572-D had been promulgated as a regulation, it would be invalid because it violates the plain language of the statute. While section 1396b(f)(1) gives the Secretary discretion to determine how the MNIL amount is to be calculated, the calculation must in fact aim at approximating the AFDC income of a family of the same size. It renders language meaningless to say, as does the panel, that the Secretary may formulate the MNIL in terms of the amount to be paid to a family of a different size. Read in this fashion, the statute provides no limits whatsoever on the Secretary's discretion and might as well say "the Secretary may set the MNIL any way he pleases."
In certain limited circumstances section 1396b(f)(1)(B)(ii) allows the Secretary to set the MNIL amount higher than 133-1/3% of the amount paid to a family of the same size under a state's AFDC plan. But this exception applies only if a state's AFDC plan provides that some large families with different numbers of children get equal amounts. The panel renders this congressionally-approved exception a nullity, since under its holding the Secretary would have authority to reach the same result under (B)(i).
Troubling as I find the panel's result, it is its method, with its explosive potential, that causes me the greatest alarm. I had thought it firmly established that a court may not roam through the vast libraries of federal agency publications, borrowing from those it likes and ignoring the rest. Agency publications come in many forms and levels of dignity — from press releases to regulations promulgated after notice and comment; and they are issued by a variety of agency officials — from GS-15 field office chiefs to Cabinet Secretaries. Regulations properly issued by a Cabinet Secretary pursuant to authority delegated to him by Congress, I had thought, took precedence over all other agency issuances and, unless held to be invalid, conclusively resolve the legal issues they address. I had also thought that an agency could safely issue internal guidance to its employees without binding itself to outsiders by: (a) confining issuances to internal procedure manuals; (b) failing to attribute them to anyone with authority to promulgate regulations; and (c) expressly disclaiming the intent to promulgate regulations.
The panel's opinion tears asunder these tenets of administrative law, replacing them with nothing more substantial than its assurance that ROM § 2572-D is "an exception." 781 F.2d at 1426. But the panel gives no accurate account of what makes it an exception; everything it says about ROM § 2572-D could fairly be said about a great many other office manuals issued by federal agencies. Lawyers in the nine western states will now have a field day prospecting for other such nuggets within the thousands of pages of internal instructions issued by agencies regulating vast areas of the law: tax, personnel, transportation, agriculture, the environment, to name a few.
Having created uncertainty where stability is vital, the panel's methodology will breed litigation which will then beget further uncertainty. While this may be a bonanza for lawyers, and give immense power to courts to meddle in the affairs of the Executive Branch, it will cause no end of headaches for agency officials who will now have to guess which of their informal, unsigned, internal guidelines will be found "exceptional" enough to outrank regulations. That door, it seems to me, was slammed shut by the Supreme Court in cases such as Schweiker v. Hansen and Federal Crop Insurance v. Merrill. It should remain that way.
2. The Deficit Reduction Act (DEFRA)
A. Perhaps not entirely comfortable with its principal ruling, the panel offers an alternative: that the Secretary's action violated DEFRA's moratorium on certain regulatory actions. During the moratorium period, the Secretary may not disapprove state Medicaid plans if such denial is premised on the fact that they violate the so-called "same methodology" or "single standard" rule. As the Supreme Court recently explained in Atkins v. Rivera, ___ U.S. ___, 106 S.Ct. 2456, 2463, 91 L.Ed.2d 131 (1986), the "same methodology" rule deals with the question of what resources are counted toward income in determining Medicaid eligibility. The statute provides that a state must use the "same methodology" in establishing an applicant's eligibility for Medicaid as it would if he were applying for other welfare programs, typically AFDC and SSI. That is, if a resource (such as interest) is counted toward establishing eligibility for AFDC or SSI, it must also be counted for purposes of Medicaid. 106 S.Ct. at 2462.
As the Supreme Court further explained, in 1981 the Secretary promulgated regulations that weakened the "same methodology" rule by permitting the States "to use eligibility standards that were unrelated to the standards used in AFDC or SSI, as long as the standards were `reasonable.'" 106 S.Ct. at 2463 (footnote omitted). Congress disapproved this approach and in 1982 passed 42 U.S.C. § 1396a(a)(10)(C)(i)(III) as part of the Tax Equity and Fiscal Responsibility Act (TEFRA) for the sole purpose of invalidating these regulations. 106 S.Ct. at 2463. Two years later, Congress had second thoughts and passed the DEFRA moratorium in order to block the operation of TEFRA for a period of 18 to 30 months. Id. at 2463 n. 10. It is this latest chapter in the "same methodology" controversy on which the panel relies as its alternative basis for striking down the Secretary's ruling.
As Atkins makes clear, however, the DEFRA moratorium, and the "same methodology" question it addresses, has nothing at all to do with this case. The "same methodology" requirement deals with what resources are counted in defining income. The question here is how much income (however defined) a family may have and still be considered medically needy. The two provisions have about as much in common as apples and pineapples: They sound vaguely similar but they grow on entirely different statutory trees.
Whether alimony is considered for the purpose of establishing Medicaid eligibility is a "same methodology" question; whether the income exceeds the maximum a family of two or a family of three could receive under AFDC is not.
In addition to fundamentally misreading the statute, the panel errs by applying DEFRA's moratorium to something that is not yet a state plan, only a proposed plan. DEFRA, by its terms, prohibits the Secretary from taking any regulatory action against a "State's [Medicaid] plan." Until approved by the Secretary, California's proposed amendment does not become part of its plan and therefore cannot benefit from the moratorium. This interpretation comports not only with the language of the statute but with common sense as well. Under the panel's interpretation, states may raise the MNIL as high as they want during the moratorium and force the federal government to contribute. So construed, DEFRA removes significant federal fiscal control over Medicaid participation, risking a hemorrhage of federal funding for this program.
Finally, I have considerable difficulty with the panel's contorting of the Secretary's decision in order to strike it down. The Secretary quite properly relied on section 1396b(f) in disapproving California's proposed plan, not on section 1396a(a)(10)(C)(i)(III), the subject of the DEFRA moratorium. In order to find fault with the Secretary's action, the panel tells the Secretary that he acted with motives he denies having and pursuant to statutory powers he denies exercising. Such psychoanalysis of administrative decision-making far exceeds the bounds of judicial review. If the Secretary properly bases his decision on a particular statutory provision, as he did here, it is irrelevant that he could (or could not) have based his decision on something else.
In law, as in life, two wrongs add up to two wrongs, nothing more. Neither of the panel's chosen pathways lead to the result it reaches; they certainly cannot not do so in tandem. In reaching that result anyway, the opinion muddies the waters in two important areas of the law. I would take the case en banc, vacate the panel's opinion, affirm the Secretary's decision and set the law of the circuit back on course.