203 F.3d 819, affirmed.
[ Rehnquist ]
[ Scalia ]
[ Ginsburg ]
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version

Scalia, J., concurring


No. 99—1848



[May 29, 2001]

Justice Scalia, with whom Justice Thomas joins, concurring.

I join the opinion of the Court in its entirety, and write to respond at greater length to the contentions of the dissent.


“Prevailing party” is not some newfangled legal term invented for use in late-20th-century fee-shifting statutes. “[B]y the long established practice and universally recognized rule of the common law, in actions at law, the prevailing party is entitled to recover a judgment for costs … .” Mansfield, C. & L. M. R. Co. v. Swan, 111 U.S. 379, 387 (1884).

“Costs have usually been allowed to the prevailing party, as incident to the judgment, since the statute 6 Edw. I, c. 1, §2, and the same rule was acknowledged in the courts of the States, at the time the judicial system of the United States was organized. . . .

“Weighed in the light of these several provisions in the Judiciary Act [of 1789], the conclusion appears to be clear that Congress intended to allow costs to the prevailing party, as incident to the judgment . . . .” The Baltimore, 8 Wall. 377, 388, 390 (1869).

The term has been found within the United States Statutes at Large since at least the Bankruptcy Act of 1867, which provided that “[t]he party prevailing in the suit shall be entitled to costs against the adverse party.” Act of Mar. 2, 1867, ch. 176, §24, 14 Stat. 528. See also Act of Mar. 3, 1887, ch. 359, §15, 24 Stat. 508 (“If the Government of the United States shall put in issue the right of the plaintiff to recover the court may, in its discretion, allow costs to the prevailing party from the time of joining such issue”). A computer search shows that the term “prevailing party” appears at least 70 times in the current United States Code; it is no stranger to the law.

At the time 42 U.S.C. § 1988 was enacted, I know of no case, state or federal, in which–either under a statutory invocation of “prevailing party,” or under the common-law rule–the “catalyst theory” was enunciated as the basis for awarding costs. Indeed, the dissent cites only one case in which (although the “catalyst theory” was not expressed) costs were awarded for a reason that the catalyst theory would support, but today’s holding of the Court would not: Baldwin v. Chesapeake & Potomac Tel. Co., 156 Md. 552, 557, 144 A. 703, 705 (1929), where costs were awarded because “the granting of [appellee’s] motion to dismiss the appeal has made it unnecessary to inquire into the merits of the suit, and the dismissal is based on an act of appellee performed after both the institution of the suit and the entry of the appeal.” And that case is irrelevant to the meaning of “prevailing party,” because it was a case in equity. While, as Mansfield observed, costs were awarded in actions at law to the “prevailing party,” see 111 U.S., at 387, an equity court could award costs “as the equities of the case might require,” Getz v. Johnston, 145 Md. 426, 433, 125 A. 689, 691 (1924). See also Horn v. Bohn, 96 Md. 8, 12—13, 53 A. 576, 577 (1902) (“The question of costs in equity cases is a matter resting in the sound discretion of the Court, from the exercise of which no appeal will lie”) (internal quotation marks and citation omitted).1 The other state or state-law cases the dissent cites as awarding costs despite the absence of a judgment all involve a judicial finding–or its equivalent, an acknowledgement by the defendant–of the merits of plaintiff’s case.2 Moreover, the dissent cites not a single case in which this Court–or even any other federal court applying federal law prior to enactment of the fee-shifting statutes at issue here–has regarded as the “prevailing party” a litigant who left
the courthouse emptyhanded. If the term means what
the dissent contends, that is a remarkable absence of authority.

That a judicial finding of liability was an understood requirement of “prevailing” is confirmed by many statutes that use the phrase in a context that presumes the existence of a judicial ruling. See, e.g., 5 U.S.C. § 1221(g)(2) (“[i]f an employee . . . is the prevailing party . . . and the decision is based on a finding of a prohibited personnel practice”); §1221(g)(3) (providing for an award of attorney’s fees to the “prevailing party,” “regardless of the basis of the decision”); §7701(b)(2)(A) (allowing the prevailing party to obtain an interlocutory award of the “relief provided in the decision”); 8 U.S.C. § 1324b(h) (permitting the administrative law judge to award an attorney’s fee to the prevailing party “if the losing party’s argument is without reasonable foundation in law and fact”); 18 U.S.C. § 1864(e) (1994 ed., Supp. V) (allowing the district court to award the prevailing party its attorney’s fee “in addition to monetary damages”).

The dissent points out, post, at 8—9, that the Prison Litigation Reform Act of 1995 limits attorney’s fees to an amount “ ‘proportionately related to the court ordered relief for the violation.’ ” This shows that sometimes Congress does explicitly “tightly bind fees to judgments,” post, at 8, inviting (the dissent believes) the conclusion that “prevailing party” does not fasten fees to judgments. That conclusion does not follow from the premise. What this statutory provision demonstrates, at most, is that use of the phrase “prevailing party” is not the only way to impose a requirement of court-ordered relief. That is assuredly true. But it would be no more rational to reject the normal meaning of “prevailing party” because some statutes produce the same result with different language, than it would be to conclude that, since there are many synonyms for the word “jump,” the word “jump” must mean something else.

It is undoubtedly true, as the dissent points out by quoting a nonlegal dictionary, see post, at 12—13, that the word “prevailing” can have other meanings in other contexts: “prevailing winds” are the winds that predominate, and the “prevailing party” in an election is the party that wins the election. But when “prevailing party” is used by courts or legislatures in the context of a lawsuit, it is a term of art. It has traditionally–and to my knowledge, prior to enactment of the first of the statutes at issue here, invariably–meant the party that wins the suit or obtains a finding (or an admission) of liability. Not the party that ultimately gets his way because his adversary dies before the suit comes to judgment; not the party that gets his way because circumstances so change that a victory on the legal point for the other side turns out to be a practical victory for him; and not the party that gets his way because the other side ceases (for whatever reason) its offensive conduct. If a nuisance suit is mooted because the defendant asphalt plant has gone bankrupt and ceased operations, one would not normally call the plaintiff the prevailing party. And it would make no difference, as far as the propriety of that characterization is concerned, if the plant did not go bankrupt but moved to a new location to avoid the expense of litigation. In one sense the plaintiff would have “prevailed”; but he would not be the prevailing party in the lawsuit. Words that have acquired a specialized meaning in the legal context must be accorded their legal meaning.

“[W]here Congress borrows terms of art in which are accumulated the legal tradition and meaning of centuries of practice, it presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken and the meaning its use will convey to the judicial mind unless otherwise instructed. In such case, absence of contrary direction may be taken as satisfaction with widely accepted definitions, not as a departure from them.” Morissette v. United States, 342 U.S. 246, 263 (1952).

The cases cited by the dissent in which we have “not treated Black’s Law Dictionary as preclusively definitive,” post, at 8, are inapposite. In both Pioneer Investment Services Co. v. Brunswick Associates Ltd. Partnership, 507 U.S. 380 (1993), and United States v. Rodgers, 466 U.S. 475 (1984), we rejected Black’s definition because it conflicted with our precedent. See Pioneer, supra, at 395—396 n. 14; Rodgers, supra, at 480. We did not, as the dissent would do here, simply reject a relevant definition of a word tailored to judicial settings in favor of a more general definition from another dictionary.


The dissent distorts the term “prevailing party” beyond its normal meaning for policy reasons, but even those seem to me misguided. They rest upon the presumption that the catalyst theory applies when “the suit’s merit led the defendant to abandon the fray, to switch rather than fight on, to accord plaintiff sooner rather than later the principal redress sought in the complaint,” post, at 1 (emphasis added). As the dissent would have it, by giving the term its normal meaning the Court today approves the practice of denying attorney’s fees to a plaintiff with a proven claim of discrimination, simply because the very merit of his claim led the defendant to capitulate before judgment. That is not the case. To the contrary, the Court approves the result in Parham v. Southwestern Bell Tel. Co., 433 F.2d 421 (CA8 1970), where attorney’s fees were awarded “after [a] finding that the defendant had acted unlawfully,” ante, at 9, and n. 9.3 What the dissent’s stretching of the term produces is something more, and something far less reasonable: an award of attorney’s fees when the merits of plaintiff’s case remain unresolved–when, for all one knows, the defendant only “abandon[ed] the fray” because the cost of litigation–either financial or in terms of public relations–would be too great. In such a case, the plaintiff may have “prevailed” as Webster’s defines that term–“gain[ed] victory by virtue of strength or superiority,” see post, at 12. But I doubt it was greater strength in financial resources, or superiority in media manipulation, rather than superiority in legal merit, that Congress intended to reward.

It could be argued, perhaps, that insofar as abstract justice is concerned, there is little to choose between the dissent’s outcome and the Court’s: If the former sometimes rewards the plaintiff with a phony claim (there is no way of knowing), the latter sometimes denies fees to the plaintiff with a solid case whose adversary slinks away on the eve of judgment. But it seems to me the evil of the former far outweighs the evil of the latter. There is all the difference in the world between a rule that denies the extraordinary boon of attorney’s fees to some plaintiffs who are no less “deserving” of them than others who receive them, and a rule that causes the law to be the very instrument of wrong–exacting the payment of attorney’s fees to the extortionist.

It is true that monetary settlements and consent decrees can be extorted as well, and we have approved the award of attorney’s fees in cases resolved through such mechanisms. See ante, at 5—6 (citing cases). Our decision that the statute makes plaintiff a “prevailing party” under such circumstances was based entirely on language in a House Report, see Maher v. Gagne, 448 U.S. 122, 129 (1980), and if this issue were to arise for the first time today, I doubt whether I would agree with that result. See Hewitt v. Helms, 482 U.S. 755, 760 (1987) (Scalia, J.) (opining that “[r]espect for ordinary language requires that a plaintiff receive at least some relief on the merits of his claim before he can be said to prevail” (emphasis added)). But in the case of court-approved settlements and consent decrees, even if there has been no judicial determination of the merits, the outcome is at least the product of, and bears the sanction of, judicial action in the lawsuit. There is at least some basis for saying that the party favored by the settlement or decree prevailed in the suit. Extending the holding
of Maher to a case in which no judicial action whatever has been taken stretches the term “prevailing party” (and the potential injustice that Maher produces) beyond what the normal meaning of that term in the litigation context can conceivably support.

The dissent points out that petitioners’ object in bringing their suit was not to obtain “a judge’s approbation,” but to “stop enforcement of a [West Virginia] rule,” post, at 13; see also Hewitt, supra, at 761. True enough. But not even the dissent claims that if a petitioner accumulated attorney’s fees in preparing a threatened complaint, but never filed it prior to the defendant’s voluntary cessation of its offending behavior, the wannabe-but-never-was plaintiff could recover fees; that would be countertextual, since the fee-shifting statutes require that there be an “action” or “proceeding,” see 42 U.S.C. § 3613(d); §1988(b) (1994 ed., Supp. V)–which in legal parlance (though not in more general usage) means a lawsuit. See post, at 23 (concluding that a party should be deemed prevailing as a result of a “postcomplaint payment or change in conduct”). Does that not leave achievement of the broad congressional purpose identified by the dissent just as unsatisfactorily incomplete as the failure to award fees when there is no decree? Just as the dissent rhetorically asks why (never mind the language of the statute) Congress would want to award fees when there is a judgment, but deny fees when the defendant capitulates on the eve of judgment; so also it is fair for us to ask why Congress would want to award fees when suit has been filed, but deny fees when the about-to-be defendant capitulates under the threat of filing. Surely, it cannot be because determination of whether suit was actually contemplated and threatened is too difficult. All the proof takes is a threatening letter and a batch of timesheets. Surely that obstacle would not deter the Congress that (according to the dissent) was willing to let district judges pursue that much more evasive will-o’-the-wisp called “catalyst.” (Is this not why we have district courts?, asks the dissent, post, at 19.) My point is not that it would take no more twisting of language to produce prelitigation attorney’s fees than to produce the decreeless attorney’s fees that the dissent favors (though that may well be true). My point is that the departure from normal usage that the dissent favors cannot be justified on the ground that it establishes a regime of logical even handedness. There must be a cutoff of seemingly equivalent entitlements to fees–either the failure to file suit in time or the failure to obtain a judgment in time. The term “prevailing party” suggests the latter rather than the former. One does not prevail in a suit that is never determined.

The dissent's ultimate worry is that today’s opinion will “impede access to court for the less well-heeled,” post, at 1. But, of course, the catalyst theory also harms the “less well-heeled,” putting pressure on them to avoid the risk of massive fees by abandoning a solidly defensible case early in litigation. Since the fee-shifting statutes at issue here allow defendants as well as plaintiffs to receive a fee award, we know that Congress did not intend to maximize the quantity of “the enforcement of federal law by private attorneys general,” ibid. Rather, Congress desired an appropriate level of enforcement–which is more likely to be produced by limiting fee awards to plaintiffs who prevail “on the merits,” or at least to those who achieve an enforceable “alteration of the legal relationship of the parties,” than by permitting the open-ended inquiry approved by the dissent.4


The dissent points out that the catalyst theory has been accepted by “the clear majority of Federal Circuits,” post, at 2. But our disagreeing with a “clear majority” of the Circuits is not at all a rare phenomenon. Indeed, our opinions sometimes contradict the unanimous and long-standing interpretation of lower federal courts. See, e.g., McNally v. United States, 483 U.S. 350, 365 (1987) (Stevens, J., dissenting) (the Court’s decision contradicted “[e]very court to consider” the question).

The dissent’s insistence that we defer to the “clear majority” of Circuit opinion is particularly peculiar in the present case, since that majority has been nurtured and preserved by our own misleading dicta (to which I, unfortunately, contributed). Most of the Circuit Court cases cited by the dissent, post, at 6, and n. 5, as reaffirming the catalyst theory after our decision in Farrar v. Hobby, 506 U.S. 103 (1992), relied on our earlier opinion in Hewitt. See Marbley v. Bane, 57 F.3d 224, 234 (CA2 1995) (relying on Hewitt to support catalyst theory); Payne v. Board of Ed., 88 F.3d 392, 397 (CA6 1996) (same); Baumgartner v. Harrisburg Housing Auth., 21 F.3d 541, 548 (CA3 1994) (explicitly rejecting Farrar in favor of Hewitt); Zinn v. Shalala, 35 F.3d 273, 274—276 (CA7 1994) (same); Beard v. Teska, 31 F.3d 942, 950—952 (CA10 1994) (same); Morris v. West Palm Beach, 194 F.3d 1203, 1207 (CA11 1999) (same). Deferring to our colleagues’ own error is bad enough; but enshrining the error that we ourselves have improvidently suggested and blaming it on the near-unanimous judgment of our colleagues would surely be unworthy.5 Informing the Courts of Appeals that our ill-considered dicta have misled them displays, it seems to me, not “disrespect,” but a most becoming (and well-deserved) humility.

* * *

The Court today concludes that a party cannot be deemed to have prevailed, for purposes of fee-shifting statutes such as 42 U.S.C. § 1988 3613(c)(2), unless there has been an enforceable “alteration of the legal relationship of the parties.” That is the normal meaning of “prevailing party” in litigation, and there is no proper basis for departing from that normal meaning. Congress is free, of course, to revise these provisions–but it is my guess that if it does so it will not create the sort of inequity that the catalyst theory invites, but will require the court to determine that there was at least a substantial likelihood that the party requesting fees would have prevailed.


1. The jurisdiction that issued Baldwin has used the phrase “prevailing party” frequently (including in equity cases) to mean the party acquiring a judgment. See Getz v. Johnston, 145 Md. 426, 434, 125 A. 689, 691—692 (1924) (an equity decision noting that “ [O]n reversal, following the usual rule, the costs will generally go to the prevailing party, that is, to the appellant” (internal quotation marks and citation omitted)). See also, e.g., Hoffman v. Glock, 20 Md. App. 284, 293, 315 A. 2d 551, 557 (1974) (“Md. Rule 604a provides: ‘Unless otherwise provided by law, or ordered by the court, the prevailing party shall be entitled to the allowance of court costs, which shall be taxed by the clerk and embraced in the judgment’ ”); Fritts v. Fritts, 11 Md. App. 195, 197, 273 A. 2d 648, 649 (1971) (“We have viewed the evidence, as we must, in a light most favorable to appellee as the prevailing party below”); Chillum-Adelphi Volunteer Fire- Dept., Inc. v. Button & Goode, Inc., 242 Md. App. 509, 516, 219 A. 2d 801, 805 (1966) (“At common law, an arbitration award became a cause of action in favor of the prevailing party”); Burch v. Scott, 1829 WL 1006, *15 (Md. Ct. App., Dec. 1829) (“[T]he demurrer being set down to be argued, the court proceeds to affirm or reverse the decree, and the prevailing party takes the deposite”).

2. Our decision to award costs in Mansfield, C. & L. M. R. Co. v. Swan, 111 U.S. 379 (1884), does not “tu[g] against the restrictive rule today’s decision installs,” post, at 9. Defendants had removed the case to federal court, and after losing on the merits, sought to have us vacate the judgment because the basis for removal (diversity of citizenship) was absent. We concluded that because defendants were responsible for the improper removal in the first place, our judgment’s “effect [was] to defeat the entire proceeding which they originated and have prosecuted,” id., at 388. In other words, plaintiffs “prevailed” because defendants’ original position as to jurisdiction was defeated. In Ficklen v. Danville, 146 Va. 426, 438—439 132 S. E. 705, 706 (1926), appellants were deemed to have “substantially prevail[ed]” on their appeal because appellees “abandoned their contention made before the lower court,” i.e., “abandoned their intention and desire to rely upon the correctness of the trial court’s decree.” In Talmage v. Monroe, 119 P. 526 (Cal. App. 1911), costs were awarded after the defendant complied with an alternative writ of mandamus; it was the writ, not the mere petition, which led to defendant’s action. Scatcherd v. Love, 166 F. 53 (CA6 1908), Wagner v. Wagner, 9 Pa. 214 (1848), and other cases cited by the dissent represent a rule adopted in some States that by settling a defendant “acknowledged his liability,” Scatcherd, supra, at 56; see also Wagner, supra, at 215. That rule was hardly uniform among the States. Compare 15 C. J. 89, §167 (1918) (citing cases from 13 States which hold that a “settlement is equivalent to a confession of judgment”), with id., at 89—90, §168, and n. a (citing cases from 11 States which hold that under a settlement “plaintiff cannot recover costs,” because “ [c]osts . . . can only follow a judgment or final determination of the action ” (internal quotation marks and citation omitted)). I do not think these state cases (and Scatcherd, a federal case applying state law) justify expanding the federal meaning of “prevailing party” (based on a “confession of judgment” fiction) to include the party accepting an out-of-court settlement–much less to expand it beyond settlements, to the domain of the “catalyst theory.” The only case cited by the dissent in which the conclusion of acknowledgment of liability was rested on something other than a settlement is Board of Ed. of Madison County v. Fowler, 192 Ga. 35, 14 S. E. 2d 478 (1941), which, in one of the states that considered settlement an acknowledgment of liability, analogized compliance with what had been sought by a mandamus suit to a settlement. This is a slim reed upon which to rest the broad conclusion of a catalyst theory.

3. The dissent incorrectly characterizes Parham as involving an undifferentiated “finding or retention of jurisdiction,” post, at 17, n. 11. In fact, Parham involved a finding that defendant had discriminated, and jurisdiction was retained so that that finding could be given effect, in the form of injunctive relief, should the defendant ever backslide in its voluntary provision of relief to plaintiffs. Jurisdiction was not retained to determine whether there had been discrimination, and I do not read the Court’s opinion as suggesting a fee award would be appropriate in those circumstances. The dissent notes that two other cases were cited in Senate legislative history (Parham is cited in legislative history from both the Senate and House) which it claims support the catalyst theory. If legislative history in general is a risky interpretive tool, legislative history from only one legislative chamber–and consisting of the citation of Court of Appeals cases that surely few if any Members of Congress read–is virtually worthless. In any event, Kopet v. Esquire Realty Co., 523 F.2d 1005 (CA2 1975), does not support the catalyst theory because defendant’s voluntary compliance was not at issue. Fees were awarded on the dubious premise that discovery uncovered some documents of potential use in other litigation, making this more a case of an award of interim fees. Thomas v. Honeybrook Mines, 428 F.2d 981 (CA3 1970), is also inapposite. There, the question was whether counsel for union members whose fruitless efforts to sue the union had nonetheless spurred the union to sue the employer, should be paid out of a fund established by the union’s victory. Whether the union members were “prevailing parties” in the union suit, or whether they were entitled to attorney’s fees as “prevailing parties” in the earlier suit against the union, was not even at issue.

4. Even the legislative history relied upon by the dissent supports the conclusion that some merit is necessary to justify a fee award. See post, at 15, n. 9 (citing a House Report for the proposition that fee-shifting statutes are “ ‘designed to give [‘victims of civil rights violation’] access to the judicial process’ ” (emphasis added)); ibid. (citing a Senate Report: “ ‘[I]f those who violate the Nation’s fundamental laws are not to proceed with impunity,’ ” fee awards are necessary (emphasis added)). And for the reasons given by the Court, see ante at 6—7, the catalyst theory’s purported “merit test”–the ability to survive a motion to dismiss for failure to state a claim, or the absence of frivolousness–is scant protection for the innocent.

5. That a few cases adopting the catalyst theory predate Hewitt v. Helms, 482 U.S. 755 (1987), see post, at 5, and n. 4, is irrelevant to my point. Absent our dicta in Hewitt, and in light of everything else we have said on this topic, see ante, at 5—6, it is unlikely that the catalyst theory would have achieved that universality of acceptance by the Courts of Appeals upon which the dissent relies.