United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued September 13, 1999 Decided December 21, 1999
No. 98-5472
Gilbert W. Galvan,
Appellant
v.
Federal Prison Industries, Inc.,
Appellee
Appeal from the United States District Court
for the District of Columbia
(No. 96cv01722)
Thomas G. Corcoran, Jr. argued the cause and was on the
briefs for appellant.
Sally M. Rider, Assistant U.S. Attorney, argued the cause
for appellee. With her on the brief were Wilma A. Lewis,
U.S. Attorney, and R. Craig Lawrence, Assistant U.S. Attor-
ney.
Before: Edwards, Chief Judge, Williams and Garland,
Circuit Judges.
Opinion for the Court filed by Circuit Judge Williams.
Williams, Circuit Judge: The False Claims Act encourages
private parties to help fight fraud on the United States by
giving them the power to bring civil actions in its name, and
by providing both the government and the private party--
known as the "relator"--a share of any financial recovery and
reimbursement for their costs, including attorneys' fees. 31
U.S.C. ss 3729-3730 (1994). Under the Act any person who
knowingly presents false or fraudulent claims to an officer or
employee of the United States may be liable. Id. s 3729(a).
Gilbert W. Galvan, an inmate at the Federal Correctional
Institution in Oxford, Wisconsin, filed such an action--often
called by its Latin shorthand, qui tam (an abbreviation of qui
tam pro domino rege quam pro se ipso in hac parte sequi-
tur)1--against his employer, Federal Prison Industries, Inc.
("FPI"). He alleged that it had falsely certified that the
communication cables and weapons parts that it produced for
the Department of Defense had been adequately tested and
met the requisite quality standards.
FPI is no ordinary employer; it is a "wholly owned govern-
ment corporation," created to further the Bureau of Prison's
goal of providing meaningful work for inmates confined in
federal institutions. See id. s 9101; 28 CFR s 345.10 (1999).
But the suit had been brought in the name of the govern-
ment, 31 U.S.C. s 3730(b)(1), and it is accordingly entitled to
intervene, 31 U.S.C. s 3730(b)(2), which it did here. This put
the Department of Justice in place as counsel on both sides of
__________
1 Black's Law Dictionary translates the phrase as "who as well
for the king as for himself sues in this matter." Black's Law
Dictionary 1262 (7th ed. 1999). There are other versions of the
complete Latin phrase, but none appears meaningfully different.
See, e.g., United States ex rel. Kelly v. Boeing Co., 9 F.3d 743, 746
n.3 (9th Cir. 1993) ("qui tam pro domino rege quam pro se imposo
sequitur"); Miami Copper Co. v. State, 149 P. 758, 761 (Ariz. 1915)
("qui tam pro domino rege, etc., quam pro se ipso in hac parte
sequitur").
the action. It then moved under s 3730(c)(2)(A) for dismissal
of the suit, arguing that the court lacked subject matter
jurisdiction because Galvan's qui tam action pitted the United
States executive branch against itself. Further, representing
the FPI itself, the government moved to dismiss on grounds
of sovereign immunity. The district court accepted the non-
justiciability argument, and never reached the issue of sover-
eign immunity. We agree with the government's sovereign
immunity defense and affirm the dismissal on that ground,
leaving for another day the question of justiciability.
Before addressing sovereign immunity we must be sure
that we may properly do so before deciding whether the suit
presents a case or controversy. Jurisdiction must be estab-
lished before a federal court may proceed to any other
question. Steel Co. v. Citizens for a Better Environment, 523
U.S. 83, 94-95 (1998). But later cases make clear what was
implicit in Steel Co.: There is an array of non-merits ques-
tions that we may decide in any order. Thus in Ruhrgas A.G.
v. Marathon Oil Co., 119 S. Ct. 1563 (1999), the Court held
that it may be perfectly proper for a court to resolve personal
jurisdiction, which is waivable, without having first deter-
mined subject matter jurisdiction. "[T]here is no unyielding
jurisdictional hierarchy." Id. at 1567. And in In re Minister
Papandreou, 139 F.3d 247, 255 (D.C. Cir. 1998), we consid-
ered an immunity defense despite considerable doubts about
the plaintiffs' standing, saying that "a court that dismisses on
other non-merits grounds ... makes no assumption of law-
declaring power that violates the separation of powers princi-
ples." Id. at 255.
Sovereign immunity questions clearly belong among the
non-merits decisions that courts may address even where
subject matter jurisdiction is uncertain. The Supreme Court
has characterized the defense as jurisdictional, FDIC v. Mey-
er, 510 U.S. 471, 475 (1994), even while recognizing that it can
be waived, id. See also Deaf Smith County Grain Pro-
cessors, Inc. v. Glickman, 162 F.3d 1206 (D.C. Cir. 1998);
First Va. Bank v. Randolph, 110 F.3d 75, 77 (D.C. Cir. 1997).
And in Papandreou itself, we resolved the case on immunity
grounds, despite the presence of a defense that we assumed
arguendo was a matter of Article III standing. 139 F.3d at
255. Similarly, we here address sovereign immunity and do
not reach justiciability.
* * *
Galvan argues that FPI is not entitled to sovereign immu-
nity because it is not, in fact, part of the sovereign. He is
mistaken. A suit is against the sovereign when "the judg-
ment sought would expend itself on the public treasury or
domain, or interfere with the public administration." Dugan
v. Rank, 372 U.S. 609, 620 (1963) (quoting Land v. Dollar, 330
U.S. 731, 738 (1947)). FPI is a wholly owned Government
corporation, see 31 U.S.C. s 9101, and all money under FPI's
control is held by the U.S. Treasury to the credit of FPI.
See 18 U.S.C. s 4126(a) (1994). Thus, any judgment in
Galvan's favor would require FPI to pay damages directly
from the public treasury. See generally Sprouse v. FPI, 480
F.2d 1, 3 (5th Cir. 1973) ("[T]hough the prisoners vehemently
deny it, 'the conclusion is inescapable that the suit is essen-
tially one designed to reach money which the government
owns.' " (quoting Mine Safety Appliances Co. v. Forrestal,
326 U.S. 371, 375, (1945))).
Pointing to 18 U.S.C. s 4126(b), which says that "[a]ll valid
claims and obligations payable out of said fund [the FPI fund
at Treasury] shall be assumed by the corporation," Galvan
characterizes the corporation as "self-sufficient." This is
quite immaterial. "Federal agencies or instrumentalities per-
forming federal functions always fall on the 'sovereign' side of
[the] fault line" between suits against the sovereign and suits
against individuals, regardless of any independence of ac-
counts. Auction Co. of America v. FDIC, 132 F.3d 746, 752
(D.C. Cir. 1997). "Diversion of resources from a private
entity created to advance federal interests has effects similar
to those of diversion of resources directly from the Treasury."
Id. In fact, as a government corporation FPI is not only a
federal instrumentality but is also an "executive agency," 5
U.S.C. s 105, and on that account deserves sovereign immu-
nity in the absence of congressional waiver. See FDIC v.
Meyer, 510 U.S. 471, 475 (1994) ("Absent a waiver, sovereign
immunity shields the Federal Government and its agencies
from suit.").
Galvan argues that Congress waived FPI's immunity both
in FPI's organic statute, 18 U.S.C. s 4121, and in the False
Claims Act, 31 U.S.C. s 3729. We first note the rather steep
incline that the Supreme Court has said a court must climb
before finding a waiver of the federal government's sovereign
immunity. Such waivers must be "unequivocally expressed in
statutory text, and will not be implied." Lane v. PeNa, 518
U.S. 187, 192 (1996) (internal citations omitted). If ambigu-
ous, statutes must be construed in favor of immunity. See
United States v. Williams, 514 U.S. 527, 531 (1995). So long
as a statute supposedly waiving immunity has a "plausible"
non-waiver reading, a finding of waiver must be rejected.
United States v. Nordic Village, Inc., 503 U.S. 30, 37 (1992)
("plausible" alternative reading is enough to establish that a
"reading imposing monetary liability on the Government is
not 'unambiguous' and therefore should not be adopted.").
With this in mind we turn to Galvan's specific claims.
FPI's Organic Statute. Congress established FPI as "a
government corporation of the District of Columbia." 18
U.S.C. s 4121. Galvan would have us read this as manifest-
ing a congressional intent to give FPI the legal characteris-
tics of an ordinary corporation established under the general
corporation law of the District of Columbia. That law states
that such corporations are "capable of suing and being sued
in any court of law or equity in the District," D.C. Code Ann.
s 29-203 (1999),2 language which if applicable would consti-
tute a waiver. See Meyer, 510 U.S. at 480; FHA v. Burr, 309
U.S. 242, 245 (1940).
On the surface (later we look below the surface) s 4121
seems capable of the meaning Galvan proposes. But there
are alternative meanings that seem plausible--namely read-
ings of s 4121 as intended to establish a different kind of link
__________
2 This provision was codified at Code D.C. s 607 when FPI was
first established.
with the District of Columbia. Thus Congress may have
intended to specify that the headquarters of FPI should be in
the District (as it in fact is, see Federal Prison Industries
1996 Annual Report 82). Congress has so provided, more
specifically to be sure, in other statutes. See, e.g., 22 U.S.C.
s 2199(a) ("The [Overseas Private Investment] Corporation
shall have its principal office in the District of Columbia.");
12 U.S.C. s 4703(a)(1) ("The offices of the [Community Devel-
opment Financial Institutions] Fund shall be in Washington,
D.C."). Or Congress may have intended to locate FPI in the
District specifically for purposes of venue, as it has for other
government corporations. See 22 U.S.C. s 2199(a) (Overseas
Private Investment Corp. deemed to be a resident of the
District of Columbia "for purposes of venue in civil actions");
22 U.S.C. s 3611(b) (Panama Canal Commission "is an inhab-
itant and resident of the District of Columbia"); cf. 7 U.S.C.
s 1506(d) ("Any suit against the [Federal Crop Insurance]
Corporation shall be brought in the District of Columbia, or
in the district wherein the plaintiff resides."). Because 28
U.S.C. s 1391(e) provides venue for suits against "an agency
of the United States" in any judicial district where the
defendant "resides," a congressional purpose simply to estab-
lish the central administration in Washington would have the
consequence of locating venue in the District for any case
against FPI brought under general waivers of sovereign
immunity such as the Tucker Act, see id. ss 1346, 1491.
In deciding on the plausibility of the above interpretations,
it is worth noting how precisely Congress has spoken in
instances where it sought to incorporate attributes estab-
lished by D.C.'s general corporation law. See 29 U.S.C.
s 1302(b) ("The [Pension Benefit Guaranty] corporation has
the powers conferred on a nonprofit corporation under the
District of Columbia Nonprofit Corporation Act."); cf. 40
U.S.C. s 875 (allowing Pennsylvania Avenue Development
Corporation to condemn property under the procedural provi-
sions of a specific subchapter of the D.C. Code). The same is
true of Section 11 of the Shipping Act of Sept. 7, 1916, 39
Stat. 728, 731, which allowed the United States Shipping
Board to "form under the laws of the District of Columbia
one or more corporations." In Sloan Shipyards Corp. v.
United States Shipping Board Emergency Fleet Corp., 258
U.S. 549 (1922), the Court found that this provision effected a
waiver. Id. at 565-68. Because of s 11's much more specific
statutory language, however, Sloan affords Galvan no aid.3
Looking at the context of s 4121, we find indicators militat-
ing against Galvan's view that Congress's reference to "a
government corporation of the District of Columbia" was
intended to incorporate the District's corporate law by refer-
ence. Congress purposefully kept FPI out of the commercial
world and limited its exposure to the courts. FPI cannot sell
its products to the public in competition with private enter-
prise, and even for its sales to the government must diversify
its operations to minimize competition with private industry
and to avoid capturing more than a reasonable share of the
government market for any specific product. See 18 U.S.C.
s 4122(a)-(b). Purchases of FPI products are considered
intragovernmental transfers, see id. s 4124(b), and FPI may
borrow money only from the Treasury itself, see id. s 4129.
When FPI's government customers challenge the "price,
quality, character, or suitability" of its products, their only
recourse is to binding arbitration before the Attorney Gener-
al, the Administrator of General Services, and the President,
or their representatives. See id. s 4124(a)-(b).
Other aspects of FPI activity are removed from judicial
influence. The Attorney General has authority to promulgate
rules and regulations governing inmates' compensation for
__________
3 The current text of FPI's organic statute was adopted as part
of the enactment of Title 18 of the United States Code. See Act of
June 25, 1948, Pub. L. No. 80-772, 62 Stat. 683, 683. The language
of the original act demonstrated even more clearly that Congress
did not intend to adopt the District's corporation laws. The Act of
June 23, 1934 authorized the President to "create a body corporate
of the District of Columbia to be known as 'Federal Prison Indus-
tries', which shall be a governmental body." Act of June 23, 1934,
Pub. L. No. 73-461, 48 Stat. 1211, 1211. Its language reinforces
FPI's status as a governmental entity and suggests that its status
as a corporation is generic rather than specific to the District of
Columbia.
injuries sustained and for work performed in connection with
FPI activities, see id. s 4126(c)(4). The Attorney General's
administrative scheme is the workers' exclusive remedy. See
28 CFR s 301.309 (1999); United States v. Demko, 385 U.S.
149, 152 (1966).
Reading s 4121 as literally making FPI a corporation
under the law of the District of Columbia creates still further
puzzles and contradictions. First, the management structure
of FPI conflicts directly with the laws in place when the
current statutory language was adopted.4 The D.C. Code
requires annual elections for the corporation's "trustees" (the
District's analogue of a director, see D.C. Code ss 29-202,
-204), see id. s 29-205, but FPI's board of directors is
appointed by, and serves at the will of, the President. See 18
U.S.C. s 4121. The District's laws also require that trustees
own stock in the company, see D.C. Code s 29-204, whereas
FPI's operating capital is maintained exclusively by the Trea-
sury of the United States. See 18 U.S.C. s 4126(a). In
addition, in 1948 (the time of adoption of FPI's organic
statute in approximately its current form), a majority of a
corporation's trustees were required to be citizens of the
District, see D.C. Code s 29-204 (1940), but it seems unlikely
that this rule has bound the President in his selection of
FPI's directors.
Second, Congress granted FPI considerably fewer powers
than an ordinary District corporation while simultaneously
imposing extra, particularly governmental, burdens. For ex-
ample, a District corporation has the specific power to mort-
gage its property, see s 29-203, and thus implicit power to
borrow money; but Congress authorized FPI to borrow
money only in 1988, and in so doing limited the source of
__________
4 The following citations refer to the current District of Columbia
and United States Codes, but each cited provision (with one excep-
tion) is identical in substance to the provision in effect when the
current version of FPI's organic statute was adopted: 18 U.S.C.
ss 4121-4128 (1948), and D.C. Code s 29-201 to -240 (1940). The
exception is that D.C. Code s 29-204 (1940) imposed a requirement
that "trustees" be citizens of the District, whereas the current
version does not.
funds to the U.S. Treasury, see Pub. L. No. 100-690, 102 Stat.
4411 (codified at 18 U.S.C. s 4129 (1994)). Similarly, in
employing its assets, FPI is required to meet the require-
ments of a government agency rather than a District corpora-
tion. It can withdraw money from its accounts "only pursu-
ant to accountable warrants or certificates of settlement
issued by the General Accounting Office," 18 U.S.C.
s 4126(a), and is required to act "in accordance with the laws
generally applicable to the expenditures of the several depart-
ments, agencies, and establishments of the Government," id.
s 4126(c).
Finally, the vast majority of the District's rules are either
indirectly superseded or have no relevance to FPI. For
example, the District required that shareholders pay 10% of
the capital stock into the corporate treasury before the
corporation could transact any business. See D.C. Code
s 29-209. FPI has no such capitalization requirement. The
other provisions of the D.C. Code concern the activities of
ordinary corporations. See, e.g., id. ss 29-209 to -212, -216
to -217, -230 to -235, -239 (capitalization and stock transac-
tions); id. ss 29-218 to -219 (payment of dividends); id.
ss 29-205, -211, -220 to -222 (stockholder liability and voting
rights). They have no apparent application to FPI. The
dismal fit between FPI and the sort of private corporation
clearly contemplated by the D.C. Code makes it vanishingly
improbable that Congress meant to make FPI a corporation
governed by that code.5
Galvan attempts to save his waiver argument by arguing
that the typical presumption in favor of sovereign immunity
should not apply to government corporations, citing Keifer &
Keifer v. Reconstruction Fin. Corp., 306 U.S. 381 (1939).
__________
5 Section 11 of the Shipping Act of 1916, in contrast, anticipated
the questions of market capitalization, stock management, and the
exercise of voting rights. 39 Stat. at 731. The Act also refrained
from placing governmental burdens on these corporations' spending
decisions. There may be some incongruities between the manage-
ment structure provisions of s 11 and the D.C. Code, but evidently
none was brought to the attention of the Supreme Court in Sloan.
There the Court found that in authorizing the Reconstruction
Finance Corporation ("RFC") to create up to twelve "regional
agricultural credit corporations" and to appoint their manage-
ment, Congress had not endowed the regional corporations
with sovereign immunity. Because the parent corporation,
the RFC, was subject to a "sue and be sued" clause, the
Court found that Congress "naturally assumed" that the
regional corporations would similarly lack immunity. Id. at
392-93. "Congress had a right to assume that the character-
istic energies for corporate enterprise with which a few
months previously it had endowed [the RFC] would now
radiate through [the RFC] to [the regional corporations]."
Id. at 393-94. Because FPI is by no means the offspring of a
non-immune government entity, the grounds for the inference
drawn in Keifer are absent here.
Moreover, the Supreme Court seems to have abandoned
Keifer's fundamental premises. The Keifer Court said that
"the government does not become the conduit of its immunity
in suits against its ... instrumentalities merely because they
do its work." Id. at 388. More recent cases have taken the
opposite view. See FDIC v. Meyer, 510 U.S. at 475 ("Absent
a waiver, sovereign immunity shields the Federal Govern-
ment and its agencies from suit."); Auction Co. of America,
132 F.3d at 752 ("Federal agencies or instrumentalities per-
forming federal functions always fall on the 'sovereign' side of
th[e] fault line."). One treatise classifies Keifer as part of the
Supreme Court's "halting and irregular" pronouncements of
its doubts about sovereign immunity, standing in contrast to
the "regular reiterations ... of the conventional position,
which has been dominant in most recent Supreme Court
decisions." Richard H. Fallon, Jr. et al., Hart & Wechsler's
The Federal Courts and the Federal System 1040 (4th ed.
1996). Obviously we must apply the currently "conventional"
position.
In short, reading s 4121 as an incorporation of the details
of the District's general corporation law (including its "sue
and be sued" clause) is not especially plausible. That law has
a variety of specific features that are either irrelevant to FPI
or contradict provisions in its organic statute; the organic
statute contemplates intra-governmental resolution of con-
flicts with FPI's primary customers; and the language of
s 4121 is nowhere near as specific as in the recognized
instances of such incorporation. More limited readings than
Galvan's, simply locating FPI in the District, are at least as
plausible. We find no waiver here.
The False Claims Act. Nor can we find a waiver in the
False Claims Act. The Act establishes liability for any
"person" who knowingly presents false or fraudulent claims.
31 U.S.C. s 3729. The term "person," however, is not de-
fined for the relevant sections of the statute. Galvan con-
tends that the term "person" should include government
corporations because 1 U.S.C. s 1 provides that the word
"person" is to include "corporations" unless "the context
indicates otherwise." The government responds by invoking
the counter canon that the term "person" does not ordinarily
include the sovereign. See Will v. Michigan Dep't of State
Police, 491 U.S. 58, 64 (1989) (citing Wilson v. Omaha Indian
Tribe, 442 U.S. 653, 667 (1979)).
The parties also point to specific contextual elements. Gal-
van urges that the language of 31 U.S.C. s 3730(b)(5), saying
that "no person other than the Government may intervene,"
implies that the term "person" includes the government and
thus resolves any ambiguity. He also argues that the exemp-
tion of certain officials from liability under s 3730(e)(2)(A)
implies that all other government entities are suable, and that
therefore government corporations fall within the definition of
person in 1 U.S.C. s 1. The government points to
s 3730(d)(2), which makes the "defendant" liable for a pre-
vailing relator's expenses, and s 3730(f), which states that the
government cannot be found liable for the relator's expenses;
taken together, these provisions arguably imply that "the
government" cannot be the defendant.
There are answers to each of these arguments. That
s 3730(b)(5) uses the term "person" in excluding all possible
intervenors other than "the Government" sheds little light on
the congressional view of proper defendants. Section
3730(e)(2)(A)'s explicit but limited immunity for individuals
holding specified offices is completely consistent with an
assumed immunity for government entities themselves; re-
covery from fraudulent individuals would not involve two
branches of the federal government running up litigation
costs so that one can collect from another. And the statutory
combination cited by the government, creating "defendant"
liability for attorneys' fees while protecting "the Government"
from such liability, can be explained by limiting the govern-
ment's exemption to appearances in its capacity as intervenor.
In the end, none of these contextual arguments seems to offer
any strong ground for interpretation of "person" one way or
the other.
We note that the circuits have split over whether "person"
under the False Claims Act includes states. Compare United
States ex rel. Long v. SCS Business & Tech. Inst., 173 F.3d
870 (D.C. Cir. 1999) (finding that it does not), with United
States ex rel. Stevens v. Vermont Agency of Natural Re-
sources, 162 F.3d 195 (2d Cir. 1998), cert. granted, 119 S. Ct.
2391 (June 24, 1999) (finding that it does), and United States
ex rel. Zissler v. Regents of the Univ. of Minn., 154 F.3d 870
(8th Cir. 1998) (same).6 The courts have necessarily resolved
this question under the presumption established in the Will
and Wilson cases; that two circuits found the presumption
overcome in that context is a source of caution. But for state
liability there was a stronger basis in the overall purpose and
legislative history for inclusion, see, e.g., Long, 173 F.3d at
875-81, and the analysis was not subject to any interpretive
guidance as strong as Nordic Village's instruction that the
presence of a "plausible" non-waiver interpretation compels
rejection of a waiver interpretation. Because the reading of
"person" to exclude agencies of the federal government is at
least plausible, we find no waiver.
__________
6 Since granting certiorari in Stevens, the Supreme Court has
added the broader question of whether a private person can have
standing to bring a qui tam action in the absence of particularized
injury attributable to the defendant's actions. See Stevens, 1999
WL 1045146, at *1 (U.S. Nov. 19, 1999) (adding standing question).
At oral argument, Galvan's counsel attempted to raise an
additional theory under which Galvan could recover pursuant
to the False Claims Act. He argued that the Tucker Act, 28
U.S.C. s 1491--and the Little Tucker Act, id. s 1346(a)(2), to
the extent that the claim did not exceed $10,000--waived
sovereign immunity: s 1491(a)(1) waives for a claim "founded
... upon ... any Act of Congress," and the Supreme Court
has understood that to encompass any statute that "can fairly
be interpreted as mandating compensation by the Federal
Government for the damages sustained." United States v.
Mitchell, 463 U.S. 206, 218 (1983). The False Claims Act,
says Galvan, is such an act. As the False Claims Act does
not specifically impose any obligation on a branch of the
United States, Galvan's argument seems quite a stretch of the
Mitchell principle. But we shall not explore Galvan's argu-
ment because it was not raised before the district court or in
Galvan's submissions to this court. We do not normally
"consider all the implications of a theory vaguely raised for
the first time at oral argument on appeal." Tarpley v.
Greene, 684 F.2d 1, 7 n.17 (D.C. Cir. 1982).
Galvan's claim must be dismissed because the FPI enjoyed
an unwaived sovereignty immunity; the judgment of the
district court is
Affirmed.