Beno v. Shalala
Opinions
Opinion by Judge GOODWIN; Dissent by Judge O’SCANNLAIN.
GOODWIN, Circuit Judge:
Plaintiffs, California residents who receive Aid to Families with Dependant Children (“AFDC”), appeal the denial of their request for a preliminary injunction enjoining California’s public benefits experiment, 853 F.Supp. 1195. Plaintiffs object to a statewide benefits cut enacted as part of an experimental work-incentive project and challenge the Secretary of Health and Human Services’ (“Secretary of HHS”)1 waiver of certain federal laws related to the project. They argue that the Secretary’s waiver violates the Administrative Procedures Act, (“APA”), 5 U.S.C. § 701 et seq. and § 211 of the HHS Appropriations Act, 42 U.S.C. § 3515b, which prohibits HHS from spending federal money on experimental projects which pose a danger to human research subjects without their informed consent. In addition, they contend that California’s project violates the Americans with Disabilities Act (“ADA”), 42 U.S.C. §§ 12131-12213, by failing to make reasonable accommodations for AFDC recipients with disabilities. We reverse.
I. CALIFORNIA’S PROGRAM
The benefits cut at issue in this appeal is part of a five-year Assistance Payments Demonstration Project (“APDP”) enacted at Cal.Welf. & Inst.Code § 11450.01 et seq. APDP includes both a “residency requirement” and a “work-incentive” program. The [1061]*1061former, which has been preliminarily enjoined on constitutional grounds, aims to discourage poor families from moving to California by limiting recent entrants’ AFDC benefits to the amount received in their state of former residence. See Green v. Anderson, 811 F.Supp. 516 (E.D.Cal.1993), aff'd, 26 F.3d 95 (9th Cir.1994). The latter, which is the subject of this appeal, aims to encourage AFDC recipients to find work by decreasing benefits and allowing recipients to keep more of their earned income.2
The work-incentive benefits cut affects all California AFDC families (approximately 826,000 families and 2.4 million persons, ER 367) without regard to family composition or disabilities,3 except for a “control group” of 5,000 families randomly selected from four counties. The control group receives AFDC benefits at their former levels and is subject to the old income-disregard rules. In order to assess the impact of the work-incentive program, the state plans to compare data about these control group families with data on 10,000 families randomly selected from the same four counties. The state does not plan to study most of the other approximately 800,000 families affected by the cut.
II. THE FEDERAL WAIVERS
The benefits program at issue, AFDC, is a cooperative federalism program created by the Social Security Act of 1935, 42 U.S.C. §§ 601-687. Participating states and the federal government jointly finance the program and state governments administer it under plans approved by the Secretary of Health and Human Services. Id. While states are not required to participate, participating states must comply with a variety of federal requirements. Alexander v. Choate, 469 U.S. 287, 289 n. 1, 105 S.Ct. 712, 714 n. 1, 83 L.Ed.2d 661 (1985); Rosado v. Wyman, 397 U.S. 397, 417, 90 S.Ct. 1207, 1220, 25 L.Ed.2d 442 (1970).
California concedes that APDP violates several of these requirements, including the “Maintenance of Effort” requirement of 42 U.S.C. § 1396a(c)(l).4 This section provides that:
the Secretary shall not approve any State plan for medical assistance if — (1) the state has in effect [AFDC] payment levels that are less than the payment levels in effect under such plan on May 1, 1988.
42 U.S.C. § 1396a(c)(l). California’s experiment reduces AFDC benefits to below their May 1988 levels. Thus, absent a waiver, the state could not implement the experiment without jeopardizing federal funding of its $14 billion Medicaid program. The California statutes enacting APDP, therefore, refer explicitly to obtaining HHS approval, and do not become effective until thirty days after state officials receive such approval.5
On September 9, 1992, California officials applied to HHS for a waiver of these various federal laws, pursuant to the Secretary’s au[1062]*1062thority under 42 U.S.C. § 1315(a), which provides that:
In the ease of any experimental, pilot, or demonstration project which, in the judgment of the Secretary, is likely to assist in promoting the objectives of subchapter I, X, XIV, XVI or XIX or this chapter, or Part A or D of subchapter IV of this chapter [the AFDC program], in a State or States—
(1) the Secretary may waive compliance with any of the requirements of section 302, 602, 654, 1202, 1352, 1382 or 1396a of this title, as the case may be, to the extent and for the period he finds necessary to enable such State or States to carry out such project.
California’s waiver application alleged that the instant benefits cut would promote the objectives of the Act by creating a “work-incentive” experiment, which would “encourage able-bodied adults” to find work. ER 1167. The state requested authority to make unlimited benefits cuts, and contended that the experiment would help determine whether such work-incentives could be effective in other states. Id.
The state’s application did not address the cut’s effect on AFDC recipients with disabilities or on child-only families, who could not respond to the work incentive. Nor did it explain why the state had decided to cut benefits to all families in order to study the response of a few recipients in four counties or offer any data about the cut’s potential impact on children or low-income families. Rather, it emphasized that the experiment will “save $420 million annually ($208 million federal, $202 million state and $10 million county, with administrative costs taken into account)” and asked HHS to take quick action on the waiver because “[g]iven the current fiscal condition of the State and Federal governments, it is imperative that these reforms be implemented as soon as possible.” ER 1167.
On September 29, 1992, less than two weeks after receiving California’s proposal, the Secretary faxed state officials a draft “Terms and Conditions” setting forth proposed conditions for approval of the experiment. The draft Terms and Conditions contains detailed provisions concerning “cost neutrality” (i.e.
Free access — add to your briefcase to read the full text and ask questions with AI
Opinion by Judge GOODWIN; Dissent by Judge O’SCANNLAIN.
GOODWIN, Circuit Judge:
Plaintiffs, California residents who receive Aid to Families with Dependant Children (“AFDC”), appeal the denial of their request for a preliminary injunction enjoining California’s public benefits experiment, 853 F.Supp. 1195. Plaintiffs object to a statewide benefits cut enacted as part of an experimental work-incentive project and challenge the Secretary of Health and Human Services’ (“Secretary of HHS”)1 waiver of certain federal laws related to the project. They argue that the Secretary’s waiver violates the Administrative Procedures Act, (“APA”), 5 U.S.C. § 701 et seq. and § 211 of the HHS Appropriations Act, 42 U.S.C. § 3515b, which prohibits HHS from spending federal money on experimental projects which pose a danger to human research subjects without their informed consent. In addition, they contend that California’s project violates the Americans with Disabilities Act (“ADA”), 42 U.S.C. §§ 12131-12213, by failing to make reasonable accommodations for AFDC recipients with disabilities. We reverse.
I. CALIFORNIA’S PROGRAM
The benefits cut at issue in this appeal is part of a five-year Assistance Payments Demonstration Project (“APDP”) enacted at Cal.Welf. & Inst.Code § 11450.01 et seq. APDP includes both a “residency requirement” and a “work-incentive” program. The [1061]*1061former, which has been preliminarily enjoined on constitutional grounds, aims to discourage poor families from moving to California by limiting recent entrants’ AFDC benefits to the amount received in their state of former residence. See Green v. Anderson, 811 F.Supp. 516 (E.D.Cal.1993), aff'd, 26 F.3d 95 (9th Cir.1994). The latter, which is the subject of this appeal, aims to encourage AFDC recipients to find work by decreasing benefits and allowing recipients to keep more of their earned income.2
The work-incentive benefits cut affects all California AFDC families (approximately 826,000 families and 2.4 million persons, ER 367) without regard to family composition or disabilities,3 except for a “control group” of 5,000 families randomly selected from four counties. The control group receives AFDC benefits at their former levels and is subject to the old income-disregard rules. In order to assess the impact of the work-incentive program, the state plans to compare data about these control group families with data on 10,000 families randomly selected from the same four counties. The state does not plan to study most of the other approximately 800,000 families affected by the cut.
II. THE FEDERAL WAIVERS
The benefits program at issue, AFDC, is a cooperative federalism program created by the Social Security Act of 1935, 42 U.S.C. §§ 601-687. Participating states and the federal government jointly finance the program and state governments administer it under plans approved by the Secretary of Health and Human Services. Id. While states are not required to participate, participating states must comply with a variety of federal requirements. Alexander v. Choate, 469 U.S. 287, 289 n. 1, 105 S.Ct. 712, 714 n. 1, 83 L.Ed.2d 661 (1985); Rosado v. Wyman, 397 U.S. 397, 417, 90 S.Ct. 1207, 1220, 25 L.Ed.2d 442 (1970).
California concedes that APDP violates several of these requirements, including the “Maintenance of Effort” requirement of 42 U.S.C. § 1396a(c)(l).4 This section provides that:
the Secretary shall not approve any State plan for medical assistance if — (1) the state has in effect [AFDC] payment levels that are less than the payment levels in effect under such plan on May 1, 1988.
42 U.S.C. § 1396a(c)(l). California’s experiment reduces AFDC benefits to below their May 1988 levels. Thus, absent a waiver, the state could not implement the experiment without jeopardizing federal funding of its $14 billion Medicaid program. The California statutes enacting APDP, therefore, refer explicitly to obtaining HHS approval, and do not become effective until thirty days after state officials receive such approval.5
On September 9, 1992, California officials applied to HHS for a waiver of these various federal laws, pursuant to the Secretary’s au[1062]*1062thority under 42 U.S.C. § 1315(a), which provides that:
In the ease of any experimental, pilot, or demonstration project which, in the judgment of the Secretary, is likely to assist in promoting the objectives of subchapter I, X, XIV, XVI or XIX or this chapter, or Part A or D of subchapter IV of this chapter [the AFDC program], in a State or States—
(1) the Secretary may waive compliance with any of the requirements of section 302, 602, 654, 1202, 1352, 1382 or 1396a of this title, as the case may be, to the extent and for the period he finds necessary to enable such State or States to carry out such project.
California’s waiver application alleged that the instant benefits cut would promote the objectives of the Act by creating a “work-incentive” experiment, which would “encourage able-bodied adults” to find work. ER 1167. The state requested authority to make unlimited benefits cuts, and contended that the experiment would help determine whether such work-incentives could be effective in other states. Id.
The state’s application did not address the cut’s effect on AFDC recipients with disabilities or on child-only families, who could not respond to the work incentive. Nor did it explain why the state had decided to cut benefits to all families in order to study the response of a few recipients in four counties or offer any data about the cut’s potential impact on children or low-income families. Rather, it emphasized that the experiment will “save $420 million annually ($208 million federal, $202 million state and $10 million county, with administrative costs taken into account)” and asked HHS to take quick action on the waiver because “[g]iven the current fiscal condition of the State and Federal governments, it is imperative that these reforms be implemented as soon as possible.” ER 1167.
On September 29, 1992, less than two weeks after receiving California’s proposal, the Secretary faxed state officials a draft “Terms and Conditions” setting forth proposed conditions for approval of the experiment. The draft Terms and Conditions contains detailed provisions concerning “cost neutrality” (i.e. whether the experiment would increase federal expenditures) and data collection, but does not address the need for a statewide cut, the proposed cut’s on children, or the need to cut benefits to disabled recipients and child-only AFDC units.
Plaintiffs’ counsel wrote HHS, raising these and other issues and asking HHS to delay approval of California’s project to enable plaintiffs to submit comments. HHS did not respond to this letter, but continued to exchange draft Terms and Conditions with California officials, adding various provisions about cancelling the program, federal funding, and data collection. The only change that addressed the project’s potential impact on AFDC recipients was the Secretary’s decision to limit the benefits cuts to 6.3% below 1988 levels. However, the Secretary gave no explanation for this figure and allowed the state to cut benefits to recent entrants by up to 80%.
On October 16, plaintiffs’ counsel sent a second letter, objecting to the benefits cut and residency requirement and submitting voluminous materials about the harm the cut would cause AFDC families and children. Counsel also noted that California could create a work incentive without cutting benefits — by simply changing the income-disregard rules — and could limit the cut to those recipients whose disabilities did not preclude work. ER 1332. According to plaintiffs, such reforms could be implemented without increasing state and federal expenditures. Id. (stating that Michigan had enacted such a program).
HHS made its final editorial changes to the draft Terms and Conditions on the day it received plaintiffs’ letter. These minor changes were not accompanied by any comment on plaintiffs’ objections or proposed alternatives. The state also submitted no response to plaintiffs’ comments. On October 29, after California submitted additional material on data collection, HHS granted California’s waiver request.
California implemented the 1.3% cut authorized by the Secretary’s waiver and Cal. Welf. & Inst.Code § 11450.01(b)(1) on De-[1063]*1063eember 1, 1992. Relying on the Secretary’s waiver, the legislature then authorized an additional 2.7% cut, which went into effect on September 1, 1993. Cal.Welf. & Inst.Code § 11450.015. Because California had previously reduced AFDC benefits by 4.5% and 4.4%, California’s AFDC benefits are now 4% below 1988 levels and some 11% below 1992 levels.6 The governor has proposed additional cuts to be implemented July, 1994.
Shortly after the first cut went into effect, plaintiffs, who are AFDC recipients subject to the benefits cuts and residency requirements, filed this class action against various state and federal officials. They sought declaratory and injunctive relief and a preliminary injunction vacating the Secretary’s waiver of the “maintenance of effort” requirement contained in 42 U.S.C. § 1396a(c)(l). Citing evidence that the proposed cut would inflict serious harm, including hunger and malnutrition, on low-income families and children, particularly given California’s unemployment rates and high cost of living,7 plaintiffs contended, inter alia, (1) that the Secretary’s waiver was “arbitrary and capricious” within the meaning of the APA, 5 U.S.C. § 701 et seq.; (2) that the experiment presented a danger to human research subjects such that federal funds cannot be spent on the project without participants’ informed consent, 42 U.S.C. § 3515b (§ 211); and (3) that California’s failure to make reasonable accommodations for recipients with disabilities violates the ADA, 42 U.S.C. §§ 12131-12213.
The district court denied plaintiffs’ motion for a preliminary injunction, finding that, while plaintiffs had shown “the possibility of irreparable harm,” they had not shown that any of these statutory claims were “likely to succeed on their merits” or that the balance of hardships tipped in their favor. Dist. Ct. Order at 38-39 853 F.Supp. at 1204-05.8 Plaintiffs timely appealed pursuant to 28 U.S.C. § 1292(a)(1).
III. STANDARD AND SCOPE OF REVIEW
In general, we review for abuse of discretion a district court order granting or denying a preliminary injunction. Miller v. California Pac. Medical Ctr., 19 F.3d 449, 455 (9th Cir.1994) (en banc); MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511, 516 (9th Cir.1993), cert. dismissed, — U.S. -, 114 S.Ct. 671, 126 L.Ed.2d 640 (1994).9 However, legal issues underlying the decision to grant or deny preliminary relief are reviewed de novo. Miller, 19 F.3d at 455; Inland Empire Pub. Lands Council v. Schultz, 992 F.2d 977, 980 (9th Cir.1993). Moreover, where the record is fully developed, the plaintiff “requested both preliminary and permanent injunctions on the issues being appealed, and the district court’s denial of injunctive relief rested primarily on interpretations of law, not on the resolution of factual issues,” we may “consider the merits of the case and enter a final judgment to the extent appropriate.” Sierra Club v. Marsh, 816 F.2d 1376, 1382 (9th Cir.1987) (citing cases); Friends of the Earth v. United States Navy, 841 F.2d 927, 931 (9th Cir.1988).
Here, the district court found, and defendants concede, that plaintiffs have shown “the possibility of irreparable harm.” Seer. Br. at 34, 853 F.Supp. at 1215.10 Both plain[1064]*1064tiffs and defendants concede that the only significant disputes in the case are legal, and that the record is adequate to address plaintiffs’ claims on their merits.11 The “parties ask us to interpret statutes and regulations in determining whether the district court erred,” Sierra Club, 816 F.2d at 1382, and “the district court’s denial of injunctive relief rested primarily on questions of law.” Friends of the Earth, 841 F.2d at 931. We therefore review de novo the district court’s legal conclusions and address plaintiffs’ claims on their merits.
IV. PLAINTIFFS’ APA CLAIM
Plaintiffs first contend that the Secretary’s decision to waive California’s compliance with the “Maintenance of Effort” requirement of 42 U.S.C. § 1396a(c)(l) is “arbitrary and capricious” and hence violates the APA, 5 U.S.C. § 706(2)(A). The Secretary counters (A) that plaintiffs lack standing to challenge the Secretary’s waiver under the APA; (B) that § 1316(a) waivers are not subject to APA review; (C) that § 1315(a) does not require her to consider the issues plaintiffs raise; and (D) that the Secretary’s decision was not “arbitrary and capricious” within the meaning of the APA. We address each of these arguments in turn.
A. Standing
We review de novo the district court’s determination that plaintiffs have standing. Ellis v. City of La Mesa, 990 F.2d 1518, 1523 (9th Cir.1993), cert. denied, — U.S. -, -, 114 S.Ct. 2707, -, 129 L.Ed.2d 834 (1994). To establish standing, a federal plaintiff must show (1) that he has suffered an “injury in fact”; (2) that there is a “causal connection between the injury and the conduct complained of’; and (3) that it is “ ‘likely,’ as opposed to merely ‘speculative,’ that the injury will be ‘redressed by a favorable decision.’ ” Lujan v. Defenders of Wildlife, — U.S. -, -, 112 S.Ct. 2130, 2136, 119 L.Ed.2d 351 (1992) (quoting Simon v. Eastern Kentucky Welfare Rights Org., 426 U.S. 26, 38, 96 S.Ct. 1917, 1924, 48 L.Ed.2d 450 (1976)).
As the Secretary concedes, plaintiffs have obviously met the injury and causal-connection prongs of this test. However, the Secretary contends that plaintiffs have not shown redressibility because, in her view, the APA gives this court jurisdiction only to remand to the Secretary for further review and neither § 1396a(c)(l) nor any other federal law makes California’s benefits cut illegal.12 [1065]*1065Therefore, she alleges, a favorable decision of this court might not restore plaintiffs’ benefits because (1) the Secretary might again grant California’s request for a waiver; (2) California might elect to continue the experiment at the risk of losing federal Medicaid funding; and (3) California might be able to continue the experiment without losing federal Medicaid money by characterizing its future Medicaid funding submissions as “plan amendments” rather than as new plans. Seer. Br. at 13 n. 12.
These arguments suffice to show that a favorable decision might not redress plaintiffs’ injury. However, to have standing, a federal plaintiff must show only that a favorable decision is likely to redress his injury, not that a favorable decision mill inevitably redress his injury. Lujan, — U.S. at -, 112 S.Ct. at 2136.13 Thus, the mere fact that, on remand, the Secretary might again issue a waiver does not defeat plaintiffs’ standing. See, e.g., Seattle Audubon Soc’y v. Espy, 998 F.2d 699, 702 (9th Cir.1993) (In a suit challenging an Environmental Impact Statement (EIS), the fact that “redrafting the EIS might not change the Secretary’s decision ... is not relevant to standing.”); Idaho Conservation League v. Mumma, 966 F.2d 1508, 1518 (9th Cir.1992) (same).
In this case, a favorable ruling is likely to redress plaintiffs’ injury. First, the California statutes explicitly require state agencies to obtain HHS approval and do not become effective until thirty days after HHS gives such approval. Arguably, then, while a favorable decision of this court might not make California’s benefits cut illegal under federal law, it might make it illegal under state law.14 It might also prevent California from implementing additional cuts.
Second, in Lujan and other cases addressing redressibility, the decision-making agencies were “not parties to the suit and [were therefor not] obliged to honor an incidental legal determination” of the suit. Lujan, — U.S. at -, 112 S.Ct. at 2141.15 However, in this case, the state decision-making agencies are parties to the action and will be bound by any legal conclusions this court or the district court may reach.
Finally, in contrast to Simon and Lujan, where invalidation of the contested decision would have only a minor effect on the relevant agencies’ budget,16 in the instant ease, California stands to lose more than $14 billion in federal Medicaid funding.17 The loss [1066]*1066of such funds would have a very significant impact on California’s budget. Given state defendants’ own statements and the political and fiscal realities in California, we doubt that state agents would risk losing such funding in order to reduce AFDC benefits. State memoranda describing APDP consistently emphasize its potential fiscal savings. See, e.g., ER 1167 (noting that the project could save $202 million in state funds and $10 million in county funds). Giving up $14 billion in order to save $202 million simply does not make sense, fiscally or otherwise — that is why the California statutes made the experiment contingent on federal approval in the first place.
While the Secretary suggests that California might evade § 1396a(e)(l) by submitting Medicaid plan amendments rather than Medicaid plans, § 1396a(e)(l) requires the Secretary to reject a state plan for medical assistance if the state’s AFDC payment levels are less than they were in May, 1988. Stowell, 976 F.2d at 69.18 The Secretary’s speculation that she and California might conspire to evade this mandatory language does not defeat plaintiffs’ standing.
B. Reviewability
The district court also correctly found that § 1315(a) waivers are subject to APA review. The Secretary contends that § 1315(a) waivers are “committed to agency discretion by law,” 5 U.S.C. § 701(a)(2), thus barring APA review. She emphasizes that § 1315(a) authorizes a waiver “to the extent and for the period [the Secretary] finds necessary” (rather than to the extent and period necessary) and allows waivers for projects which “in the judgment of the Secretary [are] likely to assist in promoting the objectives” of the Act (rather than for projects which will assist in promoting these objectives).19 42 U.S.C. § 1315(a)(1) (emphasis added). We disagree.
The APA embodies a “basic presumption of judicial review.” Lincoln v. Vigil, — U.S. -, -, 113 S.Ct. 2024, 2030, 124 L.Ed.2d 101 (1993) (quoting Abbott Labs. v. Gardner, 387 U.S. 136, 140, 87 S.Ct. 1507, 1511, 18 L.Ed.2d 681 (1967)).20 Absent an explicit statutory bar, judicial review of agency action is available except “in those rare instances where statutes are drawn in such broad terms that in a given case there is no law to apply,” Webster v. Doe, 486 U.S. 592, 599, 108 S.Ct. 2047, 2051, 100 L.Ed.2d 632 (1988),21 and “a court would have no meaningful standard against which to judge the agency’s exercise of discretion.” Heckler v. Chaney, 470 U.S. 821, 830, 105 S.Ct. 1649, 1655, 84 L.Ed.2d 714 (1985).
Whether any particular statute meets this standard is “statute specific and relates to the language of the statute and whether the general purposes of the statute would be endangered by judicial review.” Esmeralda v. Department of Energy, 925 F.2d 1216, 1218-19 (9th Cir.1991) (interpreting Webster and Heckler) (internal citations omitted). Thus, the mere fact that a statute contains discretionary language does not make agency action unreviewable. Id. In finding that the CIA’s decision to terminate an employee was not reviewable under the APA, the Webster Court relied not only on the discretionary language of the statute at issue but also on “the overall structure” of the statute, 486 U.S. at 600, 108 S.Ct. at 2052, and the fact that national security is “an area of executive action ‘in which courts have long been hesitant to intrude.’ ” Lincoln, — U.S. at -, 113 S.Ct. at 2031 (quoting Franklin v. Massachusetts, — U.S. [1067]*1067-,-, 112 S.Ct. 2767, 2785, 120 L.Ed.2d 686 (1992) (Stevens, J., concurring)).22
As the district court found, § 1815(a) does not implicate such traditionally unreviewable concerns. “[T]he granting of an exemption from statutory requirements is not an area of agency discretion traditionally unreviewable ... [and] it would be somewhat surprising were Congress to grant unreviewable discretion to the Secretary to exempt States from such an all encompassing series of statutory requirements.” 853 F.Supp. at 1205. Unlike the lump-sum appropriation found unreviewable in Lincoln, the AFDC program contains complex and detailed regulations and does not reveal a congressional commitment to the unfettered discretion of the Secretary. Lincoln, — U.S. at -, 113 S.Ct. at 2031-32 (lump-sum appropriation with little guidance suggests congressional intent to give the agency maximum flexibility and discretion). Judicial review of § 1315(a) waivers would not interfere with the purposes of § 1315(a) or “endanger” the statutory scheme of the Social Security Act. Esmeralda, 925 F.2d at 1219; cf. Bowen, 476 U.S. at 680-81, 106 S.Ct. at 2141 (HHS regulations promulgated under certain provisions of the Medicare program are subject to APA review).
Moreover, § 1315(a) provides a meaningful standard by which to judge the Secretary’s waiver. The statute does not give the Secretary unlimited discretion. It allows waivers only for the period and extent necessary to implement experimental projects which are “likely to assist in promoting the objectives” of the AFDC program; the AFDC program’s objectives are set forth with some specificity in 42 U.S.C. § 601. Cf. Webster, 486 U.S. at 600, 108 S.Ct. at 2052 (emphasizing that the statute “allows termination of an Agency employee whenever the Director ‘shall deem such termination necessary or advisable in the interest of the United States’ ”).23 Unlike in Webster, the limitations set forth in § 1315(a) and § 601 provide meaningful standards by which the Secretary’s decision can be judged.
Every court which has considered the issue has concluded that § 1315(a) waivers are subject to APA review.24 We agree.
C. The Secretary’s Obligation Under § 1315(a)
The Secretary next contends that, even if APA review is available, the instant waiver was not arbitrary and capricious because § 1315(a) does not require her to consider plaintiffs’ objections. She admits that plaintiffs and amici “raise serious concerns about California’s undertaking, including consequences stemming from the loss of income, [1068]*1068the failure to limit the study to those unable [sic] to work, and the statewide scope of the project” and “concedes that these matters deserve careful attention,” but insists that she is not required to give them such attention. Seer. Br. at 34. We cannot agree.
While we accord an agency’s interpretation of its statutory duties considerable deference, we simply cannot reconcile the Secretary’s argument with the plain language of § 1315(a), its legislative history, and the Social Security Act of which it is a part. Read in context, § 1315(a) plainly obligates the Secretary to evaluate the merits of a proposed state project, including its scope and its potential impact on AFDC recipients.
(1) § 1315 Must Be Read in Context
In interpreting § 1315, we must, of course, “follow the cardinal rule that a statute is to be read as a whole.” Conroy v. Aniskoff, — U.S. -, -, 113 S.Ct. 1562, 1565, 123 L.Ed.2d 229 (1993). “[T]he meaning of statutory language, plain or not, depends on context.” Id.
Section 1315(a) is a part of the Social Security Act, and allows the Secretary to waive a variety of federal regulations relevant not only to the AFDC program, but also to a variety of other categorical assistance programs serving elderly, blind, permanently disabled and medically needy persons.26 Each of these statutes, and the AFDC statutes in particular, contains numerous, de-tañed, specific requirements with which states must comply in order to receive federal funding. Whüe states are not required to participate in these federal programs, if they elect to participate, compliance with these regulations is mandatory.27 The Secretary is responsible for ensuring that state programs comply with these comprehensive federal regulations and is required to take certain specific steps, culminating with the loss of funding, when state plans faü to comply. See, e.g., 42 U.S.C. § 603. Moreover, under § 211 of the HHS Appropriations Act, 42 U.S.C. § 3515b, she must ensure that federal funds are not spent on experimental projects which pose a danger to human subjects’ physical, mental or emotional well-being.
This legislative scheme, with its mandatory language and detailed requirements, evidences a clear Congressional intent to take certain decisions away from state officials.28 In granting a § 1315(a) waiver, the Secretary allows the state to deviate from the minimum requirements which Congress has determined are necessary prerequisites to federal funding. While, ordinarily, the Secretary might reasonably argue that she ought to give state officials considerable discretion as to how to run a program, these federalism arguments have less weight in the context of a waiver of a congressional requirement. We are not examining the Secretary’s authority to interfere with state officials’ discretion, but rather her authority to waive compliance with federal statutes.
Whüe § 1315 obviously represents a congressional judgment that, in certain circumstances, such an override is-appropriate, we doubt that Congress would enact such comprehensive regulations, frame them in man[1069]*1069datory language, require the Secretary to enforce them, and then enact a statute allowing states to evade these requirements with little or no federal agency review. Rather, Congress intended that the Secretary would “selectively approve[ ]” state projects. S.Rep. No. 1589, 87th Cong., 2d Sess. 20, reprinted in 1962 U.S.C.C.A.N.1943, 1962. The Secretary’s own regulations and previous treatment of state projects confirm that she has plenary authority to reject state projects and to require states to modify projects to make them more consistent with federal requirements, less likely to harm recipients, and more likely to further the goals of the Social Security Act.29 Indeed, in approving California’s project, the Secretary exercised this authority to reject California’s proposed method of data collection and to require California to include a control group. ER 1192. Any argument, then, that the Secretary does not have authority to review the merits of a state proposal, or must simply accept whatever project the state submits seems not only “disingenuous,” Dist. Ct. Order at 18, but wholly inconsistent with congressional intent and the structure of the Social Security Act.
(2) The Plain Language of § 1315
Moreover, § 1315(a) plainly requires the Secretary to review state proposals. On its face, the statute allows waivers only (1) for experimental, demonstration or pilot projects, which (2) in the judgment of the Secretary are likely to assist in promoting the objectives of the Social Security Act and only (3) for the extent and period she finds necessary. Thus, while the Secretary has considerable discretion to decide which projects meet these criteria, she must, at a minimum, examine each of these issues.
(i) “An Experimental, Pilot or Demonstration Project.” First, § 1315(a) requires that the state project be an “experimental, demonstration or pilot” project. The statute was not enacted to enable states to save money or to evade federal requirements but to “test out new ideas and ways of dealing with the problems of public welfare recipients.” S.Rep. No. 1589, 87th Cong., 2d Sess. 20, reprinted in 1962 U.S.C.C.A.N. 1943, 1961. Thus, the Secretary must make some judgment that the project has a research or a demonstration value. A simple benefits cut, which might save money, but has no research or experimental goal, would not satisfy this requirement.30 Rather, the “experimental or demonstration project” language strongly implies that the Secretary must make at least some inquiry into the merits of the experiment — she must determine that the project is likely to yield useful information or demonstrate a novel approach to program administration.
(ii) “Likely To Assist in Promoting The Objectives” of the Act. Second, the Secretary must determine that the proposed project is likely to further the objectives of the AFDC program. The AFDC program was enacted:
[f]or the purpose of encouraging the care of dependent children in their own homes or in the homes of relatives by enabling each State to furnish financial assistance [1070]*1070... as far as practicable under the conditions in each State, to needy dependent children and their parents or relatives ... to help maintain and strengthen family life and to help such parents or relatives to attain or retain capability for the maximum self-support and personal independence consistent with the maintenance of continuing parental care and protection....
42 U.S.C. § 601.
As this statute, its legislative history, and courts have made clear, the AFDC program’s main objective is to support needy children. See, e.g., King v. Smith, 392 U.S. 309, 325, 88 S.Ct. 2128, 2137, 20 L.Ed.2d 1118 (1968) (“the protection of such children is the paramount goal of AFDC”).31 Thus, in determining that a state project is “likely to further the goals of the Act,” the Secretary must obviously consider the impact of the state’s project on the children and families the AFDC program was enacted to protect.32
Other federal laws and the Secretary’s own regulations support this interpretation. For example, § 211 of the HHS Appropriations Act, 42 U.S.C. § 3515b, requires the Secretary to ensure that HHS does not fund any research program or project which poses a danger to the physical, mental, or emotional well-being of a participant without the participant’s informed consent. HHS regulations implementing § 211 require most research involving human subjects to undergo independent Institutional Review Board (“IRB”) review. See 46 CFR § 46.101 et seq. Section 46.101(b)(5) of these regulations exempts research projects, including most public benefits research, “which are conducted by or subject to the approval of department or agency heads” from this requirement. According to the Secretary, IRB review of such projects is unnecessary because § 1315(a) “specifically provides that projects thereunder be consistent with the purposes of the program.” 48 Fed.Reg. 9266, 9269 (1983). Thus, IRB review of public benefits projects would be “duplicative and needlessly burdensome.” Id. This rationale implicitly acknowledges that § 1315(a) review includes an examination of the proposed project’s potential danger to participants’ physical, mental and emotional well-being. Other HHS comments have explicitly acknowledged this obligation.33
Second, as even the Secretary concedes, her duty to determine whether a state program will further the goals of the Act clearly includes some duty to examine the scope of the project. The scope of a proposed project is obviously relevant to its ability to further the objectives of the Act, including its potential to harm recipients. An experiment that eliminated AFDC benefits entirely in order to collect data about four families is less likely to further the objectives of the Act than one that reduces benefits slightly and collects data about all of the participants in the experiment. See CWRO v. Richardson, 348 F.Supp. 491, 498 (N.D.Cal.1972) (“[T]he Secretary would abuse his discretion if he were to approve a [§ 1315(a) ] project which ... subjected] an unreasonably large population to the experiment or continued] it for an unreasonably long period of time.”).
[1071]*1071Moreover, in determining that public benefits experiments need not undergo independent IRB review, the Secretary has indicated that “ethical and other problems raised by research in benefit programs will be addressed by the officials who are familiar with the programs and responsible for their successful operation.” 48 Fed.Reg. 9268 (1983). An inquiry into the ethical problems raised by research includes some evaluation of the risks the experiment poses, a review of alternative designs, and some effort to reduce risks “to those necessary to achieve the research objective.” The Belmont Report, 44 Fed.Reg. 23196 (1979). While the Secretary may defer to the state’s judgment about many aspects of the proposed experiment and exercises considerable discretion over what risks are necessary, she must make some determination that a project does not pose unnecessary risks to human subjects.
Plaintiffs contend that California’s experiment endangers needy children and is unnecessarily broad in scope — it cuts benefits statewide but collects data about only a few recipients and it imposes a “work-incentive” cut on recipients who cannot work due to disability or who are in child-only units. These objections address the project’s potential impact on welfare recipients and its scope. As the Secretary concedes, such “objections raise considerations of a type relevant to the Secretary’s decision under [§ 1315(a) ] concerning whether a proposed project serves the Act’s objectives.” Seer. Br. at 14.34 Thus, under both the Secretary’s reading of § 1315 and our independent review of the statute, the Secretary was required to consider plaintiffs’ objections.
(iii) Extent and Period Necessary. Finally, § 1315 requires the Secretary to issue a waiver only for the “extent and period she finds necessary.” The Secretary insists that this language does not obligate her to engage in any additional inquiry, once she determines that the state project furthers the goals of the Act. Rather, she contends, the phrase means only that she must determine which waivers are necessary to allow the state to implement an experiment of whatever extent and period the state has proposed. In her view, the “extent and period” inquiry is simply a nondiscretionary, rote review of which federal statutes conflict with the experiment and must be waived. According the Secretary, we must defer to this interpretation under the cases holding that agency interpretations of relevant statutes are entitled to deference.
However, such deference is not appropriate if an agency’s interpretation of a relevant statute conflicts with the statute’s plain meaning. Sullivan v. Everhart, 494 U.S. 83, 88-89, 110 S.Ct. 960, 964, 108 L.Ed.2d 72 (1990); Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 2781-82, 81 L.Ed.2d 694 (1984). Nor does it apply to statutory interpretations which appear to have been adopted for purposes of litigation, and which are not supported by any other evidence in the record. Alaniz v. Office of Personnel Management, 728 F.2d 1460, 1465 (Fed.Cir.1984). Finally, “[a]n agency interpretation of a relevant provision which conflicts with the agency’s earlier interpretation is ‘entitled to considerably less deference’ than a consistently held agency view.” INS v. Cardozo-Fonseca, 480 U.S. 421, 446 n. 30, 107 S.Ct. 1207, 1221 n. 30, 94 L.Ed.2d 434 (1987) (quoting Watt v. Alaska, 451 U.S. 259, 273, 101 S.Ct. 1673, 1681, 68 L.Ed.2d 80 (1981)).
Plaintiffs contend that the Secretary’s reading of the “extent and period” language conflicts with the statute’s plain meaning, was adopted specifically for purposes of this litigation, and is not consistent with her own regulations and previous actions. The district court agreed that the Secretary’s inter[1072]*1072pretation was “disingenuous” in light of the Secretary’s treatment of other proposed experiments, the detailed Terms and Conditions proposed in this experiment, and HHS regulations generally. 853 F.Supp. at 1207-OS. We also find nothing in the record or § 1315(a)’s 32-year history suggesting that the Secretary has previously interpreted the statute in this manner. Neither case law35 nor legislative history36 supports the Secretary’s reading. Moreover, the Secretary’s interpretation does not seem entirely consistent with the statute’s discretionary language,37 Congress’s clear intent to require states to adhere to federal guidelines except in narrow circumstances, or the Secretary’s statutory obligations to oversee AFDC programs and to ensure that federal money is not spent on projects which endanger human subjects.38 In light of the Secretary’s broad discretionary authority over federal and state cooperative programs and her obligation to enforce Congress’s comprehensive regulations, we have difficulty discerning a congressional purpose to require her to waive any and all federal regulations whenever a state proposes an “experiment” which has some ability to further the goals of program.
However, we need not resolve this issue of statutory interpretation or determine the precise meaning of § 1315(a)’s “extent and period” language. The Secretary concedes that plaintiffs’ objections are relevant to her inquiry under the first prong of the § 1315 inquiry — whether the state project is likely to further the objectives of the Act. Thus, regardless of the precise meaning of the “extent and period” clause, the issues plaintiffs raise are relevant factors for purposes of the § 1315(a) waiver decision. We must therefore determine whether the administrative record is sufficient to show that the Secretary in fact considered them.
D. The Administrative Record
Plaintiffs have indisputably shown that California’s experiment has serious problems, both as an experiment and as an attempt at welfare reform. The actors in this case — including the Secretary, the district court, and impressively credentialed am-ici 39 — agree that both the statewide scope of [1073]*1073the benefits cut as well as the decision to cut benefits to individuals who cannot work appear wholly unjustified by any legitimate experimental goal. State officials have advanced no such experimental goal, and we are unable to explain how it would advance social science to cut benefits to recipients who are not even included in the study. Amici accurately observe that such a design is “methodologically indefensible” in that it exposes a large number of subjects to potential harm, yet studies only a few. Amici Br. at 24.
Moreover, the idea of imposing a work-incentive benefits cut on individuals whose disabilities preclude work can only be called absurd. As the district court found, “the Demonstration Project was intended to create work-incentives for recipients able to work; the effect of the project on those disabled recipients who are unable to work appears unintended and serves no stated goal of the project.” 853 F.Supp. at 1213. “[T]he State could exclude from the benefit cut those of the disabled who are unable to work” and “[i]t would be humane to do so.” Id. at 1214.40
Nearly everyone also agrees that California’s experiment will put “child-only” AFDC families, AFDC families headed by adults who are too disabled to work, and families whose heads are unable to find work — or cannot work due to child-care, transportation, and other difficulties — at increased risk of homelessness, inadequate nutrition, and a variety of emotional and physical problems. Amici Br. at 21; Seer. Br. at 34. The program offers no work-training, child-care, or any other assistance designed to enable recipients to find and keep jobs. Moreover, given the minimum level of benefits already paid to AFDC families, it is difficult to imagine that the benefits cut would radically change the existing incentives to work. As plaintiffs point out, California could have accomplished its goal of increasing recipients’ incentive to work without cutting AFDC benefits at all, by simply allowing recipients to keep more of their earned income.
Given these various problems, plaintiffs’ expert contends that California’s experiment is “utterly unconscionable,” ER 51, and amici argue that it “flagrantly disregards the basic norms of research.” Amici Br. at 3.
However, we have no jurisdiction to review the wisdom of California’s undertaking. The APA does not give this court power “to substitute its judgment for that of the agency” but only to “consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgement.” Overton Park, 401 U.S. at 416, 91 S.Ct. at 823-24. We may reverse only if the decision was “arbitrary and capricious” within the meaning of the APA, 5 U.S.C. § 706(2)(A), in that
the agency has relied on factors which Congress has not intended it to consider, 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 Mfr. Ass’n v. State Farm Ins., 463 U.S. 29, 44, 103 S.Ct. 2856, 2867, 77 L.Ed.2d 443 (1983).
In examining this issue, we may not consider reasons for agency action which were not before the agency. Bowen v. American Hosp. Ass’n, 476 U.S. 610, 627, 106 S.Ct. 2101, 2112, 90 L.Ed.2d 584 (1986). Although we may “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned,” Motor Vehicle, 463 U.S. at 43, 103 S.Ct. at 2867, we cannot infer an agency’s reasoning from mere silence or where the agency failed to address significant objections and alternative proposals. Id. at 57, 103 S.Ct. at 2874.41 Rather, “an [1074]*1074agency’s action must be upheld, if at all, on the basis articulated by the agency itself.” Id. at 50, 103 S.Ct. at 2870.42 Thus, while formal findings are not required, the record must be sufficient to support the agency action, show that the agency has considered the relevant factors, and enable the court to review the agency’s decision. Florida Power & Light Co. v. Lotion, 470 U.S. 729, 744, 105 S.Ct. 1598, 1607, 84 L.Ed.2d 643 (1985).
In the present case, the record contains a rather stunning lack of evidence that the Secretary gave plaintiffs’ objections any such consideration. Except for the Secretary’s conclusory letter to plaintiffs’ counsel and possibly her decision to limit California’s authority to cut benefits, the record contains no evidence that the Secretary ever considered the danger California’s benefits cut would pose to recipients, the state’s decision to impose a statewide benefits cut, the need for cutting benefits as a work-incentive, the merits of imposing a work-incentive cut on individuals whose disabilities preclude work, or the feasibility of excluding individuals who receive federal disability benefits or have already been adjudged unable to work in the context of other government programs such as California’s Greater Avenue for Independence (“GAIN”) program. Neither the Secretary nor California ever responded to the substance of plaintiffs’ objections, and the Secretary did not revise the Terms and Conditions at all in response to plaintiffs’ comments.
Nonetheless, the district court found the record adequate (1) because the court believed the record was similar to that which the Second Circuit found adequate in Aguayo, 473 F.2d 1090; (2) because plaintiffs submitted their objections before the Secretary issued her decision; and (3) because the court believed that equity and intercounty migration justified the experiment’s scope. 853 F.Supp. at 1207-09. We disagree.
First, the Aguayo record included not only the Secretary’s decision and the plaintiffs’ objections, but also a state memoranda answering plaintiffs’ objections. Aguayo, 473 F.2d at 1106 (“the State submitted two mem-oranda ... identifying the objections and ... endeavoring to answer them”).43 In upholding the Secretary’s decision (as to most but not all of the experiment) the Court noted that New York’s attempt to refute plaintiffs’ objections was “successful at least to a substantial degree.” Id. Moreover, the Secretary required the state to submit additional information relevant to many of plaintiffs’ objections, including the quality of the childcare facilities the state would offer, the fair hearing procedures, and a statement that the state would reimburse recipients’ work-related expenditures. Id. at 1097. The instant case contains no such memoranda, no request for additional information related to the project’s impact on recipients, no statement explaining the need for a statewide benefits cut, and no indication that the Secretary had any information refuting plaintiffs’ substantial documentary evidence about the benefit cut’s danger to human subjects.
Second, a court should not infer that an agency considered an issue merely because it was raised, where there is no indication that the agency or other proponents [1075]*1075refuted the issue. See, e.g., National Wildlife Fed’n v. FERC, 801 F.2d 1505, 1512 (9th Cir.1986) (vacating and remanding for further consideration of petitioners’ objections where, as here, the agency “simply did not mention the extensive and uncontradicted evidence offered by petitioners” or explain its rejection of the options they proposed).
Here, the record contains no evidence that the Secretary considered the materials plaintiffs submitted. While the state and HHS exchanged detailed drafts of the Terms and Conditions, these drafts did not address any of plaintiffs’ objections. Even if these drafts show, as the district court found, that a “good deal of thought went into approval of the research design,” 853 F.Supp. at 1208, they do not show that HHS gave any such thought to plaintiffs’ objections or proposed alternatives. If anything, the timing of the waiver approval — HHS’s final changes to the draft Terms were made on the very day it received plaintiffs’ objections — suggests the opposite. Cf. Asarco, Inc. v. U.S.E.P.A., 616 F.2d 1153, 1162 (9th Cir.1980) (decision arbitrary and capricious where there was evidence agency had considered issues but did so in a manner which was insufficient and abstract).
Thus, district court erred in inferring that the Secretary considered plaintiffs’ objee-tions merely because plaintiffs submitted them. Id. at 1206-07 (citing Aguayo, 473 F.2d at 1105-06). Such an inference is inappropriate and would provide a perverse disincentive to advocates to refrain from raising relevant issues during the administrative process.
The district court similarly erred in inferring that the statewide scope of the project was necessary “for reasons of equity and to avoid the possible movement of AFDC recipients from counties covered by the project to those that were not.” 853 F.Supp. at 1206-07. Neither the state nor the Secretary ever mentioned these issues during the administrative process and nothing in the administrative record suggests that the Secretary in fact considered them. Thus, even if such factors were to justify California’s experiment,44 they cannot be considered. Bowen, 476 U.S. at 627, 106 S.Ct. at 2112.
The Secretary’s letter to plaintiffs’ counsel stating that she “considered the issues [plaintiffs] raised” is similarly insufficient. ER 1497.45 “Stating that a factor was considered ... is not a substitute for considering it.” See Getty v. Federal Savs. & Loan Ins. Corp., 805 F.2d 1050, 1055 (D.C.Cir.1986) (rejecting as “conclusory” a similar agency statement that all relevant factors had been considered).46
[1076]*1076Rather, the only indication that the Secretary considered the project’s potential impact on human subjects is her decision to limit California’s authority to cut benefits to 6.3%. This evidence, while limited, might ordinarily be sufficient to conclude that the Secretary in fact considered the project’s potential impact on AFDC recipients. We might fairly infer, from this decision, that she somehow determined that a cut of this magnitude would not pose a danger to human subjects.
However, the Secretary also approved cuts of up to 80% to recent entrants and chose the 6.3% cut-off without examining any data about the cost of living in California or other issues relevant to the danger determination. See National Treas. Employees Union v. Horner, 854 F.2d 490, 499 (D.C.Cir.1988) (agency action arbitrary where the record did not include “data of the sort [the agency] would had considered if it had considered [the issue] in any meaningful way”). Moreover, the Secretary’s decision to limit the cut does not explain its application to people who, according to the state’s own judgment, should be exempted from other work requirements and who are entitled to collect disability benefits. Neither the state nor HHS has offered any explanation as to why such persons were included in the “work-incentive” cut, or any evidence that it would be “administratively infeasible” to exempt them.
In the face of this extraordinarily sparse administrative record and the Secretary’s concession that the issues plaintiffs and amici raise deserve attention, we must REVERSE. The Secretary’s waiver of 42 U.S.C. § 1396a(c)(l) is VACATED 47 and the case is REMANDED to the district court with instructions to remand to the Secretary for additional consideration of plaintiffs’ objections.48
REVERSED AND REMANDED.
30 F.3d 1057 (Beno v. Shalala) — published by Counsel Stack Legal Research, free access to 12M+ legal documents.