Summary
holding that a state is not a "person" subject to suit under the FCA
Summary of this case from U.S. ex Rel. Settlemire v. Dist. of ColumbiaOpinion
No. 98-5133, 98-5149, 98-5150
Argued January 14, 1999
Decided April 2, 1999
Appeals from the United States District Court for the District of Columbia (92cv02092).
Howard L. Zwickel, Assistant Attorney General, State of New York, argued the cause for appellant/cross-appellee. With him on the briefs was Peter H. Schiff, Deputy Solicitor General.
Ronald A. Shems, Assistant Attorney General, State of Vermont, argued the cause for amici curiae State of Vermont, et al. With him on the brief was William H. Sorrell, Attorney General.
Douglas N. Letter, Appellate Litigation Counsel, United States Department of Justice, argued the cause for United States as intervenor. With him on the briefs were Frank W. Hunger, Assistant Attorney General, and Wilma A. Lewis, United States Attorney. Richard L. Cys entered an appearance.
Stuart F. Pierson argued the cause and filed the briefs for appellee/cross-appellant.
Jill A. Dunn was on the notice of joinder in brief for appellant Joseph P. Frey.
Mark B. Rotenberg was on the brief for amicus curiae The Regents of the University of Minnesota.
Before: WALD, SILBERMAN, and SENTELLE, Circuit Judges.
Opinion for the Court filed by Circuit Judge Silberman.
The question presented in this appeal is whether states are defendant persons under the False Claims Act. Contrary to the decisions of the Second and Eighth Circuits, see United States ex rel. Stevens v. Vermont Agency of Natural Resources, 162 F.3d 195 (2d Cir. 1998); United States ex rel. Zissler v. Regents of the Univ. of Minn., 154 F.3d 870 (8th Cir. 1998), we hold that they are not.
I.
Ronald Long was the Coordinator of Investigations and Audit for the Bureau of Proprietary School Supervision of the New York State Department of Education, the state agency that regulates proprietary schools. In 1989, he conducted an investigation of SCS Business and Technical Institute, which operates five business and technical schools in New York City, and discovered that SCS allegedly had made false and fraudulent claims to the federal government in return for federal funding for students attending SCS schools under tuition assistance programs. He also determined, according to his complaint subsequently filed in district court, that Joseph P. Frey, his supervisor at the Bureau, and other officials in the State Department of Education, knew about SCS's fraudulent claims and conspired with SCS to conceal the fraud in order to secure further federal funding for SCS. They did so because, after a 1990 change in New York State law, the Bureau's funding depended in substantial part on tuition assessments and fines that SCS paid to the Bureau. Long's theory was that since the Bureau received a share of the federal funds that SCS fraudulently obtained from the United States, the Bureau had every incentive to see that fraud continue. He claims that after he reported the results of his investigation to state and federal authorities, Frey and other state officials took actions to limit and subvert his investigation.
Long was taken off the investigation and then fired in 1992, shortly after SCS settled administrative charges brought against one of SCS's schools by the state education department. According to him, the settlement agreement, which did not benefit the United States in any way and grossly understated the extent of SCS's fraudulent practices, was a sweetheart deal that was but another instance of the state's conspiracy with SCS to conceal and perpetuate SCS's fraud — a conspiracy that he alleges continued until SCS filed for bankruptcy in 1995. He alleges that after the settlement, New York ignored evidence of SCS's continuing fraud and falsely represented to the United States that SCS's fraud had ceased and that it was actively monitoring SCS.
Long filed a complaint in the district court against Frey, other state officials, the State of New York, SCS, and various SCS officials. He brought his case as a qui tam relator under the False Claims Act, 31 U.S.C. §§ 3729 et seq. (1994), suing in the name of the United States for the benefit of the United States and himself. He contended that the state defendants violated the Act by conspiring with SCS to have false claims submitted to the United States and by causing false claims to be submitted. The state defendants were also alleged to have violated the whistle-blower provision of the Act by harassing and wrongfully discharging Long, and to have been unjustly enriched under state common law. The United States (the government) subsequently intervened in the case against the SCS defendants, but declined to intervene against the state defendants. The state defendants moved to dismiss the complaint on the grounds that states are not defendant persons under the Act and that, even if they were, the Eleventh Amendment to the United States Constitution would bar the suit. It was also asserted that Long's suit against the state defendants was barred by the Act because the allegations of fraud had been publicly disclosed and because Long was not an "original source" of the information.
The district court denied in part the state defendants' motion to dismiss, concluding that states are defendant persons under the Act and that the Eleventh Amendment does not bar the suit. See United States ex rel. Long v. SCS Bus. Technical Inst., 999 F. Supp. 78 (D.D.C. 1998). The state defendants filed an interlocutory appeal challenging the district court's rejection of their Eleventh Amendment defense, over which we have jurisdiction under 28 U.S.C. § 1291 (1994) and the collateral order doctrine. See Puerto Rico Aqueduct Sewer Auth. v. Metcalf Eddy, Inc., 506 U.S. 139, 144-45 (1993). We exercise pendent appellate jurisdiction over the "inextricably intertwined" statutory question, Gilda Marx, Inc. v. Wildwood Exercise, Inc., 85 F.3d 675, 679 (D.C. Cir. 1996) (quoting Swint v. Chambers County Comm'n, 514 U.S. 35, 51 (1995)), of whether states are defendant persons under the Act. Thirty-six states join as amici curiae in support of appellant New York's statutory and Eleventh Amendment arguments, and appellee Long, the relator, is joined by the government as intervenor defending the constitutionality of the Act.
The district court granted the motion to dismiss Long's whistleblower and unjust enrichment claims, the former because the Eleventh Amendment bars private suits brought against the state (although it does not bar Long's claim for prospective relief against Frey, a state official), and the latter because Long has no standing to assert the government's claim of unjust enrichment under state common law. See Long, 999 F. Supp. at 91-93.
The district court also concluded that Long's suit was not barred by the public disclosure and original source provisions of the Act. See Long, 999 F. Supp. at 87-89. Although the parties challenge aspects of those rulings on appeal, we need not address them further given our resolution of the case in favor of New York. We also decline to exercise pendent appellate jurisdiction over the statutory whistle-blower and constitutional claims against appellant Frey in his individual capacity. Although these claims are not foreclosed by anything in our opinion, they are not in any way related to the Eleventh Amendment and statutory construction questions that we decide today. And although an inextricable relation between claims is not a necessary condition for pendent appellate jurisdiction, see Jungquist v. Sheikh Sultan Bin Khalifa Al Nahyan, 115 F.3d 1020, 1027 (D.C. Cir. 1997), and efficiency interests might counsel in favor of resolving these claims now, we could not possibly terminate the entire case against Frey — even if we agreed with him — because Long also asserted § 1983 claims against him that the district court did not dismiss and from which Frey does not now seek to appeal. That, coupled with the otherwise unappealable nature of the order as to Frey, a separate appellant, see Gilda Marx, 85 F.3d at 678, and the presence of factual disputes in the briefs on the "original source" and "public disclosure" questions, see id. at 679, leads us to reject Frey's request that we resolve these claims now.
II.
To persuade us to uphold the decision below, appellees Long and the government must demonstrate that the district court correctly interpreted the term "person" (liable for making a false claim) in § 3729(a) of the False Claims Act to include states. In that respect, they have no little burden because the statute does not define the term "person" and, as the Supreme Court has remarked before, "in common usage, the term `person' does not include the sovereign, [and] statutes employing the [word] are ordinarily construed to exclude it." Will v. Michigan Dep't of State Police, 491 U.S. 58, 64 (1989) (quoting Wilson v. Omaha Indian Tribe, 442 U.S. 653, 667 (1979) (quoting United States v. Cooper Corp., 312 U.S. 600, 604 (1941))) (alteration in original); see also, e.g., Georgia v. Evans, 316 U.S. 159, 161-62 (1942).
The statute provides, in relevant part:
(a) Liability for certain acts. Any person who — . . .
(2) knowingly makes, uses, or causes to be made or used, a false record or statement to get a false or fraudulent claim paid or approved by the Government;
(3) conspires to defraud the Government by getting a false claim allowed or paid . . . is liable to the United States Government. . . .
31 U.S.C. § 3729 (1994).
As appellees observe, the Eighth Circuit rejected application of this rule in interpreting the False Claims Act on the ground that the presumption of sovereign exclusion applies only to the enacting sovereign. See Zissler, 154 F.3d at 874. However, the Court in Will applied this rule even though the enacting sovereign (the United States) was different from the state sovereigns excluded from the term person, see Will, 491 U.S. at 64, implicitly rejecting the Eighth Circuit's position as it was then articulated in Justice Brennan's dissent, see id. at 73 (Brennan, J., dissenting).
This "often-expressed understanding," Will, 491 U.S. at 64, is not a "hard and fast rule of exclusion," Wilson, 442 U.S. at 667 (quoting Cooper, 312 U.S. at 604-05), and depends in important part on the "context, the subject matter, legislative history, and executive interpretation," id. — which sounds like rather garden variety statutory interpretation. But if the Will-Wilson rule has any meaning at all, it must create at minimum a default rule states are excluded from the term person absent an affirmative contrary showing. See International Primate Protection League v. Administrators of Tulane Educ. Fund, 500 U.S. 72, 83 (1991) (noting that the "conventional reading" of person to exclude states may be "disregarded" if there is an affirmative showing of Congress' intent to include them). This interpretive principle, the Supreme Court tells us, is "particularly applicable" where, as here, "it is claimed that Congress has subjected the states to liability to which they had not been subject before." Will, 491 U.S. at 64; see also Wilson, 442 U.S. at 667. We think, therefore, that the district court had it backwards when it concluded that it found "no indication that Congress sought to create an exception for state actors to perpetrate fraud upon the federal government." Long, 999 F. Supp. at 85.
In reaching this conclusion, the district court was guided by its assumption that the "clear statement" rule of Will, 491 U.S. at 65, did not apply — an issue which we take up below. But the district court incorrectly equated Will's "clear statement" rule with the traditional rule presuming that the term person does not include states. Compare Will, 491 U.S. at 65 (clear statement rule), with id. at 64 (default rule that person does not include states). Even if the former rule were not implicated here, the latter rule — which all parties concede applies-dictates a presumption opposite to the one the district court applied.
Our review of the "legislative environment," Evans, 316 U.S. at 161, leads us to doubt appellees have met their burden. As we noted, neither the Act as currently written nor as originally passed in 1863 defines the term person. Indeed, the original Act distinguished for punishment purposes between fraudulent acts committed by "any person in the land or naval forces of the United States," Act of March 2, 1863, 37th Cong., 3d Sess., ch. 67, § 1, 12 Stat. 696, and "any person not in the military or naval forces of the United States," id. at § 3, 12 Stat. 698. Since states would not have been thought to fall within either classification, that Act can hardly be said to supply facially the requisite affirmative showing that the Will-Wilson default rule requires.
The Second Circuit explained this problem away by reasoning that the Congress; undeniable intent to include military contractors in the Act refuted any attempt to read "persons not in the military" as impliedly referring only to natural, as opposed to corporate, persons. See Stevens, 162 F.3d at 205-06. The default rule of statutory construction governing corporations as "persons," however, is precisely the opposite of the default rule that we must apply in this case. See Wilson, 442 U.S. at 666 (stating that the "word" person; for purposes of statutory construction, unless the context indicates to the contrary, is normally construed to include" corporations). Since we must look for an affirmative intent to include states, that contractors, under the default rule for corporations, could have been thought to be "person[s] not in the military" is hardly supportive of appellees; case.
Appellees nevertheless invoke the broad purposes and legislative history of the Civil War statute. We think that is not helpful because, as the Supreme Court has said, Congress' primary concern at the time — admittedly not its exclusive one — was to put an end to "frauds perpetrated by large [military] contractors during the Civil War." United States v. Bornstein, 423 U.S. 303, 309 (1976); see United States ex rel. Graber v. City of New York, 8 F. Supp. 2d 343, 352 (S.D.N Y 1998). Appellees point to the Supreme Court's statement that Congress sought to "reach all types of fraud, without qualification, that might result in financial loss to the Government." United States v. Neifert-White Co., 390 U.S. 228, 232 (1968) (holding that the term "claim" was not limited to claims submitted for payments due and owing from the government, but included claims for favorable action by the government upon applications for loans). But we think that description is too general — it was also made in an entirely different context — to answer the serious question whether states were made potential defendants under the Act. (According to appellees' reasoning, foreign governments that entered into commercial dealings with the United States would also be potential defendants.) Similarly unpersuasive is the policy proposition put forward by the Eighth Circuit, see Zissler, 154 F.3d at 874, that a truly effective anti-fraud statute would subject states to liability since states receive substantial amounts of money from the federal government. See also John T. Boese, Civil False Claims and Qui Tam Actions, at 2-91 (1993) (stating that states can be defendant persons because they are "major recipients of federal funds"). A court looks to legislative purpose under the default rule in order to locate a congressional intent "to bring state or nation within the scope of the law," Cooper, 312 U.S. at 605, not to "engraft on a statute additions which [the court] think[s] the legislature logically might or should have made," id. Even if one assumes that states commit a good deal of fraud against the federal government, it cannot seriously be argued that the very purpose of the Act would be thwarted if states were not liable under the Act. Compare California v. United States, 320 U.S. 577, 585 (1944).
Of course, Stevens, not Graber, is Second Circuit law.
In Zissler, 154 F.3d at 874, the Eighth Circuit relied on United States v. California, 297 U.S. 175, 186 (1936), for the proposition that it would be a mistake to exclude the states from an "act of Congress, all-embracing in scope and national in its purpose, which is as capable of being obstructed by state as by individual action," id. But the Supreme Court made that statement only after it had "fairly . . . inferred" that the purpose of the Federal Safety Appliance Act, albeit implicit, was to subject state-run railroads to liability. See id. If the mere use of the term person in a broad statute with national purposes, which states were equally capable of violating, were sufficient to bring the states within the statute's scope, the interpretive rule presuming the opposite would be largely ineffectual, if not wholly eviscerated.
That takes us to the legislative history. Appellees point us first to an 1862 House Committee Report that, in discussing various frauds committed during the Civil War, referred to certain state officials that had used war contracts for personal profit. See H.R. Rep. No. 2, 37th Cong., 2d Sess., at xxxviiixxxix (1862). But the report specifically stated that these examples of fraud were not committed against the United States government. See id. at xxxviii. So the prior report is a rather tenuous link to the Act Congress passed one year later. But see Stevens, 162 F.3d at 206 (concluding that "it is difficult to suppose" that Congress "had forgotten the results of this extensive investigation" when it passed the False Claims Act) (emphasis added). Even if there were a stronger tie, the Supreme Court has held that legislative history indicating an intent to impose liability on state officials is not evidence of an intent to subject the states themselves to liability. See Will, 491 U.S. at 68-69. The bottom line is that appellees have not pointed to anything in the legislative history of the 1863 Act, or in the events leading up to it, indicating that Congress actually contemplated imposing liability on the states.
Because the enacting Congress' intent is, to be charitable, rather opaque, appellees turn our attention to the 1986 amendments to the False Claims Act and to a related statute also passed in 1986. The provision of the 1986 amendments that changed 31 U.S.C. § 3729(a) from imposing liability on "[a] person not a member of an armed force of the United States" to "[a]ny person" did not, however, substantively expand the meaning of defendant persons under the Act. See Stevens, 162 F.3d at 206-07 (holding that states are persons but conceding that this change was not "envisioned as broadening the class of persons who could be held liable under the Act"); Graber, 8 F. Supp. 2d at 354-55. It is true that the amendment expanded the types of individuals subject to the Act to include those in the military. Still, that change tells one nothing about the basic meaning of the term person, or more specifically, whether Congress intended to include states within that term. The legislative history accompanying the amendment reveals Congress' extremely limited objective. See S. Rep. No. 345, 99th Cong., 2d Sess., at 17-18 (1986), reprinted in U.S.C.C.A.N. 5266, 5282-83 (explaining that the alteration of § 3729(a) was intended to provide for monetary recovery against persons in the military and that, prior to 1986, a court martial was the only available remedy). It is understandable, therefore, why appellees do not actually claim that states were made defendant persons by virtue of the 1986 amendment to § 3729(a). Instead, their argument is that states have been defendant persons all along various provisions added by the 1986 Congress — which we discuss below — simply make that clear. In other words, appellees, by relying on these recent amendments, seek to illuminate the 1863 Congress' "original intent." We are rather dubious about such an approach. As the Supreme Court has observed, such subsequent provisions are really "beside the point" because they do not "reflect any direct focus by Congress upon the meaning of the earlier enacted provisions." Almendarez-Torres v. United States, 118 S.Ct. 1219, 1227 (1998); Atkinson v. Inter-American Dev. Bank, 156 F.3d 1335, 1342 (D.C. Cir. 1998).
The Eighth Circuit thought that this amendment more broadly "evidenced consideration of whom to hold liable" under the amended Act. Zissler, 154 F.3d at 874. But there is nothing in the text of the statute or in any of the legislative history indicating that Congress' consideration of "whom to hold liable" extended beyond its intent, expressed in the statute, to bring military persons within the scope of the Act.
Be that as it may, we are not persuaded that these added provisions can bear the weight appellees would place on them. Appellees argue that Congress' decision to define "person" to include states in the Civil Investigative Demand section of the Act, see 31 U.S.C. § 3733(1)(4) (1994), indicates (some) Congress' intent to include states as persons throughout the whole Act, even though this provision applies only to the Civil Investigative Demand section. See 31 U.S.C. § 3733(l) (For purposes of this section . . .) (emphasis added). Appellees question why Congress would create a discovery tool to be used to gain information possessed by states if the Act did not already authorize false claims actions against them. See also Stevens, 162 F.3d at 207. It seems rather obvious, however, that states could provide useful evidence to establish that private contractors, for example, made false claims. Nor do appellees gain very much by pointing to the Program Fraud Civil Remedies Act, 31 U.S.C. § 3801 et seq. (1994), which Congress also passed in 1986 to create an alternative administrative remedy to lawsuits under the False Claims Act. Unlike the False Claims Act, this Act expressly defined the persons subjected to administrative liability yet omitted states from the definition. See id. at § 3801(a)(6). Appellees suggest that the exclusion of states from § 3801(a)(6) compels an inference that § 3729(a) includes states. We do not agree because the two provisions are not part of the same legislative enactment (not even the same century). See Halverson v. Slater, 129 F.3d 180, 186 (D.C. Cir. 1997) (citing Russello v. United States, 464 U.S. 16, 23 (1983)). We share appellant's view, moreover, that, since both acts proscribe essentially the same conduct, compare 31 U.S.C. § 3802(a)(1)-(2) with 31 U.S.C. § 3729(a), it would have been quite bizarre for Congress to exempt states from administrative liability if it had thought that states already were subject to the more onerous False Claims Act liability of treble damages and penalties. In sum, we are inclined to view the omission of states from the definition of person in the administrative act, to the extent it is relevant at all, as more supportive of New York's argument.
The CID section permits the government to conduct discovery of persons who "may be in possession, custody, or control of any documentary material or information relevant to a false claims investigation." 31 U.S.C. § 3733(a)(1).
Indeed, appellant and its amici, turning the blade, point out that the 1986 amendments, which increased liability from double to treble damages and increased the civil penalty, see 31 U.S.C. § 3729(a), created a form of punitive damages that would be palpably inconsistent with state liability. Congress is not thought to impose punitive damages on public entities lightly. Imposition of such a penalty has been held to be inconsistent with public policy since it gives the plaintiff a windfall at the expense of the blameless or unknowing taxpayers who must foot the bill for the government's transgressions. See City of Newport v. Fact Concerts, Inc., 453 U.S. 247, 258-71 (1981). It is true that the Supreme Court has already analyzed the Act in a related context and concluded that the statute is remedial in nature, see, e.g., Bornstein, 423 U.S. at 314-15, but as appellant rightly points out, it did so when the statute provided for double damages of which the government received a one-half share, so that the statute at that time truly did no more than make the government whole, see Graber, 8 F. Supp. 2d at 349 n. 3. Even assuming that it is possible to characterize the increased liability imposed by the 1986 amendments as remedial, that would only indicate at best that in this respect the 1986 Congress legislated in such a way that would have been consistent with state liability. The 1863 Congress, by contrast, made clear as day that it intended criminal, and a fortiori punitive, sanctions: the original statute provided for criminal penalties, including imprisonment for one to five years, for non-military persons (the class of persons said to include states) convicted under the Act, as well as fines. See § 3, 12 Stat. at 698. Those provisions are surely inconsistent with the concept of state liability.
Appellees' last sortie into the background of the 1986 amendments uncovered a piece of legislative history that they regard as the "smoking gun." They point to a Senate Report issued at the time Congress amended certain provisions of the Act that includes a section entitled "History of the False Claims Act and Court Interpretations." See S. Rep. No. 345, 99th Cong., 2d Sess., at 8 (1986), reprinted in U.S.C.C.A.N. 5266, 5273. As part of what purported to be purely descriptive history, see id. ("In its present form, the False Claims Act. . . ."), the Report states:
The False Claims Act reaches all parties who may submit false claims. The term "person" is used in its broad sense to include partnerships, associations, and corporations . . . as well as States and political subdivisions thereof. Cf. Ohio v. Helvering, 292 U.S. 360, 370 (1934); Georgia v. Evans, 316 U.S. 153, 161 (1942); Monell v. Department of Social Services of the City of New York, 436 U.S. 658 (1978).
Id. (emphasis added) (footnote omitted).
According to appellees, the Report confirms that the Congress of 1863, over a hundred years before, intended to include states as defendant persons — an argument that two of our sister circuits and the district court below accepted. See Stevens, 162 F.3d at 206-07; Zissler, 154 F.3d at 874-75; Long, 999 F. Supp. at 84-85. This portion of the Report, it should be understood, is not linked with any of the substantive amendments made by the 1986 Congress. It is instead a legislative observation about what § 3729(a), enacted by an earlier Congress, means. Courts sensibly accord such "post-enactment legislative history," arguably an outright "contradiction in terms," Sullivan v. Finklestein, 496 U.S. 617, 631 (1990) (Scalia, J., concurring), only marginal, if any, value, see Wright v. West, 505 U.S. 277, 295 n. 9 (1992) ("[T]he views of a subsequent Congress form a hazardous basis for inferring the intent of an earlier one.") (quoting Consumer Product Safety Comm'n v, GTE Sylvania, Inc., 447 U.S. 102, 117 (1980) (quoting United States v. Price, 361 U.S. 304, 313 (1960))). Post-enactment legislative history — perhaps better referred to as "legislative future" — becomes of absolutely no significance when the subsequent Congress (or more precisely, a committee of one House) takes on the role of a court and in its reports asserts the meaning of a prior statute. See Pierce v. Underwood, 487 U.S. 552, 566 (1988); In re North, 50 F.3d 42, 45-46 (D.C. Cir. 1995). The Senate Report actually was more modest; it appeared only to describe the way in which the Supreme Court had interpreted the Act. Still, its author either did not read the cited cases very carefully, or perhaps more likely, made an unforgivably misleading use of the "cf." signal. None of the cases interpreted the term "person" under the False Claims Act, and all three stand for the unremarkable proposition that governmental entities can be included in the term person when Congress so intends. In short, the Report is of no legal significance. Accord United States ex rel Graber, 8 F. Supp. 2d at 354-55.
It is unclear what appellees think they add by pointing to a 1981 General Accounting Office Report that documented recent instances of state officials defrauding the United States government — of which the Senate apparently was aware when amending the statute in 1986. See S. Rep. No. 345, 99th Cong., 2d Sess., at 2 n. 1 (1986) (citing GAO Report to Congress, Fraud in Government Programs: How Extensive Is It? How Can It Be Controlled? (1981)). Not only is evidence of an intent to impose liability on state officials (which itself would be a tenuous inference from this report) distinct from an intent to impose liability on the states themselves, see Will, 491 U.S. at 68-69, but a report documenting contemporary instances of state fraud could hardly be thought to illuminate the intent of the enacting Congress in 1863.
The Report's resort to these inapposite cases is unsurprising since, at the time of the 1986 amendments, only one decision involved a qui tam suit against the state, and that decision held that states were not persons under the Act. See United States ex rel. Weinberger v. Florida, 615 F.2d 1370, 1371 (5th Cir. 1980) (describing district court's decision to that effect and vacating on the ground that, under an older and since modified version of the present 31 U.S.C. § 3730, the district court lacked subject matter jurisdiction because the federal government had knowledge of the facts underlying the relator's suit).
The Eighth Circuit thought that 1986 amendments to § 3729(a) warranted giving the 1986 Report greater interpretive weight, even on the assumption that the Report's understanding of the pre-1986 caselaw was incorrect. See Zissler, 154 F.3d at 874. Again, the change to § 3729(a) had nothing to do with the meaning of the term person. The portion of the Report in question, moreover, makes no reference whatsoever to the slight alteration actually made to § 3729(a). It is merely a commentary on the past.
Nevertheless, appellees contend that we have asked the wrong question in searching the legislative materials for affirmative indications that Congress intended to include states as defendant persons in § 3729(a). Instead, they would have us start with the presumption that states are defendant persons and look only for some indication that Congress intended to exclude states. They justify this approach by arguing that states can be plaintiffs under § 3730(b)(1) (providing that "[a] person may bring a civil action for a violation of section 3729 for the person and for the United States Government"), and that the same statutory term, person, is used to describe the eligible class of plaintiffs. The word person is presumed to have the same meaning in different sections of the same statute. See, e.g., Commissioner v. Lundy, 516 U.S. 235, 250 (1996). Appellees, then, would use the canon of consistent meaning (following the Second and Eighth Circuits) to trump the Will-Wilson default rule. See Stevens, 162 F.3d at 205; Zissler, 154 F.3d at 875; see also Boese, supra, at 2-92 (reasoning that states are defendant persons under the Act because they are proper qui tam plaintiffs).
Although New York seemed insistent that it can have it both ways — that it can be a plaintiff but not a defendant — the states, appearing as amici, seemed quite prepared to abandon any claim that they could sue as plaintiffs' the threat of being a qui tam defendant apparently "concentrated their minds."
The consistent meaning canon is brandished as if the question whether states could be qui tam relators were a statutory given. But it is not. We recognize that other courts have assumed that states can be qui tam relators, see, e.g., United States ex rel. Woodard v. Country View Care Ctr., Inc, 797 F.2d 888 (10th Cir. 1986); United States ex rel. Wisconsin v. Dean, 729 F.2d 1100 (7th Cir. 1984), even though the term person under § 3730(b)(1) is no more clearly defined than it is under § 3729(a). The argument that states are plaintiffs is based on a provision passed in 1986 conferring jurisdiction on the district courts "over any action brought under the laws of any State for the recovery of funds paid by a State or local government if the action arises from the same transaction or occurrences as [a qui tam suit] brought under Section 3730." 31 U.S.C. § 3732(b). If states are the only parties who could bring a state law suit to recover state funds, the argument goes, and if a state is forbidden by § 3730(b)(5) from intervening in another party's qui tam suit, see id. at § 3730(b)(5) (providing that "[w]hen a person brings an action under this subsection no person other than the Government may intervene or bring a related action based on the facts underlying the pending action"), it seems to follow that the Congress which enacted § 3732(b) intended states to be qui tam relators under the Act. Otherwise, it is argued, the provision conferring jurisdiction over the state's claim under state law has little meaning. The legislative history lends some support to this reasoning. See S. Rep. No. 345, 99th Cong., 2d Sess., at 16 (1986), reprinted in 1986 U.S.C.C.A.N. 5266, 5281 (explaining that the provision was enacted in response to comments from the National Association of Attorneys General and was intended to allow "State and local governments to join State law actions with False Claims Act actions brought in Federal district court if such actions grow out of the same transaction or occurrence"); see also id. at 12-13, reprinted in 1986 U.S.C.C.A.N. at 5277-78 (disapproving of Dean decision on unrelated jurisdictional grounds but not questioning the State of Wisconsin's ability to be a qui tam plaintiff); Stevens, 162 F.3d at 204-05 (discussing Senate Report).
The more obvious reading of § 3732(b), however, is that it authorizes permissive intervention by states for recovery of state funds (creating what is in effect an exception to § 3730(b)(5)'s apparent general bar on intervention by all other parties except for the United States). See Boese, supra, at 4-13 (explaining that § 3732(b) "does not require the state to be a relator for jurisdiction to exist," noting the possibility that it permits intervention by states, but making no reference to § 3730(b)(5)). Or Congress might even have meant § 3732(b) to provide supplemental jurisdiction for a non-state relator to join a federal false claim action with an action to recover state funds under a state qui tam statute, which several states have enacted. See, e.g., Cal. Gov't Code § 12650 et seq. (West 1998); Fla. Stat Ann. § 68.081-092 (West 1998).
In any event, the argument that states are relators under § 3730(b)(1) is rather strained. To the extent it relies on the Senate Report author's knowledge of one suit by a state relator, it is no more persuasive than the analogous argument based on the Report's "recognition" of prior suits against state defendants. The argument, moreover, depends on the proposition that § 3730(b)(5) prevents all parties, except for the United States, from intervening in another relator's qui tam action. Yet it is not at all clear that this provision precludes all forms of party joinder, which would effectively limit qui tam actions to single relators. See United States ex rel. Precision Co. v. Koch Indus., Inc., 31 F.3d 1015, 1017 (10th Cir. 1994) (holding that § 3730(b)(5) does not prohibit all forms of joinder but only prevents permissive intervention in a relator's suit by unrelated parties under Fed.R.Civ.P. 24(b)(2)). If states could join as co-plaintiffs with private relators or the federal government, then § 3732(b) could be given full meaning without reading § 3730(b)(1) to include states as relators.
It should be apparent, then, that whether states can be qui tam relators presents an extraordinarily difficult question of statutory interpretation in its own right. Although appellees do not acknowledge it, their argument would require us to puzzle through that question — not squarely presented to us — in order to resolve the actual question before us (itself no easy one) in their favor. The consistent meaning canon does not have much usefulness if in order to apply it a court has to struggle that hard to determine the second meaning, against which the first is to be compared. Given the uncertainty governing the question whether states can be relators, we think the proper course is to decide only the issue before us.
Even assuming arguendo that states can be relators, we doubt that the consistent meaning canon is appropriately applied in this case. The canon itself has an important exception "[w]here the subject-matter to which the words refer is not the same in the several places where they are used." Atlantic Cleaners Dyers, Inc. v. United States, 286 U.S. 427, 433 (1932). Imposing liability is quite different from conferring a right to sue, and as we noted above, the Will-Wilson default rule has added force when the question is whether states are subject to liability as persons. See Will, 491 U.S. at 64. The canon also encounters potentially insurmountable difficulties when the "meanings" are enacted by two different Congresses — which is an obvious flaw in appellees' effort to use the 1986 amendments' effect on the term person in § 3730(b)(1) to give consistent meaning to the term person in § 3729(a), which was enacted by the 1863 Congress.
It might be argued that the 1986 amendments merely clarified that Congress has intended states to be relators since 1863, and that the consistent meaning canon really applies to the 1863 Congress alone. But this theory would require us, quite illogically, to interpret the 1986 legislative action as a declaration of what a Congress over a century earlier intended. The action of the 1986 Congress tells us, at most, what the 1986 Congress thought about states as qui tam relators (and as we noted above, it does not tell us very much); it does not purport to tell us, nor could it, what the 1863 Congress intended. See Rainwater v. United States, 356 U.S. 590, 593 (1958) (stating that 1918 amendment to the criminal provisions of the False Claims Act was at most "merely an expression of how the 1918 Congress interpreted a statute passed by another Congress more than a half century before" and had "very little, if any, significance" in interpreting the original Act's civil provisions). Although the Supreme Court occasionally says that "[s]ubsequent legislation which declares the intent of an earlier law is entitled to great weight in statutory construction," Loving v. United States, 517 U.S. 748, 770 (1996) (quoting Consumer Product Safety Comm'n v. GTE Sylvania, Inc., 447 U.S. 102, 118 n. 3 (1980) (quoting Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 380-81 (1969))), the Supreme Court's application of that principle has been rather inconsistent, see Paramount Health Sys., Inc. v. Wright, 138 F.3d 706, 709-11 (7th Cir. 1998) (comparing this rule with the competing rule that the views of a subsequent Congress in legislative history form a hazardous basis for inferring the intent of an earlier one). And we are unaware of any Supreme Court holding in which a subsequent declaration has been used, not to discern the current meaning of a statute post-declaration, see, e.g., Seatrain Shipbuilding Corp. v. Shell Oil Co., 444 U.S. 572, 595-96 (1980); Red Lion Broadcasting, 395 U.S. at 380-81, but instead to interpret the meaning of a statute prior to the declaration. Appellees' attempt to apply the consistent meaning canon to the 1863 Congress depends on precisely such a "retroactive clarification."
III.
Appellees have not persuasively demonstrated a congressional intent to include states as defendant persons under the False Claims Act. That being so, the default rule would seem to dictate that they are not. We hesitate in resting solely on this ground, however, since the Supreme Court has never explained just how much of a showing suffices to overcome the presumption against interpreting persons to include states, and indeed on occasion has employed the rule in a somewhat diluted fashion. See, e.g., Sims v. United States, 359 U.S. 108, 111-12 (1959); United States v. California, 297 U.S. 175, 186 (1936); Ohio v. Helvering, 292 U.S. at 370-71. We think there are additional considerations, however, that resolve all doubts in New York's favor.
A.
Were we to agree with appellees that states can be defendants under the False Claims Act, we would be obliged to decide whether, as appellant New York contends, the Eleventh Amendment bars a qui tam suit by a private relator against a state in federal court. The Amendment states that "[t]he Judicial Power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State." U.S. Const. amend. XI. Although it has been read to bar suits by plaintiffs not identified in the text of the amendment itself, such as citizens of the state being sued, see Hans v. Louisiana, 134 U.S. 1, 10-11 (1890), and foreign sovereigns, see Principality of Monaco v. Mississippi, 290 U.S. 313, 330-32 (1934), it is well settled that it poses no bar to a suit by the United States against a state in federal court. The states' consent to such suits is thought to be inherent in the constitutional plan and necessary to the very permanence of the Union. See, e.g., West Virginia v. United States, 479 U.S. 305, 311 (1987); Monaco, 292 U.S. at 329; United States v. Texas, 143 U.S. 621, 641-46 (1892). Reasoning from this unobjectionable proposition, three of our sister circuits have held that, since a qui tam suit against a state is essentially a suit by and for the United States, the Eleventh Amendment does not preclude a qui tam suit in federal court. See Stevens, 162 F.3d at 201-03; United States ex rel. Rodgers v. Arkansas, 154 F.3d 865, 868 (8th Cir. 1998); United States ex rel. Milam v. University of Texas M.D. Anderson Cancer Ctr., 961 F.2d 46, 50 (4th Cir. 1992); see also United States ex rel. Fine v. Chevron, U.S.A., Inc., 39 F.3d 957, 962-63 (9th Cir. 1994), vacated on other grounds, 72 F.3d 740 (9th Cir. 1995) (en banc).
We think our sister circuits have paid insufficient attention to the Supreme Court's decision in Blatchford v. Native Village of Noatak, 501 U.S. 775 (1991). In Blatchford, the Court held that a statute giving federal district courts original jurisdiction of suits brought by an Indian tribe involving federal law did not constitute a delegation to the tribes of the United States' ability, free from the Eleventh Amendment bar, to sue the states as the tribes' trustee. See id. at 785-86. Although the Court held that Congress intended no delegation in the jurisdictional statute, the Court was dubious that such a delegation would have been constitutionally permissible:
We doubt . . . that that sovereign exemption can be delegated — even if one limits the permissibility of delegation . . . to persons on whose behalf the United States itself might sue. The consent, "inherent in the convention," to suit by the United States — at the instance and under the control of responsible federal officers — is not consent to suit by anyone whom the United States might select; and even consent to suit by the United States for a particular persons's benefit is not consent to suit by that person himself.
Id. at 785 (emphasis added).
It seems to us that permitting a qui tam relator to sue a state in federal court based on the government's exemption from the Eleventh Amendment bar involves just the kind of delegation that Blatchford so plainly questioned. See Rodgers, 154 F.3d at 869 (Panner, J., dissenting). Nor are we persuaded by the argument that the Court in Blatchford was concerned about a possible delegation of the United States `Eleventh Amendment exemption just because the injury to be remedied was the tribe's and not the United States'. See Stevens, 162 F.3d at 203. The problems inherent in expanding the states' consent to suit by the United States to suits "by anyone whom the United States might select," Blatchford, 501 U.S. at 785, are no less troublesome where, as here, the injury on which the suit is premised is a pecuniary injury to the United States. One should bear in mind that the United States' ability to sue is broad; it is not limited to suits to protect the federal fisc. See, e.g., In re Debs, 158 U.S. 564, 584 (1895), disapproved of on other grounds Bloom v. Illinois, 391 U.S. 194, 208 (1968). Indeed, the United States' very ability to sue as the tribes' trustee, which was unquestioned in Blatchford, depended on an injury to the United States as sovereign when injury was inflicted on the tribes. See United States v. Minnesota, 270 U.S. 181, 194 (1926). It does not seem reasonable, therefore, to distinguish Blatchford as an anti-delegation principle applicable only where the "injury" is an injury to someone other than the United States. The problem in either case is whether, consistent with the constitutional plan, the United States can delegate its own exemption from the Eleventh Amendment bar to another party.
Whatever the ultimate resolution of the question, we think it presents a serious constitutional issue. It is quite a stretch to claim that such a delegation was part of the inherent constitutional design, or that the permanence of the union somehow depends on giving the United States broad latitude to permit private parties to sue the states in the federal courts on the United States' behalf. Compare United States v. Texas, 143 U.S. at 644-45. To assume that the United States possesses plenary power to do what it will with its Eleventh Amendment exemption is to acknowledge that Congress can make an end-run around the limits that that Amendment imposes on its legislative choices. Imagine that Congress is contemplating a new statute, to be enacted pursuant to its Article I powers, which would create a private cause of action against the states in federal court. Since the Court's decision in Seminole Tribe of Florida v. Florida, 517 U.S. 44 (1996), Congress would not be able to enact such a statute, irrespective of its clarity in imposing liability against the states, because Congress is without constitutional power to abrogate the states' Eleventh Amendment immunity under its Article I powers. See id. at 57-73. Yet if Congress is permitted to use the qui tam device to create a private cause of action against the states brought on behalf and in the name of the United States, it can reach precisely the same end without constitutional impediment. See Jonathan R. Siegel, The Hidden Source of Congress's Power to Abrogate State Sovereign Immunity, 73 Tex. L. Rev. 539, 556-64 (1995) (approving of this outcome); see also Blatchford, 501 U.S. at 785-86 (noting that the tribe's "delegation theory" was designed to avoid the constraints on congressional abrogation of the states' Eleventh Amendment immunity). Admittedly, Congress could have imposed liability against the states if it chose to put enforcement of the statute "at the instance and under the control of responsible federal officers." Blatchford, 501 U.S. at 785; see also Seminole Tribe, 517 U.S. at 71 n. 14. But the quite different legislative choice of authorizing private parties to haul sovereign states into federal court against their will, ordinarily foreclosed unless Congress successfully abrogates the states' immunity, suddenly becomes an all too easy legislative option.
Long and the government would avoid the Blatchford delegation difficulty by asserting that in qui tam suits the United States is the real party in interest; a qui tam suit is therefore essentially a suit by and for the United States. See, e.g., Stevens, 162 F.3d at 202; Milam, 961 F.2d at 49 (concluding that the United States is the real party in interest because of "the structure of the qui tam procedure, the extensive benefit flowing to the government from any recovery, and the extensive power the government has to control the litigation"). This argument appears to us merely to sidestep the core problem because it ignores the relator's undisputed role as a party with a cause of action under the Act. The "real party in interest" rule ordinarily requires that the suit be brought by the "person who, according to the governing substantive law, is entitled to enforce the right." 6A Charles Alan Wright et al., Federal Practice Procedure § 1543, at 334 (2d ed. 1990); see Fed R. Civ. P. 17(a) (stating that "every action shall be prosecuted in the name of the real party in interest"). There is no question that the False Claims Act gives such a right to the relator, see 31 U.S.C. § 3730(b) ("A person may bring a civil action for a violation of section 3729 for the person and for the United States Government.") (emphasis added), and the statutory right to bring suit is sufficient to satisfy the real party in interest requirement, even if the suit is brought for the benefit of some other party, see Fed.R.Civ.P. 17(a) (second sentence); Wright et al., § 1550, at 384. In any event, contrary to the suggestion of the district court, see Long, 999 F. Supp. at 83-84, a qui tam action is brought for the benefit of both the relator and the United States, not for the benefit of the United States alone. See 31 U.S.C. § 3730(b) (authorizing qui tam suit "for the person and for the United States Government"). Nor does it make any difference that the False Claims Act requires the relator to sue "in the name of the Government," 31 U.S.C. § 3730(b), because the procedural question of in whose name the suit must be brought is distinct from the substantive legal question whether the plaintiff has a cause of action. See Wright Et Al. § 1544, at 340.
The district court concluded that Long's claim under the whistle-blower provision of the False Claims Act, 31 U.S.C. § 3730(h), was barred by the Eleventh Amendment because, unlike a qui tam suit under § 3730(b) brought in the name of the United States, a claim under § 3730(h) is a true "private right of action." Long, 999 F. Supp. at 92. We disagree; a qui tam suit under § 3730(b) is no less a cause of action, and the relator is no less a party prosecuting that action, because the action is brought in the name of the United States.
Accordingly, we do not think the relator's technical status as a "real party in interest" is inconsistent with the conclusion of our sister circuits that the United States is a "real party in interest" as well. See, e.g., Stevens, 162 F.3d at 202; Rodgers, 154 F.3d at 868; United States ex rel. Hyatt v. Northrop Corp., 91 F.3d 1211, 1217 n. 8 (9th Cir. 1996); Milam, 961 F.2d at 49. It is, after all, not unheard of for there to be two real parties in interest to a cause of action. See Wright et al., § 1545, at 351-53 (in cases of partial assignments, the assignor and assignee are both real parties in interest); id. § 1546, at 360 (same for partial subrogation). More important, although we are aware of a variant of the doctrine used in a related Eleventh Amendment context, see, e.g., Ford Motor Co. v. Department of Treasury, 323 U.S. 459, 464 (1945) (analyzing whether a state defendant is the "real party in interest" such that a suit against a state entity, though not nominally against the state, would be barred by the Eleventh Amendment), we do not see how the doctrine can be used to convert a party with a statutory cause of action into a "nonparty-party." In short, we think the real party in interest doctrine is plainly irrelevant to the Eleventh Amendment question presented in this case. See Rodgers, 154 F.3d at 869 (Panner, J., dissenting).
One of the principal concerns motivating the Eleventh Amendment inquiry into whether the state is the "real party in interest" defendant (or in other words that the actual defendant is an "arm of the state") is that an individual plaintiff's recovery will be paid out of the state treasury. See Regents of the University of California v. Doe, 117 S.Ct. 900, 904 (1997). That is the precise concern presented by a private relator recovering against a state defendant in a qui tam suit.
Nor do we think, as appellees suggest, that the government's control over a relator's suit alters the result. We acknowledge that the government takes the greater share of any recovery, see 31 U.S.C. § 3730(d)(1),(2), and that the statute gives the United States considerable control over the relator's suit, see, e.g., id. at § 3730(b)(2) (providing that the government can intervene in the suit as of right within sixty days after receiving the relator's complaint, evidence, and information); id. at § 3730(b)(1) (relator cannot dismiss his own suit without written consent of the court and the Attorney General); id. at § 3730(c)(3)-(4) (even if the government does not intervene, it may monitor the proceedings and stay discovery in certain situations); id. at 3730(c)(3) (government can intervene at any time upon a showing of good cause); id. § at 3730(c)(2)(A) (government may dismiss the suit after notice to the relator and a hearing); id. at § 3730(c)(2)(B) (government may settle the suit with the defendant over the relator's objection if the court approves after a hearing). Still, we simply do not see how the government's potential exercise of its power renders the relator any less a party. Whatever the degree of control the United States exercises, we think it is telling that, although there are some intimations to that effect, no court has actually held that the relator is not a party to the qui tam suit merely because of the United States' potential ability to control the prosecution of the suit.
There are, however, substantial restrictions on the United States' power incorporated within these provisions. See Stevens, 162 F.3d at 223-24 (Weinstein, J., dissenting).
The relator appears to remain a party whether or not the United States intervenes. In either situation, the relator's rights must be protected under the statute. See 31 U.S.C. § 3730(c)(3) (providing that the court may permit the United States to intervene for good cause but must not "limit[ ] the status and rights of the person initiating the action"); id. at § 3730(c)(1) (providing that the relator "shall have the right to continue as a party to the action," subject to certain limitations, even after the United States intervenes). This is important because the Eleventh Amendment must be satisfied for every claim in the suit, see Pennhurst State Sch. Hosp. v. Halderman, 465 U.S. 89, 121 (1984), and the presence of the United States as a co-plaintiff does not ordinarily remove the Eleventh Amendment bar for claims by other plaintiffs, see id. at 103 n. 12; but see Rodgers, 154 F.3d at 870 (Panner, J., dissenting) (distinguishing for Eleventh Amendment purposes between cases in which the United States intervenes from those in which it does not). But assuming arguendo that the Eleventh Amendment would not pose a problem in cases in which the United States actually intervenes in a suit against a state, the government did not do so in the present case. That fact, coupled with the government's intervention limited to the claim against the private defendants, suggests that the government does not lightly take on the task of probing into the internal operations of the sovereign states, and may well think it better to leave such politically unpalatable tasks for the qui tam relators of the world. Yet, the government wishes the option to sit back while the relator brings an action against a state, thus removing itself from direct accountability and from the subtle political pressures that might have precluded the lawsuit in the first place had the United States been more actively involved from the start. See Stevens, 162 F.3d at 225-29 (Weinstein, J., dissenting). That seems quite at odds with the obvious purpose of the Eleventh Amendment since such a suit is emphatically not one brought "at the instance and under the control of responsible federal officers." Blatchford, 501 U.S. at 785. We seriously doubt that the government, under the Eleventh Amendment, is entitled to transfer all of the benefits that accrue to it as a plaintiff in the federal courts when it chooses to watch from the sidelines. That could be described as allowing the government to have its constitutional cake and eat it too.
It has also been contended that, despite the clear statutory language giving relators a cause of action and treating them as parties vested with rights and protections, relators should be seen instead as self-appointed government counsel. See Stevens, 162 F.3d at 202; Milam, 961 F.2d at 49 ("Congress has let loose a posse of ad hoc deputies to uncover and prosecute frauds against the government."); Siegel, supra, 73 Tex. L. Rev. at 556-57; Evan Caminker, The Constitutionality of Qui Tam Actions, 99 Yale L. J. 341, 353 (1989). It has even been suggested that the relator's economic interest in the lawsuit makes him more like a contingency fee lawyer than a party. See Stevens, 162 F.3d at 202 (acknowledging that the qui tam plaintiff has an interest in the action's outcome, but stating that "his interest is less like that of a party than that of an attorney working for a contingent fee" and citing cases noting that relators' primary motivation is a monetary reward and not the public good). We simply do not understand the analogy typically both the client and the attorney have an economic interest in litigation. In this sense, a relator looks no different to us than, let us say, an applicant for a broadcast license. It is therefore not possible to contend that the False Claims Act is an open-ended letter of engagement from the government as client to a posse of prospective attorneys. See United States ex rel. Farrell v. SKF, USA, Inc., 32 F. Supp. 2d 617, 617-18 (W.D.N.Y. 1999) (rejecting contention by qui tam defendant that, since the relator is only the United States' lawyer and the United States always remains a party litigant, the defendant was entitled to discovery from the United States even though the United States had not intervened in the suit). To accept the "private Attorneys General" characterization as anything more than an inapt convention would run headlong into the problems of how a party with a statutory right to sue on his own behalf can be thought to be acting in a representational capacity, see 31 U.S.C. § 3730(b), why the client would need the court's permission to intervene in his own suit, see id. at § 3730(c)(3), or to dismiss the lawyer's "suit," see id. at § 3730(c)(2)(A), and why the lawyer's "status and rights" would be worthy of statutory protection in the event the client chooses to intervene in the lawyer's action, see id. at § 3730(c)(3).
Of course, if the government actually hired a lawyer to bring its own cause of action, the Blatchford delegation problem would not arise. But as we have explained at length, that is not what the False Claims Act does.
B.
Although, as we have indicated, we have profound doubts that the Eleventh Amendment permits this lawsuit against New York even if Congress implicitly authorized relators to bring suits against the states, we do not rest our decision on an interpretation of the Constitution. Instead, bearing in mind that we must decide this difficult constitutional issue only if the term person in the Act is interpreted as including states, and that it seems quite dubious that Congress intended that result, the appropriate course seems to us to interpret "person" as not including states.
The venerable doctrine of construing statutes in such a way as to avoid serious constitutional questions has two important prerequisites. First, the "statute must be genuinely susceptible to two constructions," and this determination must be made "after, and not before, [the statute's] complexities are unraveled." Almendarez-Torres, 118 S.Ct. at 1228; see also United States v. Espy, 145 F.3d 1369, 1372 (D.C. Cir. 1998); Association of Am. Physicians Surgeons, Inc. v. Clinton, 997 F.2d 898, 906, 910-11 (D.C. Cir. 1993). Furthermore, the constitutional question must be one that presents a "serious likelihood that the statute will be held unconstitutional." Almendarez-Torres, 118 S.Ct. at 1228; see also Association of Am. Physicians Surgeons, 997 F.2d at 906 (constitutional question must be a "grave" one); Espy, 145 F.3d at 1372.
It is obvious from what we have said already that these requirements are satisfied in this case. As we have just explained at length, the Eleventh Amendment question is, at bare minimum, a serious one. It could not be suggested, moreover, that we are distorting the language of the statute in order to avoid a constitutional question. The more obvious reading is to exclude states from "person." The more difficult task is to demonstrate that the inclusion of states as defendant persons is a fair reading of the statute. There can be no objection to avoiding a constitutional question that is implicated only by a rather strained reading of the statute.
We think it relevant — if not decisive — to observe that the avoidance canon coincides in this case with two additional related canons of construction that impose upon Congress an obligation of specificity. When "Congress intends to alter the usual constitutional balance between the States and the Federal Government,'" federal courts insist that Congress "make its intention to do so `unmistakably clear in the language of the statute.'" Will, 491 U.S. at 65 (quoting Atascadero State Hosp. v. Scanlon, 473 U.S. 234, 242 (1985)); see also Gregory v. Ashcroft, 501 U.S. 452, 464 (1991) (linking clear statement rule with constitutional avoidance canon). The Court in Will derived this "clear statement" rule from the Eleventh Amendment cases requiring an explicit textual intent to abrogate a state's Eleventh Amendment immunity, but noted its applicability in a range of contexts in which Congress alters the federal-state balance of power. See Will, 491 U.S. at 65. In Gregory, the Court applied this "plain statement" principle where Congress' imposition of liability under the Age Discrimination in Employment Act would "upset the usual constitutional balance" by interfering with the states' fundamental role in defining the qualifications of their state judges. Id. at 460-61; see id. at 464-67 (holding that Congress did not make a sufficiently clear statement in the ADEA that state judges are within the Act's coverage). It cannot seriously be disputed that if Congress were required to make its intentions "clear and manifest," Will, 491 U.S. at 65, in order to impose False Claims Act liability on the states, it has failed to do so.
Indeed, in Will itself the Eleventh Amendment was not a concern because the question whether states were persons under § 1983 arose in the context of a state court case, and the Eleventh Amendment does not apply in state courts. See Will, 491 U.S. at 63-64.
Appellees contend that there is no justification for applying this clear statement rule of Will or Gregory because treating states as defendant persons would not actually alter the constitutional balance of powers between the federal and state governments. Such an alteration occurs, for example, when Congress seeks to remove the states' sovereign immunity in their own courts, as in Will, 491 U.S. at 67, or when Congress attempts to interfere with an essential governmental function, as in Gregory, 501 U.S. at 460. Since this case arose in federal court and because the fraudulent conduct proscribed cannot be thought an essential governmental function, appellees argue that neither Will nor Gregory apply.
We are unpersuaded by various crabbed analyses of the Court's "clear statement" jurisprudence that we have seen. To characterize the relevant state function at issue, as the Second Circuit did, as fraudulent conduct, see, e.g., Stevens, 162 F.3d at 204 ("The States have no right or authority, traditional or otherwise, to engage in [fraudulent] conduct."), is to assume the conclusion that the function is not an essential one. Using that logic, the Court in Gregory would have declined to apply a clear statement rule because it is not essential for the state to discriminate against elderly judges. Appellees, for their part, describe the governmental function at issue in this case as the process by which a state receives federal funding — which they argue cannot possibly be described as an essential state function. The state, in other words, is simply a supplicant coming to the federal sovereign. That characterization, in our view, is still too narrow because the Act's imposition of liability necessarily interferes with a state's sovereign performance of a range of indisputably essential functions, such as the administration of a state education department involved in the present case. See Ambach v. Norwick, 441 U.S. 68, 76 (1979) ("Public education, like the police function, `fulfills a most fundamental obligation of government to its constituency.'") (quoting Foley v. Connelie, 435 U.S. 291, 297 (1978)); Brown v. Board of Educ., 347 U.S. 483, 493 (1954) ("[E]ducation is perhaps the most important function of state and local governments."). That the federal government funds in part that function does not destroy its essentiality to the state. To accept that hypothesis, given present tax and spending mechanisms, would go a long way toward burying federalism.
The Supreme Court has applied Gregory as we do, focusing on the state functions necessarily affected by operation of the statute, and not exclusively on the actual conduct proscribed by Congress. See Gregory, 501 U.S. at 463 (essential state function with which ADEA liability would interfere was the "authority of the people of the States to determine the qualifications of their most important government officials" in their state Constitutions); see also Pennsylvania Dep't of Corrections v. Yeskey, 118 S.Ct. 1952, 1953-54 (1998) (assuming that imposition of ADA liability against state prisons would interfere with the essential state function of "exercising ultimate control over the management of state prisons"); BFP v. Resolution Trust Corp., 511 U.S. 531, 544 n. 8 (1994) (applying Gregory to Bankruptcy Code provisions governing constructively fraudulent transfers, and explaining that the state function was not general authority over debtor-creditor law, but the "essential sovereign interest in the security and stability of title to land" necessarily affected by application of the Bankruptcy Code to foreclosure sales). We thus do not think it is appropriate to look myopically only to the state's formal submission of the claim to the government and to ignore the underlying governmental functions to which the claim relates.
It could be argued, we suppose, that because False Claims Act liability is only triggered when the state requests money from the federal government, it brings any interference with its essential functions on itself. But we do not see any basis in Gregory for eliminating the need for a clear statement simply because the liability imposed is conditioned on a voluntary act by the state. The clear statement rule of Pennhurst State School and Hospital v. Halderman, 451 U.S. 1, 17 (1981) — which requires a clear statement when Congress imposes conditions on grants of federal money — seems flatly inconsistent with such an argument.
Appellees similarly give an overly restrictive reading of Will. It is true that the Court in Will pointed to the states' sovereign immunity in their own courts as a supporting reason for concluding that Congress did not intend to make states persons under 42 U.S.C. § 1983. See Will, 491 U.S. at 66-67. But the Court nowhere even suggested that abrogation of state sovereign immunity was the only alteration of the constitutional balance that justified use of the clear statement rule, nor did it rely on the idea of essential state functions implicit in the later decision in Gregory. Will could be read to suggest — although we are uncertain of this — that it was the very imposition of a new liability against the state that would have altered the constitutional balance of powers.
Whether or not Will or Gregory can be taken as far as we have suggested, there is a second related clear statement canon that bears on our case. In cases involving congressional abrogation of a state's Eleventh Amendment immunity, the applicability of the clear statement rule is well-established and the uncertainties in defining the scope of the Will and Gregory versions of that rule disappear. See Dellmuth v. Muth, 491 U.S. 223, 230 (1989). Appellees contend that that rule does not apply, however, because they conclude that the Eleventh Amendment is not a bar to a qui tam suit (and thus that no abrogation is necessary). But it seems highly artificial to conclude that Congress labors under an obligation of utmost textual specificity when it seeks to abrogate the states' Eleventh Amendment immunity when that immunity is otherwise certain, but that liability against the states — potentially implicating the Eleventh Amendment — can be imposed willy-nilly, using as imprecise a term as "person." We think there is significant conceptual overlap — though admittedly not an identity — between the abrogation inquiry and the statutory construction question whether Congress intended to include states as defendant persons. The Supreme Court's conclusion that Congress has failed to abrogate with the requisite specificity is often based on Congress' failure explicitly to provide for suits against the states in federal court — the precise failing of the False Claims Act that raises the question in this appeal. See, e.g., Dellmuth, 491 U.S. at 231-32; Atascadero State Hosp., 473 U.S. at 245-46. So although we recognize that the Eleventh Amendment's clear statement rule has always been applied to an abrogation inquiry — rather than to a threshold question as to whether the Eleventh Amendment applies — we do not think it wholly irrelevant to the latter.
The government would have us instead limit the Court's clear statement rules because of the significant reliance interests created by Congress' and the federal agencies' assumption that states, to whom they entrusted large sums of money, are covered by the Act. For the proposition that reliance interests can trump clear statement rules, the government relies on Hilton v. South Carolina Public Railways Comm'n, 502 U.S. 197, 205-07 (1991) (holding that Will is a rule of statutory construction, not of constitutional law, and that the reliance interests created by the Court's prior decision interpreting the Federal Employers' Liability Act to include state-owned railroads warranted adherence to stare decisis rather than to the clear statement rule). That the Court feels obliged to disregard the clear statement rule because of reliance interests that it created through its own precedent is of course quite different from the government's contention. Any reliance interests in this case are not the judiciary's doing, but rather stem from the legislature's and the federal agencies' assumption, based on weak post-enactment legislative history, that states were, or ought to be, covered by the Act. Since Congress easily could have included states within the definition of person if it so intended, the government can hardly be heard to complain now (on behalf of Congress) that Congress, in effect, wrote the states a blank check.
Appellees' argument against using the Eleventh Amendment's clear statement rule follows from their prior conclusion that the Eleventh Amendment does not apply to this case. Appellees therefore assume that states are persons for the purpose of rejecting New York's Eleventh Amendment defense, and then proceed to reject the Eleventh Amendment's clear statement rule when actually interpreting the statute previously assumed to include states — sort of a divide and conquer strategy. The statutory construction issue is, however, inextricably linked with the jurisdictional one, which is precisely why we decline to assume that states are persons in order to conduct an Eleventh Amendment inquiry that could be avoided if the assumption were not made in the first place. We think the correct resolution is to read the Act in such a way that avoids the serious constitutional question whether the Eleventh Amendment bars qui tam suits against the state in federal court. In so doing, we rely on the constitutional avoidance canon buttressed by the family of "clear statement" rules applicable when Congress attempts to legislate in the way that appellees contend it has legislated.
New York would also have us apply the clear statement rule of Pennhurst, 451 U.S. at 17, under which Congress must unambiguously set forth conditions it imposes on the grant of federal money when it exercises its spending power. Because we have enough — more than enough — clear statement rules to resolve this case, we need not decide whether False Claims Act liability can be seen as a condition imposed on a grant of federal money.
* * * *
In the end it comes to this: if we must decide whether states constitutionally can be defendants in federal court under the Act, Congress must make its intent clear. The decision of the district court is therefore reversed.
So ordered.