In the latest chapter of this long-litigated original-jurisdiction case, the Quechan Tribe (Tribe) and the United States on the Tribe's behalf assert claims for increased rights to water from the Colorado River. These claims are based on the contention that the Fort Yuma (Quechan) Indian Reservation encompasses some 25,000 acres of disputed boundary lands not attributed to that reservation in earlier stages of the litigation. In this decision, we resolve a threshold question regarding these claims to additional water rights: Are the claims precluded by this Court's prior decision in Arizona v. California, 373 U.S. 546 (1963) (Arizona I), or by a consent judgment entered by the United States Claims Court in 1983? The Special Master has prepared a report recommending that the Court reject the first ground for preclusion but accept the second. We reject both grounds for preclusion and remand the case to the Special Master for consideration of the claims for additional water rights appurtenant to the disputed boundary lands.

* This litigation began in 1952 when Arizona invoked our original jurisdiction to settle a dispute with California over the extent of each State's right to use water from the Colorado River system. Nevada intervened, seeking a determination of its water rights, and Utah and New Mexico were joined as defendants. The United States intervened and sought water rights on behalf of various federal establishments, including five Indian reservations: the Chemehuevi Indian Reservation, the Cocopah Indian Reservation, the Fort Yuma (Quechan) Indian Reservation, the Colorado River Indian Reservation, and the Fort Mojave Indian Reservation. The Court appointed Simon Rifkind as Special Master.

The first round of the litigation culminated in our opinion in Arizona I. We agreed with Special Master Rifkind that the apportionment of Colorado River water was governed by the Boulder Canyon Project Act of 1928, 43 U.S.C. § 617 et seq., and by contracts entered into by the Secretary of the Interior pursuant to the Act. We further agreed that the United States had reserved water rights for the five reservations under the doctrine of Winters v. United States, 207 U.S. 564 (1908). See Arizona I, 373 U.S., at 565, 599_601. Because the Tribes' water rights were effective as of the time each reservation was created, the rights were considered present perfected rights and given priority under the Act. Id., at 600. We also agreed with the Master that the reservations' water rights should be based on the amount of practicably irrigable acreage on each reservation and sustained his findings as to the relevant acreage for each reservation. Ibid. Those findings were incorporated in our decree of March 9, 1964, which specified the quantities and priorities of the water entitlements for the States, the United States, and the Tribes. Arizona v. California, 376 U.S. 340. The Court rejected as premature, however, Master Rifkind's recommendation to determine the disputed boundaries of the Fort Mojave and Colorado River Indian Reservations; we ordered, instead, that water rights for those two reservations "shall be subject to appropriate adjustment by agreement or decree of this Court in the event that the boundaries of the respective reservations are finally determined." Id., at 345.

In 1978, the United States and the State parties jointly moved this Court to enter a supplemental decree identifying present perfected rights to the use of mainstream water in each State and their priority dates. The Tribes then filed motions to intervene, and the United States ultimately joined the Tribes in moving for additional water rights for the five reservations. Again, the Court deferred resolution of reservation boundary disputes and allied water rights claims. The supplemental decree we entered in 1979 set out the water rights and priority dates for the five reservations under the 1964 decree, but added that the rights for all five reservations (including the Fort Yuma Indian Reservation at issue here) "shall continue to be subject to appropriate adjustment by agreement or decree of this Court in the event that the boundaries of the respective reservations are finally determined." Arizona v. California, 439 U.S. 419, 421 (per curiam). The Court then appointed Senior Circuit Judge Elbert P. Tuttle as Special Master and referred to him the Tribes' motions to intervene and other pending matters.

Master Tuttle issued a report recommending that the Tribes be permitted to intervene, and concluding that various administrative actions taken by the Secretary of the Interior constituted "final determinations" of reservation boundaries for purposes of allocating water rights under the 1964 decree. (Those administrative actions included a 1978 Secretarial Order, discussed in greater detail infra, at 9, which recognized the Quechan Tribe's entitlement to the disputed boundary lands of the Fort Yuma Reservation.) Master Tuttle also concluded that certain lands within the undisputed reservation boundaries but for which the United States had not sought water rights in Arizona I-the so-called "omitted lands"-had in fact been practicably irrigable at the time of Arizona I and were thus entitled to water. On these grounds, Master Tuttle recommended that the Court reopen the 1964 decree to award the Tribes additional water rights.

In Arizona v. California, 460 U.S. 605 (1983) (Arizona II), the Court permitted the Tribes to intervene, but otherwise rejected Master Tuttle's recommendations. The Secretary's determinations did not qualify as "final determinations" of reservation boundaries, we ruled, because the States, agencies, and private water users had not had an opportunity to obtain judicial review of those determinations. Id., at 636_637. In that regard, we noted that California state agencies had initiated an action in the United States District Court for the Southern District of California challenging the Secretary's decisions, and that the United States had moved to dismiss that action on various grounds, including sovereign immunity. "There will be time enough," the Court stated, "if any of these grounds for dismissal are sustained and not overturned on appellate review, to determine whether the boundary issues foreclosed by such action are nevertheless open for litigation in this Court." Id., at 638. The Court also held that the United States was barred from seeking water rights for the lands omitted from presentation in the proceedings leading to Arizona I; "principles of res judicata," we said, "advise against reopening the calculation of the amount of practicably irrigable acreage." 460 U.S., at 626. In 1984, in another supplemental decree, the Court again declared that water rights for all five reservations "shall be subject to appropriate adjustments by agreement or decree of this Court in the event that the boundaries of the respective reservations are finally determined." Arizona v. California, 466 U.S. 144, 145.

The district court litigation proceeded with the participation of eight parties: the United States, the States of Arizona and California, the Metropolitan Water District of Southern California, the Coachella Valley Water District, and the Quechan, Fort Mojave, and Colorado River Indian Tribes. The District Court rejected the United States' sovereign immunity defense; taking up the Fort Mojave Reservation matter first, the court voided the Secretary's determination of that reservation's boundaries. Metropolitan Water Dist. of S. Cal. v. United States, 628 F. Supp. 1018 (SD Cal. 1986). The Court of Appeals for the Ninth Circuit, however, accepted the United States' plea of sovereign immunity, and on that ground reversed and remanded with instructions to dismiss the entire case. Specifically, the Court of Appeals held that the Quiet Title Act, 28 U.S.C. § 2409a preserved the United States' sovereign immunity from suits challenging the United States' title "to trust or restricted Indian lands," §2409a(a), and therefore blocked recourse to the District Court by the States and state agencies. Metropolitan Water Dist. of S. Cal. v. United States, 830 F.2d 139 (1987). We granted certiorari and affirmed the Ninth Circuit's judgment by an equally divided Court. California v. United States, 490 U.S. 920 (1989) (per curiam).

The dismissal of the district court action dispelled any expectation that a "final determination" of reservation boundaries would occur in that forum. The State parties then moved to reopen the 1964 decree, asking the Court to determine whether the Fort Yuma Indian Reservation and two other reservations were entitled to claim additional boundary lands and, if so, additional water rights. Neither the United States nor the Tribes objected to the reopening of the decree, and the Court granted the motion. Arizona v. California, 493 U.S. 886 (1989). After the death in 1990 of the third Special Master, Robert McKay, the Court appointed Frank J. McGarr as Special Master. Special Master McGarr has now filed a report and recommendation (McGarr Report), a full understanding of which requires a discussion of issues and events specific to the Fort Yuma Indian Reservation. We now turn to those issues and events.

The specific dispute before us has its roots in an 1884 Executive Order signed by President Chester A. Arthur, designating approximately 72 square miles of land along the Colorado River in California as the Fort Yuma Indian Reservation (Reservation) for the benefit of the Quechan Tribe. The Tribe, which had traditionally engaged in farming, offered to cede its rights to a portion of the Reservation to the United States in exchange for allotments of irrigated land to individual Indians. In 1893, the Secretary of the Interior concluded an agreement with the Tribe (1893 Agreement), which Congress ratified in 1894. The 1893 Agreement provided for the Tribe's cession of a 25,000-acre tract of boundary lands on the Reservation. Language in the agreement, however, could be read to condition the cession on the performance by the United States of certain obligations, including construction within three years of an irrigation canal, allotment of irrigated land to individual Indians, sale of certain lands to raise revenues for canal construction, and opening of certain lands to the public domain.

Doubts about the validity and effect of the 1893 Agreement arose as early as 1935. In that year the construction of the All-American Canal, which prompted the interstate dispute in Arizona I, see 373 U.S., at 554_555, also sparked a controversy concerning the Fort Yuma Reservation. When the Department of the Interior's Bureau of Reclamation sought to route the canal through the Reservation, the Department's Indian Office argued that the Bureau had to pay compensation to the Tribe for the right-of-way. The Secretary of the Interior submitted the matter to the Department's Solicitor, Nathan Margold. In 1936, Solicitor Margold issued an opinion (Margold Opinion) stating that, under the 1893 Agreement, the Tribe had unconditionally ceded the lands in question to the United States. 1 Dept. of Interior, Opinions of the Solicitor Relating to Indian Affairs 596, 600 (No. M_28198, Jan. 8, 1936). The Margold Opinion remained the position of the Federal Government for 42 years.

In 1946, Congress enacted the Indian Claims Commission Act, 60 Stat. 1049, 25 U.S.C. § 70 et seq. (1976 ed.), establishing an Article I tribunal with power to decide claims of Indian tribes against the United States.1 See generally United States v. Dann, 470 U.S. 39 (1985). The Tribe filed an action before the Commission in 1951, challenging the validity and effect of the 1893 Agreement. In that action, referred to by the parties as Docket No. 320, the Tribe relied principally on two mutually exclusive grounds for relief. First, the Tribe alleged that the 1893 Agreement was obtained through fraud, coercion, and/or inadequate consideration, rendering it "wholly nugatory." Petition for Loss of Reservation in Docket No. 320 (Ind. Cl. Comm'n.), ¶¶15_16, reprinted in Brief for United States in Support of Exception, pp. 11a_27a. At the very least, contended the Tribe, the United States had failed to perform the obligations enumerated in the 1893 Agreement, rendering the cession void. Id., at ¶31. In either event, the Tribe claimed continuing title to the disputed lands and sought damages essentially for trespass. Alternatively, the Tribe alleged that the 1893 Agreement was contractually valid but constituted an uncompensated taking of tribal lands, an appropriation of lands for unconscionable consideration, and/or a violation of standards of fair and honorable dealing, for which §§2(3)_(5) of the Act authorized recovery. Id., at ¶¶19, 22, 25. According to this theory of recovery, the 1893 Agreement had indeed vested in the United States unconditional title to the disputed lands, and the Tribe sought damages as compensation for that taking. During the more than quarter-century of litigation in Docket No. 320, the Tribe vacillated between these two grounds for relief, sometimes emphasizing one and sometimes the other. See Quechan Tribe of Fort Yuma Reservation v. United States, 26 Ind. Cl. Comm'n. 15 (1971), reprinted in Brief for United States in Support of Exception, pp. 29a_34a.

The Commission conducted a trial on liability, but stayed further proceedings in 1970 because legislation had been proposed in Congress that would have restored the disputed lands to the Tribe. The legislation was not enacted, and the Commission vacated the stay. In 1976, the Commission transferred the matter to the Court of Claims.

In the meantime, the Tribe had asked the Department of the Interior to reconsider its 1936 Margold Opinion regarding the 1893 Agreement. In 1977, Interior Solicitor Scott Austin concluded, in accord with the 1936 opinion, that the 1893 Agreement was valid and that the cession of the disputed lands had been unconditional. Opinion of the Solicitor, No. M_36886 (Jan. 18, 1977), 84 I. D. 1 (1977) (Austin Opinion). It soon became clear both to the Tribe and to interested Members of Congress, however, that the Austin Opinion had provoked controversy within the Department, and, after the election of President Carter, the Department revisited the issue and reversed course. In 1978, without notice to the parties, Solicitor Leo Krulitz issued an opinion concluding that the 1893 Agreement had provided for a conditional cession of the disputed lands, that the conditions had not been met by the United States, and that "[t]itle to the subject property is held by the United States in trust for the Quechan Tribe." Opinion of the Solicitor, No. M_36908 (Dec. 20, 1978), 86 I. D. 3, 22 (1979) (Krulitz Opinion). On December 20, 1978, the Secretary of the Interior issued a Secretarial Order adopting the Krulitz Opinion and confirming the Tribe's entitlement to the disputed lands, with the express exception of certain lands that the United States had acquired pursuant to Act of Congress or had conveyed to third parties.

The 1978 Secretarial Order caused the United States to change its position both in Docket No. 320, which was still pending in the Claims Court, and in the present litigation. Because the Secretarial Order amounted to an admission that the 1893 Agreement had been ineffective to transfer title and that the Tribe enjoyed beneficial ownership of the disputed boundary lands, the United States no longer opposed the Tribe's claim for trespass in Docket No. 320. In the present litigation, the Secretarial Order both prompted the United States to file a water rights claim for the affected boundary lands and provided the basis for the Tribe's intervention to assert a similar, albeit larger, water rights claim. See Arizona II, 460 U.S., at 632_633. Those water rights claims are the subject of the current proceedings.

In August 1983, a few months after this Court decided in Arizona II that the 1978 Secretarial Order did not constitute a final determination of reservation boundaries, see supra, at 4, the United States and the Tribe entered into a settlement of Docket No. 320, which the Court of Claims approved and entered as its final judgment. Under the terms of that settlement, the United States agreed to pay the Tribe $15 million in full satisfaction of "all rights, claims, or demands which plaintiff [i.e., the Tribe] has asserted or could have asserted with respect to the claims in Docket 320." Final Judgment, Docket No. 320 (Aug. 11, 1983). The judgment further provided that "plaintiff shall be barred thereby from asserting any further rights, claims, or demands against the defendant and any future action on the claims encompassed on Docket 320." Ibid. The United States and the Tribe also stipulated that the "final judgment is based on a compromise and settlement and shall not be construed as an admission by either party for the purposes of precedent or argument in any other case." Ibid. Both the Tribe and the United States continue to recognize the Tribe's entitlement to the disputed boundary lands.

Master McGarr has issued a series of orders culminating in the report and recommendation now before the Court. He has recommended that the Court reject the claims of the United States and the Tribe seeking additional water rights for the Fort Yuma Indian Reservation. The Master rejected the State parties' contention that this Court's Arizona I decision precludes the United States and the Tribe from seeking water rights for the disputed boundary lands. He concluded, however, that the United States and the Tribe are precluded from pursuing those claims by operation of the 1983 Claims Court consent judgment. The State parties have filed an exception to the first of these preclusion recommendations, and the United States and the Tribe have filed exceptions to the second. In Part III_A, infra, we consider the exception filed by the State parties, and in Part III_B we address the exceptions filed by the United States and the Tribe. The Special Master has also recommended that the Court approve the parties' proposed settlements respecting the Fort Mojave and Colorado River Indian Reservations. No party has filed an exception to those recommendations; we address them in Part III_C, infra.

* The States of Arizona and California, the Coachella Valley Water District, and the Metropolitan Water District of Southern California (State parties) argued before Special Master McGarr, and repeat before this Court, that the water rights claims associated with the disputed boundary lands of the Fort Yuma Reservation are precluded by the finality rationale this Court employed in dismissing the "omitted lands" claims in Arizona II. See supra, at 4. According to the State parties, the United States could have raised a boundary lands claim for the Fort Yuma Reservation in the Arizona I proceedings based on facts known at that time, just as it did for the Fort Mojave and Colorado River Reservations, but deliberately decided not to do so, just as it did with respect to the "omitted lands." In Arizona II, this Court rejected the United States' claim for water rights for the "omitted lands," emphasizing that "[c]ertainty of rights is particularly important with respect to water rights in the Western United States" and noting "the strong interest in finality in this case." 460 U.S., at 620. Observing that the 1964 decree determined "the extent of irrigable acreage within the uncontested boundaries of the reservations," id., at 621, n. 12, the Court refused to reconsider issues "fully and fairly litigated 20 years ago," id., at 621. The Court concomitantly held that the Tribes were bound by the United States' representation of them in Arizona I. Id., at 626_627.

The Special Master rejected the State parties' preclusion argument. He brought out first the evident reason why the United States did not assert water rights claims for the Fort Yuma Reservation boundary lands in Arizona I. At that point in time, the United States was bound to follow the 1936 Margold Opinion, see supra, at 6_7, which maintained that the Tribe had no claim to those lands. "[I]t is clear," the Master stated, "that the later Secretary of the Interior opinion arbitrarily changing [the Margold] decision was a circumstance not known in 1964, thus constituting an exception to the application of the rule of res adjudicata." Special Master McGarr Memorandum Opinion and Order No. 4, pp. 6_7 (Sept. 6, 1991). Characterizing the question as "close," the Master went on to conclude that "the Tribe is not precluded from asserting water rights based on boundary land claims on [sic] this proceeding, because although the U.S. on behalf of the Tribe failed to assert such claims in the proceeding leading to the 1964 decree, a later and then unknown circumstance bars the application of the doctrine of res judicata to this issue." Id., at 7.

While the Special Master correctly recognized the relevance of the Margold Opinion to the litigating stance of the United States, he ultimately relied on an improper ground in rejecting the State parties' preclusion argument. The Department of the Interior's 1978 Secretarial Order recognizing the Tribe's beneficial ownership of the boundary lands, see supra, at 9, does not qualify as a "later and then unknown circumstance" that can overcome otherwise applicable preclusion principles. The 1978 Order did not change the underlying facts in dispute; it simply embodied one party's changed view of the import of unchanged facts. Moreover, the Tribe can hardly claim to have been surprised by the Government's shift in assessment of the boundary lands ownership question, for the Tribe had been advocating just such a shift for decades.

The United States and the Tribe, however, urge other grounds on which to reject the State parties' argument regarding the preclusive effect of Arizona I. The United States and the Tribe maintain that the preclusion rationale the Court applied to the "omitted lands" in Arizona II is not equally applicable to the disputed boundary lands,2 and that, in any event, the State parties have forfeited their preclusion defense. We agree that the State parties' preclusion defense is inadmissible at this late date, and therefore we do not reach the merits of that plea. The State parties could have raised the defense in 1979 in response to the United States' motion for a supplemental decree granting additional water rights for the Fort Yuma Reservation. The State parties did not do so then, nor did they raise the objection in 1982 when Arizona II was briefed and argued.3 Unaccountably, they raised the preclusion argument for the first time in 1989, when they initiated the current round of proceedings. See Exception and Brief for the State Parties 16; Motion of the State Parties to Reopen Decree in Arizona v. California, O. T. 1989, No. 8 Orig., p. 6, n. 2. The State parties had every opportunity, and every incentive, to press their current preclusion argument at earlier stages in the litigation, yet failed to do so.4

"[W]hile the technical rules of preclusion are not strictly applicable [in the context of a single ongoing original action], the principles upon which these rules are founded should inform our decision." Arizona II, 460 U.S., at 619. Those principles rank res judicata an affirmative defense ordinarily lost if not timely raised. See Fed. Rule Civ. Proc. 8(c). Counsel for the State parties conceded at oral argument that "no preclusion argument was made with respect to boundary lands" in the proceedings leading up to Arizona II, and that "after this Court's decision in Arizona II and after the Court's later decision in [Nevada v. United States, 463 U.S. 110 (1983)], the light finally dawned on the State parties that there was a valid preclusion-or res judicata argument here with respect to Fort Yuma." Tr. of Oral Arg. 46_47. We disapprove the notion that a party may wake up because a "light finally dawned," years after the first opportunity to raise a defense, and effectively raise it so long as the party was (though no fault of anyone else) in the dark until its late awakening.

The State parties assert that our prior pronouncements in this case have expressly recognized the possibility that future boundary lands claims for the Fort Yuma Reservation might be precluded. If anything, the contrary is true. Nothing in the Arizona II decision hints that the Court believed the boundary lands issue might ultimately be held precluded. Rather, the Court expressly found it "necessary to decide whether any or all of these boundary disputes have been `finally determined' within the meaning of Article II(D)(5) _ ." 460 U.S., at 631 (emphasis added). That Arizona II contains no discussion of preclusion with respect to the disputed lands is hardly surprising, given that the State parties neglected to raise that issue until six years later.

The Court did note in Arizona II that in the district court proceedings the United States had asserted defenses based on "lack of standing, the absence of indispensable parties, sovereign immunity and the applicable statute of limitations," and added that "[t]here will be time enough, if any of these grounds for dismissal are sustained and not overturned on appellate review, to determine whether the boundary issues foreclosed by such [lower court] action are nevertheless open for litigation in this Court." 460 U.S., at 638 (emphasis added). This passage, however, is most sensibly read to convey that the defenses just mentioned-standing, indispensable parties, sovereign immunity, and the statute of limitations-would not necessarily affect renewed litigation in this Court. The passage contains no acknowledgment, express or implied, of a lurking preclusion issue stemming from our Arizona I disposition.

Moreover, and of large significance, the 1979 and 1984 supplemental decrees anticipated that the disputed boundary issues for all five reservations, including the Fort Yuma Reservation, would be "finally determined" in some forum, not by preclusion but on the merits. See 1984 Supplemental Decree, Art. II(D)(5), Arizona v. California, 466 U.S., at 145 (Water rights for all five reservations "shall be subject to appropriate adjustments by agreement or decree of this Court in the event that the boundaries of the respective reservations are finally determined."); 1979 Supplemental Decree, Art. II(D)(5), Arizona v. California, 439 U.S., at 421 (same).

The State parties themselves stipulated to the terms of the supplemental decree we entered in 1979. They also appear to have litigated the Arizona II proceedings on the understanding that the boundary disputes should be resolved on the merits. See Arizona II, 460 U.S., at 634 ("[The State parties] argued _ that the boundary controversies were ripe for judicial review, and they urged the Special Master to receive evidence, hear legal arguments, and resolve each of the boundary disputes, but only for the limited purpose of establishing additional Indian water rights, if any."); Report of Special Master Tuttle, O. T. 1981, No. 8 Orig., p. 57 (describing the State parties' contention "that the boundaries [of all five Reservations] have not been finally determined and that I should make a de novo determination of the boundaries for recommendation to the Court"). As late as 1988, the State parties asked the Court to appoint a new Special Master and direct him "to conclude his review of the boundary issues as expeditiously as possible and to submit a recommended decision to the Court." Brief for Petitioners in California v. United States, O. T. 1987, No. 87_1165, p. 49.

Finally, the State parties argue that even if they earlier failed to raise the preclusion defense, this Court should raise it now sua sponte. Judicial initiative of this sort might be appropriate in special circumstances. Most notably, "if a court is on notice that it has previously decided the issue presented, the court may dismiss the action sua sponte, even though the defense has not been raised. This result is fully consistent with the policies underlying res judicata: it is not based solely on the defendant's interest in avoiding the burdens of twice defending a suit, but is also based on the avoidance of unnecessary judicial waste." United States v. Sioux Nation, 448 U.S. 371, 432 (1980) (Rehnquist, J., dissenting) (citations omitted). That special circumstance is not present here: While the State parties contend that the Fort Yuma boundary dispute could have been decided in Arizona I, this Court plainly has not "previously decided the issue presented." Therefore we do not face the prospect of redoing a matter once decided. Where no judicial resources have been spent on the resolution of a question, trial courts must be cautious about raising a preclusion bar sua sponte, thereby eroding the principle of party presentation so basic to our system of adjudication.

In view of the State parties' failure to raise the preclusion argument earlier in the litigation, despite ample opportunity and cause to do so, we hold that the claims of the United States and the Tribe to increased water rights for the disputed boundary lands of the Fort Yuma Reservation are not foreclosed by our decision in Arizona I.

The State parties also assert that the instant water rights claims are precluded by the 1983 consent judgment in the Claims Court proceeding, Docket No. 320. Special Master McGarr agreed, noting the consent judgment's declaration that the Tribe would "be barred thereby from asserting any further rights, claims or demands against the defendant and any future action encompassed on docket no. 320." See Special Master McGarr Memorandum Opinion and Order No. 4, pp. 9_10 (Sept. 6, 1991). On reconsideration, the Special Master provided a fuller account of his recommendation. The settlement, he concluded, had extinguished the Tribe's claim to title in the disputed boundary lands, vesting that title in the United States against all the world: "The only viable basis for a damage or trespass claim [in Docket No. 320] was that the 1893 taking was illegal and that title therefore remained with the Tribe. When the Tribe accepted money in settlement of this claim, it relinquished its claim to title." Id., No. 7, p. 5 (May 5, 1992). See also id., No. 13, p. 3 (Apr. 13, 1993) ("[T]he relinquishment of all future claims regarding the subject matter of Docket No. 320 in exchange for a sum of money extinguished the Tribe's title in the subject lands _ ."). Because the settlement extinguished the Tribe's title to the disputed boundary lands, the Master reasoned, the United States and the Tribe cannot now seek additional water rights based on the Tribe's purported beneficial ownership of those lands.

Under standard preclusion doctrine, the Master's recommendation cannot be sustained. As already noted, the express terms of the consent judgment in Docket No. 320 barred the Tribe and the United States from asserting against each other any claim or defense they raised or could have raised in that action. See supra, at 10. As between the parties to Docket No. 320, then, the settlement indeed had, and was intended to have, claim-preclusive effect-a matter the United States and the Tribe readily concede. Exception and Brief for the United States 36; Exception and Brief for the Quechan Indian Tribe 20. But settlements ordinarily occasion no issue preclusion (sometimes called collateral estoppel), unless it is clear, as it is not here, that the parties intend their agreement to have such an effect. "In most circumstances, it is recognized that consent agreements ordinarily are intended to preclude any further litigation on the claim presented but are not intended to preclude further litigation on any of the issues presented. Thus consent judgments ordinarily support claim preclusion but not issue preclusion." 18 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §4443, p. 384_385 (1981). This differentiation is grounded in basic res judicata doctrine. It is the general rule that issue preclusion attaches only "[w]hen an issue of fact or law is actually litigated and determined by a valid and final judgment, and the determination is essential to the judgment." Restatement (Second) of Judgments §27, p. 250 (1982). "In the case of a judgment entered by confession, consent, or default, none of the issues is actually litigated. Therefore, the rule of this Section [describing issue preclusion's domain] does not apply with respect to any issue in a subsequent action." Id., comment e, p. 257.

This Court's decision in United States v. International Building Co., 345 U.S. 502 (1953), is illustrative. In 1942, the Commissioner of Internal Revenue assessed deficiencies against a taxpayer for the taxable years 1933, 1938, and 1939, alleging that the taxpayer had claimed an excessive basis for depreciation. Id., at 503. After the taxpayer filed for bankruptcy, however, the Commissioner and the taxpayer filed stipulations in the pending Tax Court proceedings stating that there was no deficiency for the taxable years in question, and the Tax Court entered a formal decision to that effect. Id., at 503_504. In 1948, the Commissioner assessed deficiencies for the years 1943, 1944, and 1945, and the taxpayer defended on the ground that the earlier Tax Court decision was preclusive on the issue of the correct basis for depreciation. We disagreed, holding that the Tax Court decision, entered pursuant to the parties' stipulations, did not accomplish an "estoppel by judgment," i.e., it had no issue-preclusive effect:

"We conclude that the decisions entered by the Tax Court for the years 1933, 1938, and 1939 were only a pro forma acceptance by the Tax Court of an agreement between the parties to settle their controversy for reasons undisclosed . _ Perhaps, as the Court of Appeals inferred, the parties did agree on the basis for depreciation. Perhaps the settlement was made for a different reason, for some exigency arising out of the bankruptcy proceeding. As the case reaches us, we are unable to tell whether the agreement of the parties was based on the merits or on some collateral consideration. Certainly the judgments entered are res judicata of the tax claims for the years 1933, 1938, and 1939, whether or not the basis of the agreements on which they rest reached the merits . _ Estoppel by judgment includes matters in a second proceeding which were actually presented and determined in an earlier suit. A judgment entered with the consent of the parties may involve a determination of questions of fact and law by the court. But unless a showing is made that that was the case, the judgment has no greater dignity, so far as collateral estoppel is concerned, than any judgment entered only as a compromise of the parties." Id., at 505_506 (citations omitted).

The State parties, perhaps recognizing the infirmity of their argument as a matter of standard preclusion doctrine, assert that common-law principles of issue preclusion do not apply in the special context of Indian land claims. Instead, they argue, §22 of the Indian Claims Commission Act created a special regime of "statutory preclusion."5 According to the State parties, the payment of a Commission judgment for claims to aboriginal or trust lands automatically and universally extinguishes title to the Indian lands upon which the claim is based and creates a statutory bar to further assertion of claims against either the United States or third parties based on the extinguished title. The State parties point to several decisions of the Ninth Circuit in support of this contention. See Reply Brief for State Parties 17 (citing United States v. Pend Oreille Pub. Util. Dist. No. 1, 926 F.2d 1502 (CA9 1991)); id., at 15 (citing United States v. Dann, 873 F.2d 1189 (CA9 1989)); id., at 11 (citing United States v. Gemmill, 535 F.2d 1145 (CA9 1976)).

We need not decide whether, in the distinctive context of the Indian Claims Commission Act, some consent judgments might bar a tribe from asserting title even in discrete litigation against third parties, for the 1983 settlement of Docket No. 320 plainly could not qualify as such a judgment. Not only was the issue of ownership of the disputed boundary lands not actually litigated and decided in Docket No. 320, but, most notably, the Tribe proceeded on alternative and mutually exclusive theories of recovery. Had the case proceeded to final judgment upon trial, the Tribe might have won damages for a taking, indicating that title was in the United States. Alternatively, however, the Tribe might have obtained damages for trespass, indicating that title remained in the Tribe. The consent judgment embraced all of the Tribe's claims. There was no election by the Tribe of one theory over the other, nor was any such election required to gain approval for the consent judgment. The Special Master's assumption that the settlement necessarily and universally relinquished the Tribe's claim to title was thus unwarranted. Certainly, if the $15 million payment constituted a discharge of the Tribe's trespass claim, it would make scant sense to say that the acceptance of the payment extinguished the Tribe's title. In contrast, the Ninth Circuit cases cited by the State parties (the correctness of which we do not address) all involved Indian Claims Commission Act petitions in which tribes claimed no continuing title, choosing instead to seek compensation from the United States for the taking of their lands. See, e.g., Pend Oreille, 926 F.2d, at 1507_1508; Dann, 873 F.2d, at 1192, 1194; Gemmill, 535 F.2d, at 1149, and n. 6.

The United States invites us to look behind the consent judgment in Docket No. 320 at presettlement stipulations and memoranda purportedly demonstrating that the judgment was grounded on the parties' shared view, after the 1978 Secretarial Order, that the disputed lands belong to the Tribe. We need not accept the Government's invitation. On the matter of issue preclusion, it suffices to observe that the settlement was ambiguous as between mutually exclusive theories of recovery. Like the Tax Court settlement in International Building Co., then, the consent judgment in the Tribe's Claims Court action is too opaque to serve as a foundation for issue preclusion. Accordingly, we hold that the claims of the United States and the Tribe to increased water rights for the disputed boundary lands of the Fort Yuma Reservation are not precluded by the consent judgment in Docket No. 320.

The Special Master has recommended that the Court approve the parties' proposed settlement of the dispute respecting the Fort Mojave Reservation. The claim to additional water for the Fort Mojave Reservation arises out of a dispute over the accuracy of a survey of the so-called Hay and Wood Reserve portion of the Reservation. See Arizona II, 460 U.S., at 631_632. The parties agreed to resolve the matter through an accord that (1) specifies the location of the disputed boundary; (2) preserves the claims of the parties regarding title to and jurisdiction over the bed of the last natural course of the Colorado River within the agreed-upon boundary; (3) awards the Tribe the lesser of an additional 3,022 acre-feet of water or enough water to supply the needs of 468 acres; (4) precludes the United States and the Tribe from claiming additional water rights from the Colorado River for lands within the Hay and Wood Reserve; and (5) disclaims any intent to affect any private claims to title to or jurisdiction over any lands. See McGarr Report 8_9. We accept the Master's uncontested recommendation and approve the proposed settlement.

The Master has also recommended that the Court approve the parties' proposed settlement of the dispute respecting the Colorado River Indian Reservation. The claim to additional water for that reservation stems principally from a dispute over whether the reservation boundary is the ambulatory west bank of the Colorado River or a fixed line representing a past location of the River. See Arizona II, 460 U.S., at 631. The parties agreed to resolve the matter through an accord that (1) awards the Tribes the lesser of an additional 2,100 acre-feet of water or enough water to irrigate 315 acres; (2) precludes the United States or the Tribe from seeking additional reserved water rights from the Colorado River for lands in California; (3) embodies the parties' intent not to adjudicate in these proceedings the correct location of the disputed boundary; (4) preserves the competing claims of the parties to title to or jurisdiction over the bed of the Colorado River within the reservation; and (5) provides that the agreement will become effective only if the Master and the Court approve the settlement. See McGarr Report 9_10. The Master expressed concern that the settlement does not resolve the location of the disputed boundary, but recognized that it did achieve the ultimate aim of determining water rights associated with the disputed boundary lands. Id., at 10_12, 13_14. We again accept the Master's recommendation and approve the proposed settlement.6

* * *

For the foregoing reasons, we remand the outstanding water rights claims associated with the disputed boundary lands of the Fort Yuma Indian Reservation to the Special Master for determination on the merits. Those claims are the only ones that remain to be decided in Arizona v. California; their resolution will enable the Court to enter a final consolidated decree and bring this case to a close.

With respect to the Fort Mojave and Colorado River Reservations, the Special Master has submitted a proposed supplemental decree to carry the parties' accords into effect. That decree is reproduced as the Appendix to this opinion, infra, at 26_27. The parties are directed to submit to the Clerk of this Court, before August 22, 2000, any objections to the proposed supplemental decree.

It is so ordered.

A. Paragraph (4) of Article II(D) of the Decree in this case entered on March 9, 1964 (376 U.S. 340, 344_345) is hereby amended to read as follows:

(4) The Colorado River Indian Reservation in annual quantities not to exceed (i) 719,248 acre-feet of diversions from the mainstream or (ii) the quantity of mainstream water necessary to supply the consumptive use required for irrigation of 107,903 acres and for the satisfaction of related uses, whichever of (i) or (ii) is less, with priority dates of March 3, 1865, for lands reserved by the Act of March 3, 1865 (13 Stat. 541, 559); November 22, 1873, for lands reserved by the Executive Order of said date; November 16, 1874, for lands reserved by the Executive Order of said date, except as later modified; May 15, 1876, for lands reserved by the Executive Order of said date; November 22, 1915, for lands reserved by the Executive Order of said date.

B. Paragraph (5) of Article II(D) of the Decree in this case entered on March 9, 1964 (376 U.S. 340, 345) and supplemented on April 16, 1984 (466 U.S. 144, 145) is hereby amended to read as follows:

(5) The Fort Mojave Indian Reservation in annual quantities not to exceed (i) 132,789 acre-feet of diversions from the mainstream or (ii) the quantity of mainstream water necessary to supply the consumptive use required for irrigation of 20,544 acres and for the satisfaction of related uses, whichever of (i) or (ii) is less, with priority dates of September 19, 1890, for lands transferred by the Executive Order of said date; February 2, 1911, for lands reserved by the Executive Order of said date.

C. Paragraph (5) of the introductory conditions to the Supplemental Decree in this case entered on January 9, 1979 (439 U.S. 419, 421_423) is hereby amended by adding the following exception at the end of the concluding proviso in the first sentence of that paragraph: "except for the western boundaries of the Fort Mojave and Colorado River Indian Reservations in California."

D. Paragraph II(A)(24) of the Decree of January 9, 1979 (439 U.S. 419, 428) is hereby amended to read as follows:

Colorado River Indian Reservation 10,745 1,612 Nov. 22, 1873 40,241 6,037 Nov. 16, 1874

E. Paragraph II(A)(25) of the Decree of January 9, 1979 (439 U.S. 419, 428) is hereby amended to read as follows:

Fort Mojave Indian Reservation 16,720 2,587 Sept. 18, 1890

F. Except as otherwise provided herein, the Decree entered on March 9, 1964, and the Supplemental Decrees entered on January 9, 1979, and April 16, 1984, shall remain in full force and effect.

G. The Court shall retain jurisdiction herein to order such further proceedings and enter such supplemental decree as may be deemed appropriate.

Opinion of Rehnquist, C. J.

Chief Justice Rehnquist, with whom Justice O'Connor and Justice Thomas join, concurring in part and dissenting in part.

I believe that the United States and Quechan Tribe's claim for additional water rights is barred by the principles of res judicata, and therefore I dissent. The Special Master concluded that an exception to the general preclusion rule applied and that, therefore, the United States' claim was not barred. The Court rejects the Special Master's reasoning but concludes that the State Parties' res judicata defense is not properly before the Court. While I agree that the Special Master erred in finding the 1978 Order of the Secretary of the Interior a "new fact" justifying an exception to the application of preclusion, I disagree with the Court's refusal to reach the merits of the State Parties' defense.

The Court first concludes that the State Parties lost the defense because they failed to assert it in a timely manner. While the State Parties concede that they did not raise their claim of res judicata until 1989, it does not automatically follow that the defense is lost. Federal Rule of Civil Procedure 8(c) provides that res judicata shall be pleaded as an affirmative defense. But the only "pleadings" in this case were filed in the 1950's, at which time no claim of res judicata could have been made. The motions filed by the State Parties in 1977 and 1979 were not in any sense comprehensive pleadings, purporting to set forth all of the claims and defenses of the parties. More importantly, neither Special Master Tuttle nor this Court focused on the merits of the boundary dispute during the proceedings in Arizona v. California, 460 U.S. 605 (1983)(Arizona II). Rather, the Master only decided whether the Secretary's order was a final boundary determination, and, similarly, this Court simply determined that the Secretary's order was subject to challenge and encouraged the parties to assert their legal claims and defenses in another forum. Consequently, it is likely that the State Parties' res judicata claim would not have been resolved in Arizona II even if it had been raised.

The State Parties did expressly raise the defense of res judicata in their 1989 motion, and neither the United States nor the Tribe objected to its consideration. The Tribe contested the merits of the State Parties' res judicata claim and argued that its water rights' claim was not precluded. In so doing, the Tribe asserted that the State Parties had not argued res judicata during the Arizona II proceedings. But neither the Tribe nor the United States contended, in response to the State Parties' motion, that the Court could not decide the res judicata issue because it was not timely raised. We granted the motion, and Master McGarr considered the claim on the merits. Under these circumstances, I believe that the State Parties did not lose their res judicata defense by failing to assert it in the earlier proceedings.

The Court also concludes that this Court's 1979 and 1984 supplemental decrees "anticipated" that the boundary dispute would be finally resolved in some forum. See, ante, at 16. To reach this conclusion, the Court reads too much into the simple language of the supplemental decrees and ignores language in our Arizona II opinion. The supplemental decrees stated that water rights for the five reservations "shall be subject to appropriate adjustments by agreement or decree of this Court in the event that the boundaries of the respective reservations are finally determined." 1984 Supplemental Decree, Art. II(D)(5), Arizona v. California, 466 U.S. 144, 145 (1984); 1979 Supplemental Decree, Art. II(D)(5), Arizona v. California, 439 U.S. 419, 421 (1979) (per curiam). These decrees can best be interpreted as merely providing that the reservation's water quantity can be adjusted if the boundary changes, without deciding whether the boundary relied on in the 1964 decree could be properly challenged, and without indicating that the boundary necessarily would be "finally determined" at some future point. This reading is supported by language in Arizona II. In discussing the pending District Court action, we explained: "We note that the United States has moved to dismiss the action filed by the agencies based on lack of standing, the absence of indispensable parties, sovereign immunity, and the applicable statute of limitations. There will be time enough, if any of these grounds for dismissal are sustained and not overturned on appellate review, to determine whether the boundary issues foreclosed by such action are nevertheless open for litigation in this Court." 460 U.S., at 638 (emphasis added; footnote omitted). As is evident from this language, we did not "anticipate" that the dispute would be finally resolved. Instead, we explicitly left open the question whether the dispute could be litigated in this Court.

The Court disregards this language in Arizona II because it does not mention a potential preclusion defense. However, the point is not that this Court anticipated the State Parties' preclusion defense. Rather, it is that this Court recognized the possibility that the boundary issue would not be judicially resolved at all, and left open the question whether there was some defense precluding this Court's review. What that defense might be was not before the Court.

Now that the question is squarely before us, I would hold that the United States' claim for additional water rights is barred by the principles of res judicata. Res judicata not only bars relitigation of claims previously litigated, but also precludes claims that could have been brought in earlier proceedings. Under the doctrine of res judicata, "when a final judgment has been entered on the merits of a case, `[i]t is a finality as to the claim or demand in controversy, concluding parties and those in privity with them, not only as to every matter which was offered and received to sustain or defeat the claim or demand, but as to any other admissible matter which might have been offered for that purpose." Nevada v. United States, 463 U.S. 110, 129_130 (1983) (quoting Cromwell v. County of Sac, 94 U.S. 351, 352 (1877)).

In Arizona II, we recognized that the general principles of res judicata apply to our 1964 decree even though the decree expressly provided for modification in appropriate circumstances. In so doing, we noted the importance of the certainty of water rights in the Western United States. "A major purpose of this litigation, from its inception to the present day, has been to provide the necessary assurance to States of the Southwest and to various private interests, of the amount of water they can anticipate to receive from the Colorado River system. . . . If there is no surplus of water in the Colorado River, an increase in federal reserved water rights will require a `gallon-for-gallon reduction in the amount of water available for water-needy state and private appropriators.'" 460 U.S., at 620_621 (quoting United States v. New Mexico, 438 U.S. 696, 699 (1978)). Thus, we concluded that allowing recalculation of the amount of practicably irrigable acreage "runs directly counter to the strong interest in finality in this case." Id., at 620. We also noted that treating the 1964 calculation as final comported with the clearly expressed intention of the parties and was consistent with our previous treatment of original actions, allowing modifications after a change in the relevant circumstances.

This reasoning is equally applicable to the United States and the Tribe's claim for additional water for the disputed boundary lands. Even though the exact claim was not actually litigated in Arizona v. California, 373 U.S. 546 (1963) (Arizona I), the United States could have raised the boundary claim and failed to do so. Indeed, in the proceedings before Special Master Rifkind, the counsel for the United States affirmatively represented that "[t]he testimony . . . as reflected by these maps and by the other testimony will define the maximum claim which the United States is asserting in this case." Earlier in the proceedings, the Master explicitly warned the United States about the preclusive effect of failing to assert potential claims: "In an action or a decree quieting title, you cut out all claims not asserted. . . . I just want you to be aware of the fact that the mere fact that it has not been asserted does not mean that you may not lose it . . . ." Exception by State Parties to Report of Special Master and Supporting Brief 8_9 (colloquy between counsel for the United States and the Special Master). Thus, under the general principles of res judicata, the United States would clearly be barred from now asserting the claim for additional water rights.

Master McGarr concluded that the United States' claim was not precluded because it fell within an exception to the bar of res judicata. Wisely abandoning the Master's reasoning, the United States instead defends the Master's ruling on the ground that these claims "are not precluded, under basic principles of res judicata, because [they] were not decided, and could not have been decided, in the prior proceedings." Reply Brief for United States in Response to Exception of the State Parties 21. But this argument fares no better.

The issue before the Master in Arizona I was the amount of water from the Colorado River to which the Quechan Tribe was entitled. The Master made an allotment to the reservation based on the evidence then before him as to the amount of irrigable acreage within the reservation boundary, which was undisputed at the time. Only years after that decree was confirmed by this Court in Arizona I did the United States assert a larger claim to water for the reservation based on a claim for a larger amount of irrigable acreage-not because of a miscalculation as to the irrigability of acreage already claimed, but because of a claimed extension of the boundaries of the reservation. But, at the time of Arizona I, the United States had in its possession all of the facts that it later asserted in 1979 in Arizona II, and it could have litigated the larger claim before Master Rifkind.

The United States offers no support for its contention that the boundary dispute could not have been decided in Arizona I except for the fact that this Court rejected the Master's resolution of the Fort Mojave Reservation and Colorado River Reservation boundary disputes. However, those boundary disputes are different. While we did not explain in Arizona I why we believed it was improper to decide the boundary disputes, California's objection was based on the fact that necessary parties were not participating in the proceedings. Specifically, California argued that it lacked the authority to represent private individuals claiming title to the disputed lands and maintained that "it would be unfair to prejudice any of the parties in future litigation over land titles or political jurisdiction by approving findings on a tangential issue never pleaded by the United States." Arizona II, supra, at 629. The Fort Yuma Reservation boundary dispute, on the other hand, is solely between the United States and the Quechan Tribe-there are no private parties claiming title to the land. Thus, the United States could have raised this claim in Arizona I, and the Master could have decided it.

Because I believe that the State Parties' res judicata defense is properly before the Court and that the United States' claim for additional water rights is precluded, I see no need to remand for further proceedings. I agree with the Court that we should approve the proposed settlements of the remaining claims in this case and direct the parties to submit any objections to the proposed supplemental decree.

SUPPLEMENTAL DECREE ORDERED, ADJUDGED, AND DECREED

STATE OF ARIZONA, COMPLAINANT
v.
STATE OF CALIFORNIA et al.

The Special Master has submitted a proposed supplemental decree in this case to carry the parties' accords into effect. The proposed decree was reproduced as an appendix to the Court's opinion dated June 19, 2000, and any objections were called for. No objections were filed with the Clerk. Accordingly, the proposed supplemental decree with respect to the Fort Mojave and
Colorado River Reservations is approved and entered.

A. Paragraph (4) of Article II(D) of the Decree in this case entered on March 9, 1964 (376 U.S. 340, 344_345) is hereby amended to read as follows:

(4) The Colorado River Indian Reservation in annual quantities not to exceed (i) 719,248 acre-feet of diversions from the mainstream or (ii) the quantity of mainstream water necessary to supply the consumptive use required for irrigation of 107,903 acres and for the satisfaction of related uses, whichever of (i) or (ii) is less, with priority dates of March 3, 1865, for lands reserved by the Act of March 3, 1865 (13 Stat. 541, 559); November 22, 1873, for lands reserved by the Executive Order of said date; November 16, 1874, for lands reserved by the Executive Order of said date, except as later modified; May 15, 1876, for lands reserved by the Executive Order of said date; November 22, 1915, for lands reserved by the Executive Order of said date.

B. Paragraph (5) of Article II(D) of the Decree in this case entered on March 9, 1964 (376 U.S. 340, 345) and supplemented on April 16, 1984 (466 U.S. 144, 145) is hereby amended to read as follows:

(5) The Fort Mojave Indian Reservation in annual quantities not to exceed (i) 132,789 acre-feet of diversions from the mainstream or (ii) the quantity of mainstream water necessary to supply the consumptive use required for irrigation of 20,544 acres and for the satisfaction of related uses, whichever of (i) or (ii) is less, with priority dates of September 19, 1890, for lands transferred by the Executive Order of said date; February 2, 1911, for lands reserved by the Executive Order of said date.

C. Paragraph (5) of the introductory conditions to the Supplemental Decree in this case entered on January 9, 1979 (439 U.S. 419, 421_423) is hereby amended by adding the following exception at the end of the concluding proviso in the first sentence of that paragraph: "except for the western boundaries of the Fort Mojave and Colorado River Indian Reservations in California."

D. Paragraph II(A)(24) of the Decree of January 9, 1979 (439 U.S. 419, 428) is hereby amended to read as follows:

Colorado River Indian Reservation 10,745 1,612 Nov. 22, 1873

E. Paragraph II(A)(25) of the Decree of January 9, 1979 (439 U.S. 419, 428) is hereby amended to read as follows:

Fort Mojave Indian Reservation 16,720 2,587 Sept. 18, 1890

F. Except as otherwise provided herein, the Decree entered on March 9, 1964, and the Supplemental Decrees entered on January 9, 1979, and April 16, 1984, shall remain in full force and effect.

G. The Court shall retain jurisdiction herein to order such further proceedings and enter such supplemental decree as may be deemed appropriate

This case presents the question whether the federal mail fraud statute, 18 U.S.C. § 1341 reaches false statements made in an application for a state license. Section 1341 proscribes use of the mails in furtherance of "any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises." Petitioner Carl W. Cleveland and others were prosecuted under this federal measure for making false statements in applying to the Louisiana State Police for permission to operate video poker machines. We conclude that permits or licenses of this order do not qualify as "property" within §1341's compass. It does not suffice, we clarify, that the object of the fraud may become property in the recipient's hands; for purposes of the mail fraud statute, the thing obtained must be property in the hands of the victim. State and municipal licenses in general, and Louisiana's video poker licenses in particular, we hold, do not rank as "property," for purposes of §1341, in the hands of the official licensor.

* Louisiana law allows certain businesses to operate video poker machines. La. Rev. Stat. Ann. §§27:301 to 27:324 (West Supp. 2000). The State itself, however, does not run such machinery. The law requires prospective owners of video poker machines to apply for a license from the State. §27:306. The licenses are not transferable, §27:311(G), and must be renewed annually, La. Admin. Code, tit. 42, §2405(B)(3) (2000). To qualify for a license, an applicant must meet suitability requirements designed to ensure that licensees have good character and fiscal integrity. La. Rev. Stat. Ann. §27:310 (West Supp. 2000).

In 1992, Fred Goodson and his family formed a limited partnership, Truck Stop Gaming, Ltd. (TSG), in order to participate in the video poker business at their truck stop in Slidell, Louisiana. Cleveland, a New Orleans lawyer, assisted Goodson in preparing TSG's application for a video poker license. The application required TSG to identify its partners and to submit personal financial statements for all partners. It also required TSG to affirm that the listed partners were the sole beneficial owners of the business and that no partner held an interest in the partnership merely as an agent or nominee, or intended to transfer the interest in the future.

TSG's application identified Goodson's adult children, Alex and Maria, as the sole beneficial owners of the partnership. It also showed that Goodson and Cleveland's law firm had loaned Alex and Maria all initial capital for the partnership and that Goodson was TSG's general manager. In May 1992, the State approved the application and issued a license. TSG successfully renewed the license in 1993, 1994, and 1995 pursuant to La. Admin. Code, tit. 42, §2405(B)(3) (2000). Each renewal application identified no ownership interests other than those of Alex and Maria.

In 1996, the FBI discovered evidence that Cleveland and Goodson had participated in a scheme to bribe state legislators to vote in a manner favorable to the video poker industry. The Government charged Cleveland and Goodson with multiple counts of money laundering under 18 U.S.C. § 1957 as well as racketeering and conspiracy under §1962. Among the predicate acts supporting these charges were four counts of mail fraud under §1341.1 The indictment alleged that Cleveland and Goodson had violated §1341 by fraudulently concealing that they were the true owners of TSG in the initial license application and three renewal applications mailed to the State. They concealed their ownership interests, according to the Government, because they had tax and financial problems that could have undermined their suitability to receive a video poker license. See La. Rev. Stat. Ann. §27:310(B)(1) (suitability requirements).

Before trial, Cleveland moved to dismiss the mail fraud counts on the ground that the alleged fraud did not deprive the State of "property" under §1341. The District Court denied the motion, concluding that "licenses constitute property even before they are issued." 951 F. Supp. 1249, 1261 (ED La. 1997). A jury found Cleveland guilty on two counts of mail fraud (based on the 1994 and 1995 license renewals) and on money laundering, racketeering, and conspiracy counts predicated on the mail fraud. The District Court sentenced Cleveland to 121 months in prison.

On appeal, Cleveland again argued that Louisiana had no property interest in video poker licenses, relying on several Court of Appeals decisions holding that the government does not relinquish "property" for purposes of §1341 when it issues a permit or license. See United States v. Shotts, 145 F.3d 1289, 1296 (CA11 1998) (license to operate a bail bonds business); United States v. Schwartz, 924 F.2d 410, 418 (CA2 1991) (arms export license); United States v. Granberry, 908 F.2d 278, 280 (CA8 1990) (school bus operator's permit); Toulabi v. United States, 875 F.2d 122, 125 (CA7 1989) (chauffeur's license); United States v. Dadanian, 856 F.2d 1391, 1392 (CA9 1988) (gambling license); United States v. Murphy, 836 F.2d 248, 254 (CA6 1988) (license to conduct charitable bingo games).

The Court of Appeals for the Fifth Circuit nevertheless affirmed Cleveland's conviction and sentence, United States v. Bankston, 182 F.3d 296, 309 (1999), considering itself bound by its holding in United States v. Salvatore, 110 F.3d 1131, 1138 (1997), that Louisiana video poker licenses constitute "property" in the hands of the State. Two other Circuits have concluded that the issuing authority has a property interest in unissued licenses under §1341. United States v. Bucuvalas, 970 F.2d 937, 945 (CA1 1992) (entertainment and liquor license); United States v. Martinez, 905 F.2d 709, 715 (CA3 1990) (medical license).

We granted certiorari to resolve the conflict among the Courts of Appeals, 529 U.S. 1017 (2000), and now reverse the Fifth Circuit's judgment.

In McNally v. United States, 483 U.S. 350, 360 (1987), this Court held that the federal mail fraud statute is "limited in scope to the protection of property rights." McNally reversed the mail fraud convictions of two individuals charged with participating in "a self-dealing patronage scheme" that defrauded Kentucky citizens of "the right to have the Commonwealth's affairs conducted honestly." Id., at 352. At the time McNally was decided, federal prosecutors had been using §1341 to attack various forms of corruption that deprived victims of "intangible rights" unrelated to money or property.2 Reviewing the history of §1341, we concluded that "the original impetus behind the mail fraud statute was to protect the people from schemes to deprive them of their money or property." Id., at 356.

As first enacted in 1872, §1341 proscribed use of the mails to further " 'any scheme or artifice to defraud.' " Ibid. In 1896, this Court held in Durland v. United States, 161 U.S. 306, 313, that the statute covered fraud not only by "representations as to the past or present," but also by "suggestions and promises as to the future." In 1909, Congress amended §1341 to add after "any scheme or artifice to defraud" the phrase "or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises." McNally, 483 U.S., at 357. We explained in McNally that the 1909 amendment "codified the holding of Durland," id., at 357, and "simply made it unmistakable that the statute reached false promises and misrepresentations as to the future as well as other frauds involving money or property," id., at 359. Rejecting the argument that "the money-or-property requirement of the latter phrase does not limit schemes to defraud to those aimed at causing deprivation of money or property," id., at 358, we concluded that the 1909 amendment signaled no intent by Congress to "depar[t] from [the] common understanding" that "the words 'to defraud' commonly refer 'to wronging one in his property rights,' " id., at 358 359 (quoting Hammerschmidt v. United States, 265 U.S. 182, 188 (1924)).

Soon after McNally, in Carpenter v. United States, 484 U.S. 19, 25 (1987), we again stated that §1341 protects property rights only. Carpenter upheld convictions under §1341 and the federal wire fraud statute, 18 U.S.C. § 1343 of defendants who had defrauded the Wall Street Journal of confidential business information. Citing decisions of this Court as well as a corporate law treatise, we observed that "[c]onfidential business information has long been recognized as property." 484 U.S., at 26.

The following year, Congress amended the law specifically to cover one of the "intangible rights" that lower courts had protected under §1341 prior to McNally: "the intangible right of honest services." Anti-Drug Abuse Act of 1988, §7603(a), 18 U.S.C. § 1346. Significantly, Congress covered only the intangible right of honest services even though federal courts, relying on McNally, had dismissed, for want of monetary loss to any victim, prosecutions under §1341 for diverse forms of public corruption, including licensing fraud.3

In this case, there is no assertion that Louisiana's video poker licensing scheme implicates the intangible right of honest services. The question presented is whether, for purposes of the federal mail fraud statute, a government regulator parts with "property" when it issues a license. For the reasons we now set out, we hold that §1341 does not reach fraud in obtaining a state or municipal license of the kind here involved, for such a license is not "property" in the government regulator's hands. Again, as we said in McNally, "[i]f Congress desires to go further, it must speak more clearly than it has." 483 U.S., at 360.

To begin with, we think it beyond genuine dispute that whatever interests Louisiana might be said to have in its video poker licenses, the State's core concern is regulatory. Louisiana recognizes the importance of "public confidence and trust that gaming activities are conducted honestly and are free from criminal and corruptive elements." La. Rev. Stat. Ann. §27:306(A)(1) (West Supp. 2000). The video poker licensing statute accordingly asserts the State's "legitimate interest in providing strict regulation of all persons, practices, associations, and activities related to the operation of establishments licensed to offer video draw poker devices." Ibid. The statute assigns the Office of State Police, a part of the Department of Public Safety and Corrections, the responsibility to promulgate rules and regulations concerning the licensing process. §27:308(A). It also authorizes the State Police to deny, condition, suspend, or revoke licenses, to levy fines of up to $1,000 per violation of any rule, and to inspect all premises where video poker devices are offered for play. §§27:308(B), (E)(1). In addition, the statute defines criminal penalties for unauthorized use of video poker devices, §27:309, and prescribes detailed suitability requirements for licensees, §27:310.

In short, the statute establishes a typical regulatory program. It licenses, subject to certain conditions, engagement in pursuits that private actors may not undertake without official authorization. In this regard, it resembles other licensing schemes long characterized by this Court as exercises of state police powers. E.g., Ziffrin, Inc. v. Reeves, 308 U.S. 132, 138 (1939) (license to transport alcoholic beverages); Hall v. Geiger-Jones Co., 242 U.S. 539, 558 (1917) (license to sell corporate stock); Fanning v. Gregoire, 16 How. 524, 534 (1854) (ferry license); License Cases, 5 How. 504, 589 (1847) (license to sell liquor) (opinion of McLean, J.), overruled on other grounds, Leisy v. Hardin, 135 U.S. 100 (1890).

Acknowledging Louisiana's regulatory interests, the Government offers two reasons why the State also has a property interest in its video poker licenses. First, the State receives a substantial sum of money in exchange for each license and continues to receive payments from the licensee as long as the license remains in effect. Second, the State has significant control over the issuance, renewal, suspension, and revocation of licenses.

Without doubt, Louisiana has a substantial economic stake in the video poker industry. The State collects an upfront "processing fee" for each new license application, La. Rev. Stat. Ann. §27:311(H)(2) (West Supp. 2000) ($10,000 for truck stops), a separate "processing fee" for each renewal application, §27:311(H)(4) ($1,000 for truck stops), an "annual fee" from each device owner, §27:311(A)(4) ($2,000), an additional "device operation" fee, §27:311(A)(5)(c) ($1,000 for truck stops), and, most importantly, a fixed percentage of net revenue from each video poker device, §27:311(D)(1)(b) (32.5% for truck stops). It is hardly evident, however, why these tolls should make video poker licenses "property" in the hands of the State. The State receives the lion's share of its expected revenue not while the licenses remain in its own hands, but only after they have been issued to licensees. Licenses pre-issuance do not generate an ongoing stream of revenue. At most, they entitle the State to collect a processing fee from applicants for new licenses. Were an entitlement of this order sufficient to establish a state property right, one could scarcely avoid the conclusion that States have property rights in any license or permit requiring an upfront fee, including drivers' licenses, medical licenses, and fishing and hunting licenses. Such licenses, as the Government itself concedes, are "purely regulatory." Tr. of Oral Arg. 24 25.

Tellingly, as to the character of Louisiana's stake in its video poker licenses, the Government nowhere alleges that Cleveland defrauded the State of any money to which the State was entitled by law. Indeed, there is no dispute that TSG paid the State of Louisiana its proper share of revenue, which totaled more than $1.2 million, between 1993 and 1995. If Cleveland defrauded the State of "property," the nature of that property cannot be economic.

Addressing this concern, the Government argues that Cleveland frustrated the State's right to control the issuance, renewal, and revocation of video poker licenses under La. Rev. Stat. Ann. §§27:306, 27:308 (West Supp. 2000). The Fifth Circuit has characterized the protected interest as "Louisiana's right to choose the persons to whom it issues video poker licenses." Salvatore, 110 F.3d, at 1140. But far from composing an interest that "has long been recognized as property," Carpenter, 484 U.S., at 26, these intangible rights of allocation, exclusion, and control amount to no more and no less than Louisiana's sovereign power to regulate. Notably, the Government overlooks the fact that these rights include the distinctively sovereign authority to impose criminal penalties for violations of the licensing scheme, La. Rev. Stat. Ann. §27:309 (West Supp. 2000), including making false statements in a license application, §27:309(A). Even when tied to an expected stream of revenue, the State's right of control does not create a property interest any more than a law licensing liquor sales in a State that levies a sales tax on liquor. Such regulations are paradigmatic exercises of the States' traditional police powers.

The Government compares the State's interest in video poker licenses to a patent holder's interest in a patent that she has not yet licensed. Although it is true that both involve the right to exclude, we think the congruence ends there. A patent not only confers the right to exclude others from using an invention, it also protects the holder's right to use, make, or sell the invention herself. 35 U.S.C. § 154 271(d)(1). Louisiana does not conduct gaming operations itself, it does not hold video poker licenses to reserve that prerogative, and it does not "sell" video poker licenses in the ordinary commercial sense. Furthermore, while a patent holder may sell her patent, see §261 ("patents shall have the attributes of personal property"), the State may not sell its licensing authority. Instead of a patent holder's interest in an unlicensed patent, the better analogy is to the Federal Government's interest in an unissued patent. That interest, like the State's interest in licensing video poker operations, surely implicates the Government's role as sovereign, not as property holder. See U.S. Const., Art. I, §8, cl. 8.

The Government also compares the State's licensing power to a franchisor's right to select its franchisees. On this view, Louisiana's video poker licensing scheme represents the State's venture into the video poker business. Although the State could have chosen to run the business itself, the Government says, it decided to franchise private entities to carry out the operations instead. However, a franchisor's right to select its franchisees typically derives from its ownership of a trademark, brand name, business strategy, or other product that it may trade or sell in the open market. Louisiana's authority to select video poker licensees rests on no similar asset. It rests instead upon the State's sovereign right to exclude applicants deemed unsuitable to run video poker operations. A right to exclude in that governing capacity is not one appropriately labeled "property." See Tr. of Oral Arg. 25. Moreover, unlike an entrepreneur or business partner who shares both losses and gains arising from a business venture, Louisiana cannot be said to have put its labor or capital at risk through its fee-laden licensing scheme. In short, the State did not decide to venture into the video poker business; it decided typically to permit, regulate, and tax private operators of the games.

We reject the Government's theories of property rights not simply because they stray from traditional concepts of property. We resist the Government's reading of §1341 as well because it invites us to approve a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress. Equating issuance of licenses or permits with deprivation of property would subject to federal mail fraud prosecution a wide range of conduct traditionally regulated by state and local authorities. We note in this regard that Louisiana's video poker statute typically and unambiguously imposes criminal penalties for making false statements on license applications. La. Rev. Stat. Ann. §27:309(A) (West Supp. 2000). As we reiterated last Term, " 'unless Congress conveys its purpose clearly, it will not be deemed to have significantly changed the federal-state balance' in the prosecution of crimes." Jones v. United States, 529 U.S. 848, 858 (2000) (quoting United States v. Bass, 404 U.S. 336, 349 (1971)).

Moreover, to the extent that the word "property" is ambiguous as placed in §1341, we have instructed that "ambiguity concerning the ambit of criminal statutes should be resolved in favor of lenity." Rewis v. United States, 401 U.S. 808, 812 (1971). This interpretive guide is especially appropriate in construing §1341 because, as this case demonstrates, mail fraud is a predicate offense under RICO, 18 U.S.C. § 1961(1) (1994 ed., Supp. IV), and the money laundering statute, §1956(c)(7)(A). In deciding what is "property" under §1341, we think "it is appropriate, before we choose the harsher alternative, to require that Congress should have spoken in language that is clear and definite." United States v. Universal C. I. T. Credit Corp., 344 U.S. 218, 222 (1952).

Finally, in an argument not raised below but urged as an alternate ground for affirmance, the Government contends that §1341, as amended in 1909, defines two independent offenses: (1) "any scheme or artifice to defraud" and (2) "any scheme or artifice for obtaining money or property by means of false or fraudulent pretenses, representations, or promises." Because a video poker license is property in the hands of the licensee, the Government says, Cleveland "obtain[ed] property" and thereby committed the second offense even if the license is not property in the hands of the State.

Although we do not here question that video poker licensees may have property interests in their licenses,4 we nevertheless disagree with the Government's reading of §1341. In McNally, we recognized that "[b]ecause the two phrases identifying the proscribed schemes appear in the disjunctive, it is arguable that they are to be construed independently." 483 U.S., at 358. But we rejected that construction of the statute, instead concluding that the second phrase simply modifies the first by "ma[king] it unmistakable that the statute reached false promises and misrepresentations as to the future as well as other frauds involving money or property." Id., at 359. Indeed, directly contradicting the Government's view, we said that "the mail fraud statute had its origin in the desire to protect individual property rights, and any benefit which the Government derives from the statute must be limited to the Government's interests as property holder." Id., at 359, n. 8 (emphasis added). We reaffirm our reading of §1341 in McNally. See Hilton v. South Carolina Public Railways Comm'n, 502 U.S. 197, 205 (1991) ("stare decisis is most compelling" where "a pure question of statutory construction" is involved). Were the Government correct that the second phrase of §1341 defines a separate offense, the statute would appear to arm federal prosecutors with power to police false statements in an enormous range of submissions to state and local authorities. For reasons already stated, see supra, at 11 12, we decline to attribute to §1341 a purpose so encompassing where Congress has not made such a design clear.

We conclude that §1341 requires the object of the fraud to be "property" in the victim's hands and that a Louisiana video poker license in the State's hands is not "property" under §1341. Absent clear statement by Congress, we will not read the mail fraud statute to place under federal superintendence a vast array of conduct traditionally policed by the States. Our holding means that Cleveland's §1341 conviction must be vacated. Accordingly, the judgment of the United States Court of Appeals for the Fifth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Section 404(a) of the Clean Water Act (CWA or Act), 86 Stat. 884, as amended, 33 U.S.C. § 1344(a), regulates the discharge of dredged or fill material into "navigable waters." The United States Army Corps of Engineers (Corps), has interpreted §404(a) to confer federal authority over an abandoned sand and gravel pit in northern Illinois which provides habitat for migratory birds. We are asked to decide whether the provisions of §404(a) may be fairly extended to these waters, and, if so, whether Congress could exercise such authority consistent with the Commerce Clause, U.S. Const., Art. I, §8, cl. 3. We answer the first question in the negative and therefore do not reach the second.

Petitioner, the Solid Waste Agency of Northern Cook County (SWANCC), is a consortium of 23 suburban Chicago cities and villages that united in an effort to locate and develop a disposal site for baled nonhazardous solid waste. The Chicago Gravel Company informed the municipalities of the availability of a 533-acre parcel, bestriding the Illinois counties Cook and Kane, which had been the site of a sand and gravel pit mining operation for three decades up until about 1960. Long since abandoned, the old mining site eventually gave way to a successional stage forest, with its remnant excavation trenches evolving into a scattering of permanent and seasonal ponds of varying size (from under one-tenth of an acre to several acres) and depth (from several inches to several feet).

The municipalities decided to purchase the site for disposal of their baled nonhazardous solid waste. By law, SWANCC was required to file for various permits from Cook County and the State of Illinois before it could begin operation of its balefill project. In addition, because the operation called for the filling of some of the permanent and seasonal ponds, SWANCC contacted federal respondents (hereinafter respondents), including the Corps, to determine if a federal landfill permit was required under §404(a) of the CWA, 33 U.S.C. § 1344(a).

Section 404(a) grants the Corps authority to issue permits "for the discharge of dredged or fill material into the navigable waters at specified disposal sites." Ibid. The term "navigable waters" is defined under the Act as "the waters of the United States, including the territorial seas." §1362(7). The Corps has issued regulations defining the term "waters of the United States" to include

"waters such as intrastate lakes, rivers, streams (including intermittent streams), mudflats, sandflats, wetlands, sloughs, prairie potholes, wet meadows, playa lakes, or natural ponds, the use, degradation or destruction of which could affect interstate or foreign commerce _ ." 33 CFR § 328.3(a)(3) (1999).

In 1986, in an attempt to "clarify" the reach of its jurisdiction, the Corps stated that §404(a) extends to instrastate waters:

"a. Which are or would be used as habitat by birds protected by Migratory Bird Treaties; or

"b. Which are or would be used as habitat by other migratory birds which cross state lines; or

"c. Which are or would be used as habitat for endangered species; or

"d. Used to irrigate crops sold in interstate commerce." 51 Fed. Reg. 41217.

This last promulgation has been dubbed the "Migratory Bird Rule."1

The Corps initially concluded that it had no jurisdiction over the site because it contained no "wetlands," or areas which support "vegetation typically adapted for life in saturated soil conditions," 33 CFR § 328.3(b) (1999). However, after the Illinois Nature Preserves Commission informed the Corps that a number of migratory bird species had been observed at the site, the Corps reconsidered and ultimately asserted jurisdiction over the balefill site pursuant to subpart (b) of the "Migratory Bird Rule." The Corps found that approximately 121 bird species had been observed at the site, including several known to depend upon aquatic environments for a significant portion of their life requirements. Thus, on November 16, 1987, the Corps formally "determined that the seasonally ponded, abandoned gravel mining depressions located on the project site, while not wetlands, did qualify as `waters of the United States' _ based upon the following criteria: (1) the proposed site had been abandoned as a gravel mining operation; (2) the water areas and spoil piles had developed a natural character; and (3) the water areas are used as habitat by migratory bird [sic] which cross state lines." U.S. Army Corps of Engineers, Chicago District, Dept. of Army Permit Evaluation and Decision Document, Lodging of Petitioner, Tab No. 1, p. 6.

During the application process, SWANCC made several proposals to mitigate the likely displacement of the migratory birds and to preserve a great blue heron rookery located on the site. Its balefill project ultimately received the necessary local and state approval. By 1993, SWANCC had received a special use planned development permit from the Cook County Board of Appeals, a landfill development permit from the Illinois Environmental Protection Agency, and approval from the Illinois Department of Conservation.

Despite SWANCC's securing the required water quality certification from the Illinois Environmental Protection Agency, the Corps refused to issue a §404(a) permit. The Corps found that SWANCC had not established that its proposal was the "least environmentally damaging, most practicable alternative" for disposal of nonhazardous solid waste; that SWANCC's failure to set aside sufficient funds to remediate leaks posed an "unacceptable risk to the public's drinking water supply"; and that the impact of the project upon area-sensitive species was "unmitigatable since a landfill surface cannot be redeveloped into a forested habitat." Id., at 87.

Petitioner filed suit under the Administrative Procedure Act, 5 U.S.C. § 701 et seq., in the Northern District of Illinois challenging both the Corps' jurisdiction over the site and the merits of its denial of the §404(a) permit. The District Court granted summary judgment to respondents on the jurisdictional issue, and petitioner abandoned its challenge to the Corps' permit decision. On appeal to the Court of Appeals for the Seventh Circuit, petitioner renewed its attack on respondents' use of the "Migratory Bird Rule" to assert jurisdiction over the site. Petitioner argued that respondents had exceeded their statutory authority in interpreting the CWA to cover nonnavigable, isolated, intrastate waters based upon the presence of migratory birds and, in the alternative, that Congress lacked the power under the Commerce Clause to grant such regulatory jurisdiction.

The Court of Appeals began its analysis with the constitutional question, holding that Congress has the authority to regulate such waters based upon "the cumulative impact doctrine, under which a single activity that itself has no discernible effect on interstate commerce may still be regulated if the aggregate effect of that class of activity has a substantial impact on interstate commerce." 191 F.3d 845, 850 (CA7 1999). The aggregate effect of the "destruction of the natural habitat of migratory birds" on interstate commerce, the court held, was substantial because each year millions of Americans cross state lines and spend over a billion dollars to hunt and observe migratory birds.2 Ibid. The Court of Appeals then turned to the regulatory question. The court held that the CWA reaches as many waters as the Commerce Clause allows and, given its earlier Commerce Clause ruling, it therefore followed that respondents' "Migratory Bird Rule" was a reasonable interpretation of the Act. See id., at 851_852.

We granted certiorari, 529 U.S. 1129 (2000), and now reverse.

Congress passed the CWA for the stated purpose of "restor[ing] and maintain[ing] the chemical, physical, and biological integrity of the Nation's waters." 33 U.S.C. § 1251(a). In so doing, Congress chose to "recognize, preserve, and protect the primary responsibilities and rights of States to prevent, reduce, and eliminate pollution, to plan the development and use (including restoration, preservation, and enhancement) of land and water resources, and to consult with the Administrator in the exercise of his authority under this chapter." §1251(b). Relevant here, §404(a) authorizes respondents to regulate the discharge of fill material into "navigable waters," 33 U.S.C. § 1344(a), which the statute defines as "the waters of the United States, including the territorial seas," §1362(7). Respondents have interpreted these words to cover the abandoned gravel pit at issue here because it is used as habitat for migratory birds. We conclude that the "Migratory Bird Rule" is not fairly supported by the CWA.

This is not the first time we have been called upon to evaluate the meaning of §404(a). In United States v. Riverside Bayview Homes, Inc., 474 U.S. 121 (1985), we held that the Corps had §404(a) jurisdiction over wetlands that actually abutted on a navigable waterway. In so doing, we noted that the term "navigable" is of "limited import" and that Congress evidenced its intent to "regulate at least some waters that would not be deemed `navigable' under the classical understanding of that term." Id., at 133. But our holding was based in large measure upon Congress' unequivocal acquiescence to, and approval of, the Corps' regulations interpreting the CWA to cover wetlands adjacent to navigable waters. See id., at 135_139. We found that Congress' concern for the protection of water quality and aquatic ecosystems indicated its intent to regulate wetlands "inseparably bound up with the `waters' of the United States." Id., at 134.

It was the significant nexus between the wetlands and "navigable waters" that informed our reading of the CWA in Riverside Bayview Homes. Indeed, we did not "express any opinion" on the "question of the authority of the Corps to regulate discharges of fill material into wetlands that are not adjacent to bodies of open water _ ." Id., at 131_132, n. 8. In order to rule for respondents here, we would have to hold that the jurisdiction of the Corps extends to ponds that are not adjacent to open water. But we conclude that the text of the statute will not allow this.

Indeed, the Corps' original interpretation of the CWA, promulgated two years after its enactment, is inconsistent with that which it espouses here. Its 1974 regulations defined §404(a)'s "navigable waters" to mean "those waters of the United States which are subject to the ebb and flow of the tide, and/or are presently, or have been in the past, or may be in the future susceptible for use for purposes of interstate or foreign commerce." 33 CFR § 209.120(d)(1). The Corps emphasized that "[i]t is the water body's capability of use by the public for purposes of transportation or commerce which is the determinative factor." §209.260(e)(1). Respondents put forward no persuasive evidence that the Corps mistook Congress' intent in 1974.3

Respondents next contend that whatever its original aim in 1972, Congress charted a new course five years later when it approved the more expansive definition of "navigable waters" found in the Corps' 1977 regulations. In July 1977, the Corps formally adopted 33 CFR § 323.2(a)(5) (1978), which defined "waters of the United States" to include "isolated wetlands and lakes, intermittent streams, prairie potholes, and other waters that are not part of a tributary system to interstate waters or to navigable waters of the United States, the degradation or destruction of which could affect interstate commerce." Respondents argue that Congress was aware of this more expansive interpretation during its 1977 amendments to the CWA. Specifically, respondents point to a failed House bill, H. R. 3199, that would have defined "navigable waters" as "all waters which are presently used, or are susceptible to use in their natural condition or by reasonable improvement as a means to transport interstate or foreign commerce." 123 Cong. Rec. 10420, 10434 (1977).4 They also point to the passage in §404(g)(1) that authorizes a State to apply to the Environmental Protection Agency for permission "to administer its own individual and general permit program for the discharge of dredged or fill material into the navigable waters (other than those waters which are presently used, or are susceptible to use in their natural condition or by reasonable improvement as a means to transport interstate or foreign commerce _ including wetlands adjacent thereto) within its jurisdiction _ ." 33 U.S.C. § 1344(g)(1). The failure to pass legislation that would have overturned the Corps' 1977 regulations and the extension of jurisdiction in §404(g) to waters "other than" traditional "navigable waters," respondents submit, indicate that Congress recognized and accepted a broad definition of "navigable waters" that includes nonnavigable, isolated, intrastate waters.

Although we have recognized congressional acquiescence to administrative interpretations of a statute in some situations, we have done so with extreme care.5 "[F]ailed legislative proposals are `a particularly dangerous ground on which to rest an interpretation of a prior statute.' " Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U.S. 164, 187 (1994) (quoting Pension Benefit Guaranty Corporation v. LTV Corp., 496 U.S. 633, 650 (1990)). A bill can be proposed for any number of reasons, and it can be rejected for just as many others. The relationship between the actions and inactions of the 95th Congress and the intent of the 92d Congress in passing §404(a) is also considerably attenuated. Because "subsequent history is less illuminating than the contemporaneous evidence," Hagen v. Utah, 510 U.S. 399, 420 (1994), respondents face a difficult task in overcoming the plain text and import of §404(a).

We conclude that respondents have failed to make the necessary showing that the failure of the 1977 House bill demonstrates Congress' acquiescence to the Corps' regulations or the "Migratory Bird Rule," which, of course, did not first appear until 1986. Although respondents cite some legislative history showing Congress' recognition of the Corps' assertion of jurisdiction over "isolated waters,"6 as we explained in Riverside Bayview Homes, "[i]n both Chambers, debate on the proposals to narrow the definition of navigable waters centered largely on the issue of wetlands preservation." 474 U.S., at 136. Beyond Congress' desire to regulate wetlands adjacent to "navigable waters," respondents point us to no persuasive evidence that the House bill was proposed in response to the Corps' claim of jurisdiction over nonnavigable, isolated, intrastate waters or that its failure indicated congressional acquiescence to such jurisdiction.

Section 404(g) is equally unenlightening. In Riverside Bayview Homes we recognized that Congress intended the phrase "navigable waters" to include "at least some waters that would not be deemed `navigable' under the classical understanding of that term." Id., at 133. But §404(g) gives no intimation of what those waters might be; it simply refers to them as "other _ waters." Respondents conjecture that "other _ waters" must incorporate the Corps' 1977 regulations, but it is also plausible, as petitioner contends, that Congress simply wanted to include all waters adjacent to "navigable waters," such as nonnavigable tributaries and streams. The exact meaning of §404(g) is not before us and we express no opinion on it, but for present purposes it is sufficient to say, as we did in Riverside Bayview Homes, that "§404(g)(1) does not conclusively determine the construction to be placed on the use of the term `waters' elsewhere in the Act (particularly in §502(7), which contains the relevant definition of `navigable waters') _ ." Id., at 138, n. 11.7

We thus decline respondents' invitation to take what they see as the next ineluctable step after Riverside Bayview Homes: holding that isolated ponds, some only seasonal, wholly located within two Illinois counties, fall under §404(a)'s definition of "navigable waters" because they serve as habitat for migratory birds. As counsel for respondents conceded at oral argument, such a ruling would assume that "the use of the word navigable in the statute _ does not have any independent significance." Tr. of Oral Arg. 28. We cannot agree that Congress' separate definitional use of the phrase "waters of the United States" constitutes a basis for reading the term "navigable waters" out of the statute. We said in Riverside Bayview Homes that the word "navigable" in the statute was of "limited effect" and went on to hold that §404(a) extended to nonnavigable wetlands adjacent to open waters. But it is one thing to give a word limited effect and quite another to give it no effect whatever. The term "navigable" has at least the import of showing us what Congress had in mind as its authority for enacting the CWA: its traditional jurisdiction over waters that were or had been navigable in fact or which could reasonably be so made. See, e.g., United States v. Appalachian Elec. Power Co., 311 U.S. 377, 407_408 (1940).

Respondents-relying upon all of the arguments addressed above-contend that, at the very least, it must be said that Congress did not address the precise question of §404(a)'s scope with regard to nonnavigable, isolated, intrastate waters, and that, therefore, we should give deference to the "Migratory Bird Rule." See, e.g., Chevron U.S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). We find §404(a) to be clear, but even were we to agree with respondents, we would not extend Chevron deference here.

Where an administrative interpretation of a statute invokes the outer limits of Congress' power, we expect a clear indication that Congress intended that result. See Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Constr. Trades Council, 485 U.S. 568, 575 (1988). This requirement stems from our prudential desire not to needlessly reach constitutional issues and our assumption that Congress does not casually authorize administrative agencies to interpret a statute to push the limit of congressional authority. See ibid. This concern is heightened where the administrative interpretation alters the federal-state framework by permitting federal encroachment upon a traditional state power. See United States v. Bass, 404 U.S. 336, 349 (1971) ("[U]nless Congress conveys its purpose clearly, it will not be deemed to have significantly changed the federal-state balance"). Thus, "where an otherwise acceptable construction of a statute would raise serious constitutional problems, the Court will construe the statute to avoid such problems unless such construction is plainly contrary to the intent of Congress." DeBartolo, supra, at 575.

Twice in the past six years we have reaffirmed the proposition that the grant of authority to Congress under the Commerce Clause, though broad, is not unlimited. See United States v. Morrison, 529 U.S. 598 (2000); United States v. Lopez, 514 U.S. 549 (1995). Respondents argue that the "Migratory Bird Rule" falls within Congress' power to regulate intrastate activities that "substantially affect" interstate commerce. They note that the protection of migratory birds is a "national interest of very nearly the first magnitude," Missouri v. Holland, 252 U.S. 416, 435 (1920), and that, as the Court of Appeals found, millions of people spend over a billion dollars annually on recreational pursuits relating to migratory birds. These arguments raise significant constitutional questions. For example, we would have to evaluate the precise object or activity that, in the aggregate, substantially affects interstate commerce. This is not clear, for although the Corps has claimed jurisdiction over petitioner's land because it contains water areas used as habitat by migratory birds, respondents now, post litem motam, focus upon the fact that the regulated activity is petitioner's municipal landfill, which is "plainly of a commercial nature." Brief for Federal Respondents 43. But this is a far cry, indeed, from the "navigable waters" and "waters of the United States" to which the statute by its terms extends.

These are significant constitutional questions raised by respondents' application of their regulations, and yet we find nothing approaching a clear statement from Congress that it intended §404(a) to reach an abandoned sand and gravel pit such as we have here. Permitting respondents to claim federal jurisdiction over ponds and mudflats falling within the "Migratory Bird Rule" would result in a significant impingement of the States' traditional and primary power over land and water use. See, e.g., Hess v. Port Authority Trans-Hudson Corporation, 513 U.S. 30, 44 (1994) ("[R]egulation of land use [is] a function traditionally performed by local governments"). Rather than expressing a desire to readjust the federal-state balance in this manner, Congress chose to "recognize, preserve, and protect the primary responsibilities and rights of States _ to plan the development and use _ of land and water resources _ ." 33 U.S.C. § 1251(b). We thus read the statute as written to avoid the significant constitutional and federalism questions raised by respondents' interpretation, and therefore reject the request for administrative deference.8

We hold that 33 CFR § 328.3(a)(3) (1999), as clarified and applied to petitioner's balefill site pursuant to the "Migratory Bird Rule," 51 Fed. Reg. 41217 (1986), exceeds the authority granted to respondents under §404(a) of the CWA. The judgment of the Court of Appeals for the Seventh Circuit is therefore

Reversed.

Justice Stevens, with whom Justice Souter, Justice Ginsburg, and Justice Breyer join, dissenting.

In 1969, the Cuyahoga River in Cleveland, Ohio, coated with a slick of industrial waste, caught fire. Congress responded to that dramatic event, and to others like it, by enacting the Federal Water Pollution Control Act (FWPCA) Amendments of 1972, 86 Stat. 817, as amended 33 U.S.C. § 1251 et seq., commonly known as the Clean Water Act (Clean Water Act, CWA, or Act).1 The Act proclaimed the ambitious goal of ending water pollution by 1985. §1251(a). The Court's past interpretations of the CWA have been fully consistent with that goal. Although Congress' vision of zero pollution remains unfulfilled, its pursuit has unquestionably retarded the destruction of the aquatic environment. Our Nation's waters no longer burn. Today, however, the Court takes an unfortunate step that needlessly weakens our principal safeguard against toxic water.

It is fair to characterize the Clean Water Act as "watershed" legislation. The statute endorsed fundamental changes in both the purpose and the scope of federal regulation of the Nation's waters. In §13 of the Rivers and Harbors Appropriation Act of 1899 (RHA), 30 Stat. 1152, as amended, 33 U.S.C. § 407 Congress had assigned to the Army Corps of Engineers (Corps) the mission of regulating discharges into certain waters in order to protect their use as highways for the transportation of interstate and foreign commerce; the scope of the Corps' jurisdiction under the RHA accordingly extended only to waters that were "navigable." In the CWA, however, Congress broadened the Corps' mission to include the purpose of protecting the quality of our Nation's waters for esthetic, health, recreational, and environmental uses. The scope of its jurisdiction was therefore redefined to encompass all of "the waters of the United States, including the territorial seas." §1362(7). That definition requires neither actual nor potential navigability.

The Court has previously held that the Corps' broadened jurisdiction under the CWA properly included an 80- acre parcel of low-lying marshy land that was not itself navigable, directly adjacent to navigable water, or even hydrologically connected to navigable water, but which was part of a larger area, characterized by poor drainage, that ultimately abutted a navigable creek. United States v. Riverside Bayview Homes, Inc., 474 U.S. 121 (1985).2 Our broad finding in Riverside Bayview that the 1977 Congress had acquiesced in the Corps' understanding of its jurisdiction applies equally to the 410-acre parcel at issue here. Moreover, once Congress crossed the legal watershed that separates navigable streams of commerce from marshes and inland lakes, there is no principled reason for limiting the statute's protection to those waters or wetlands that happen to lie near a navigable stream.

In its decision today, the Court draws a new jurisdictional line, one that invalidates the 1986 migratory bird regulation as well as the Corps' assertion of jurisdiction over all waters except for actually navigable waters, their tributaries, and wetlands adjacent to each. Its holding rests on two equally untenable premises: (1) that when Congress passed the 1972 CWA, it did not intend "to exert anything more than its commerce power over navigation," ante, at 7, n. 3; and (2) that in 1972 Congress drew the boundary defining the Corps' jurisdiction at the odd line on which the Court today settles.

As I shall explain, the text of the 1972 amendments affords no support for the Court's holding, and amendments Congress adopted in 1977 do support the Corps' present interpretation of its mission as extending to so-called "isolated" waters. Indeed, simple common sense cuts against the particular definition of the Corps' jurisdiction favored by the majority.

* The significance of the FWPCA Amendments of 1972 is illuminated by a reference to the history of federal water regulation, a history that the majority largely ignores. Federal regulation of the Nation's waters began in the 19th century with efforts targeted exclusively at "promot[ing] water transportation and commerce." Kalen, Commerce to Conservation: The Call for a National Water Policy and the Evolution of Federal Jurisdiction Over Wetlands, 69 N. D. L. Rev. 873, 877 (1993). This goal was pursued through the various Rivers and Harbors Acts, the most comprehensive of which was the RHA of 1899.3 Section 13 of the 1899 RHA, commonly known as the Refuse Act, prohibited the discharge of "refuse" into any "navigable water" or its tributaries, as well as the deposit of "refuse" on the bank of a navigable water "whereby navigation shall or may be impeded or obstructed" without first obtaining a permit from the Secretary of the Army. 30 Stat. 1152.

During the middle of the 20th century, the goals of federal water regulation began to shift away from an exclusive focus on protecting navigability and toward a concern for preventing environmental degradation. Kalen, 69 N. D. L. Rev., at 877_879, and n. 30. This awakening of interest in the use of federal power to protect the aquatic environment was helped along by efforts to reinterpret §13 of the RHA in order to apply its permit requirement to industrial discharges into navigable waters, even when such discharges did nothing to impede navigability. See, e.g., United States v. Republic Steel Corp., 362 U.S. 482, 490_491 (1960) (noting that the term "refuse" in §13 was broad enough to include industrial waste).4 Seeds of this nascent concern with pollution control can also be found in the FWPCA, which was first enacted in 1948 and then incrementally expanded in the following years.5

The shift in the focus of federal water regulation from protecting navigability toward environmental protection reached a dramatic climax in 1972, with the passage of the CWA. The Act, which was passed as an amendment to the existing FWPCA, was universally described by its supporters as the first truly comprehensive federal water pollution legislation. The "major purpose" of the CWA was "to establish a comprehensive long-range policy for the elimination of water pollution." S. Rep. No. 92_414, p. 95 (1971), reprinted in 2 Legislative History of the Water Pollution Control Act Amendments of 1972 (Committee Print compiled for the Senate Committee on Public Works by the Library of Congress), Ser. No. 93_1, p. 1511 (1971) (hereinafter Leg. Hist.) (emphasis added). And "[n]o Congressman's remarks on the legislation were complete without reference to [its] `comprehensive' nature _ ." Milwaukee v. Illinois, 451 U.S. 304, 318 (1981) (Rehnquist, J.). A House sponsor described the bill as "the most comprehensive and far-reaching water pollution bill we have ever drafted," 1 Leg. Hist. 369 (Rep. Mizell), and Senator Randolph, Chairman of the Committee on Public Works, stated: "It is perhaps the most comprehensive legislation that the Congress of the United States has ever developed in this particular field of the environment." 2 id., at 1269. This Court was therefore undoubtedly correct when it described the 1972 amendments as establishing "a comprehensive program for controlling and abating water pollution." Train v. City of New York, 420 U.S. 35, 37 (1975).

Section 404 of the CWA resembles §13 of the RHA, but, unlike the earlier statute, the primary purpose of which is the maintenance of navigability, §404 was principally intended as a pollution control measure. A comparison of the contents of the RHA and the 1972 Act vividly illustrates the fundamental difference between the purposes of the two provisions. The earlier statute contains pages of detailed appropriations for improvements in specific navigation facilities, 30 Stat. 1121_1149, for studies concerning the feasibility of a canal across the Isthmus of Panama, id., at 1150, and for surveys of the advisability of harbor improvements at numerous other locations, id., at 1155_1161. Tellingly, §13, which broadly prohibits the discharge of refuse into navigable waters, contains an exception for refuse "flowing from streets and sewers . . . in a liquid state." Id., at 1152.

The 1972 Act, in contrast, appropriated large sums of money for research and related programs for water pollution control, 86 Stat. 816_833, and for the construction of water treatment works, id., at 833_844. Strikingly absent from its declaration of "goals and policy" is any reference to avoiding or removing obstructions to navigation. Instead, the principal objective of the Act, as stated by Congress in §101, was "to restore and maintain the chemical, physical, and biological integrity of the Nation's waters." 33 U.S.C. § 1251. Congress therefore directed federal agencies in §102 to "develop comprehensive programs for preventing, reducing, or eliminating the pollution of the navigable waters and ground waters and improving the sanitary condition of surface and underground waters." 33 U.S.C. § 1252. The CWA commands federal agencies to give "due regard," not to the interest of unobstructed navigation, but rather to "improvements which are necessary to conserve such waters for the protection and propagation of fish and aquatic life and wildlife [and] recreational purposes." Ibid.

Because of the statute's ambitious and comprehensive goals, it was, of course, necessary to expand its jurisdictional scope. Thus, although Congress opted to carry over the traditional jurisdictional term "navigable waters" from the RHA and prior versions of the FWPCA, it broadened the definition of that term to encompass all "waters of the United States." §1362(7).6 Indeed, the 1972 conferees arrived at the final formulation by specifically deleting the word "navigable" from the definition that had originally appeared in the House version of the Act.7 The majority today undoes that deletion.

The Conference Report explained that the definition in §502(7) was intended to "be given the broadest possible constitutional interpretation." S. Conf. Rep. No. 92_1236, p. 144 (1972), reprinted in 1 Leg. Hist. 327. The Court dismisses this clear assertion of legislative intent with the back of its hand. Ante, at 7, n. 3. The statement, it claims, "signifies that Congress intended to exert [nothing] more than its commerce power over navigation." Ibid.

The majority's reading drains all meaning from the conference amendment. By 1972, Congress' Commerce Clause power over "navigation" had long since been established. The Daniel Ball, 10 Wall. 557 (1871); Gilman v. Philadelphia, 3 Wall. 713 (1866); Gibbons v. Ogden, 9 Wheat. 1 (1824). Why should Congress intend that its assertion of federal jurisdiction be given the "broadest possible constitutional interpretation" if it did not intend to reach beyond the very heartland of its commerce power? The activities regulated by the CWA have nothing to do with Congress' "commerce power over navigation." Indeed, the goals of the 1972 statute have nothing to do with navigation at all.

As we recognized in Riverside Bayview, the interests served by the statute embrace the protection of " `significant natural biological functions, including food chain production, general habitat, and nesting, spawning, rearing and resting sites' " for various species of aquatic wildlife. 474 U.S., at 134_135. For wetlands and "isolated" inland lakes, that interest is equally powerful, regardless of the proximity of the swamp or the water to a navigable stream. Nothing in the text, the stated purposes, or the legislative history of the CWA supports the conclusion that in 1972 Congress contemplated-much less commanded-the odd jurisdictional line that the Court has drawn today.

The majority accuses respondents of reading the term "navigable" out of the statute. Ante, at 11. But that was accomplished by Congress when it deleted the word from the §502(7) definition. After all, it is the definition that is the appropriate focus of our attention. Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U.S. 687, 697_698, n. 10 (1995) (refusing to be guided by the common-law definition of the term "take" when construing that term within the Endangered Species Act of 1973 and looking instead to the meaning of the terms contained in the definition of "take" supplied by the statute). Moreover, a proper understanding of the history of federal water pollution regulation makes clear that-even on respondents' broad reading-the presence of the word "navigable" in the statute is not inexplicable. The term was initially used in the various Rivers and Harbors Acts because (1) at the time those statutes were first enacted, Congress' power over the Nation's waters was viewed as extending only to "water bodies that were deemed `navigable' and therefore suitable for moving goods to or from markets," Power 513; and (2) those statutes had the primary purpose of protecting navigation. Congress' choice to employ the term "navigable waters" in the 1972 Clean Water Act simply continued nearly a century of usage. Viewed in light of the history of federal water regulation, the broad §502(7) definition, and Congress' unambiguous instructions in the Conference Report, it is clear that the term "navigable waters" operates in the statute as a shorthand for "waters over which federal authority may properly be asserted."

As the majority correctly notes, ante, at 7, when the Corps first promulgated regulations pursuant to §404 of the 1972 Act, it construed its authority as being essentially the same as it had been under the 1899 RHA.8 The reaction to those regulations in the federal courts,9 in the Environmental Protection Agency (EPA),10 and in Congress,11 convinced the Corps that the statute required it "to protect water quality to the full extent of the [C]ommerce [C]lause" and to extend federal regulation over discharges "to many areas that have never before been subject to Federal permits or to this form of water quality protection." 40 Fed. Reg. 31320 (1975).

In 1975, the Corps therefore adopted the interim regulations that we upheld in Riverside Bayview. As we noted in that case, the new regulations understood "the waters of the United States" to include, not only navigable waters and their tributaries, but also "nonnavigable intrastate waters whose use or misuse could affect interstate commerce." 474 U.S., at 123. The 1975 regulations provided that the new program would become effective in three phases: phase 1, which became effective immediately, encompassed the navigable waters covered by the 1974 regulation and the RHA; phase 2, effective after July 1, 1976, extended Corps jurisdiction to nonnavigable tributaries, freshwater wetlands adjacent to primary navigable waters, and lakes; and phase 3, effective after July 1, 1977, extended Corps jurisdiction to all other waters covered under the statute, including any waters not covered by phases 1 and 2 (such as "intermittent rivers, streams, tributaries, and perched wetlands that are not contiguous or adjacent to navigable waters") that "the District Engineer determines necessitate regulation for the protection of water quality." 40 Fed. Reg. 31325_31326 (1975). The final version of these regulations, adopted in 1977, made clear that the covered waters included "isolated lakes and wetlands, intermittent streams, prairie potholes, and other waters that are not part of a tributary system to interstate waters or to navigable waters of the United States, the degradation or destruction of which could affect interstate commerce."12

The Corps' broadened reading of its jurisdiction provoked opposition among some Members of Congress. As a result, in 1977, Congress considered a proposal that would have limited the Corps' jurisdiction under §404 to waters that are used, or by reasonable improvement could be used, as a means to transport interstate or foreign commerce and their adjacent wetlands. H. R. 3199, 95th Cong., 1st Sess., §16(f) (1977). A bill embodying that proposal passed the House but was defeated in the Senate. The debates demonstrate that Congress was fully aware of the Corps' understanding of the scope of its jurisdiction under the 1972 Act. We summarized these debates in our opinion in Riverside Bayview:

"In both Chambers, debate on the proposals to narrow the definition of navigable waters centered largely on the issue of wetlands preservation. See [123 Cong. Rec.], at 10426_10432 (House debate); id., at 26710_26729 (Senate debate). Proponents of a more limited §404 jurisdiction contended that the Corps' assertion of jurisdiction over wetlands and other nonnavigable `waters' had far exceeded what Congress had intended in enacting §404. Opponents of the proposed changes argued that a narrower definition of `navigable waters' for purposes of §404 would exclude vast stretches of crucial wetlands from the Corps' jurisdiction, with detrimental effects on wetlands ecosystems, water quality, and the aquatic environment generally. The debate, particularly in the Senate, was lengthy. In the House, the debate ended with the adoption of a narrowed definition of `waters'; but in the Senate the limiting amendment was defeated and the old definition retained. The Conference Committee adopted the Senate's approach: efforts to narrow the definition of `waters' were abandoned; the legislation as ultimately passed, in the words of Senator Baker, `retain[ed] the comprehensive jurisdiction over the Nation's waters exercised in the 1972 Federal Water Pollution Control Act.' " 474 U.S., at 136_137.

The net result of that extensive debate was a congressional endorsement of the position that the Corps maintains today. We explained in Riverside Bayview:

"[T]he scope of the Corps' asserted jurisdiction over wetlands was specifically brought to Congress' attention, and Congress rejected measures designed to curb the Corps' jurisdiction in large part because of its concern that protection of wetlands would be unduly hampered by a narrowed definition of `navigable waters.' Although we are chary of attributing significance to Congress' failure to act, a refusal by Congress to overrule an agency's construction of legislation is at least some evidence of the reasonableness of that construction, particularly where the administrative construction has been brought to Congress' attention through legislation specifically designed to supplant it." Id., at 137.

Even if the majority were correct that Congress did not extend the Corps' jurisdiction in the 1972 CWA to reach beyond navigable waters and their nonnavigable tributaries, Congress' rejection of the House's efforts in 1977 to cut back on the Corps' 1975 assertion of jurisdiction clearly indicates congressional acquiescence in that assertion. Indeed, our broad determination in Riverside Bayview that the 1977 Congress acquiesced in the very regulations at issue in this case should foreclose petitioner's present urgings to the contrary. The majority's refusal in today's decision to acknowledge the scope of our prior decision is troubling. Compare id., at 136 ("Congress acquiesced in the [1975] administrative construction [of the Corps' jurisdiction]"), with ante, at 9 ("We conclude that respondents have failed to make the necessary showing that the failure of the 1977 House bill demonstrates Congress' acquiescence to the Corps' regulations _").13 Having already concluded that Congress acquiesced in the Corps' regulatory definition of its jurisdiction, the Court is wrong to reverse course today. See Dickerson v. United States, 530 U.S. 428, 443 (2000) (Rehnquist, C. J.) (" `[T]he doctrine [of stare decisis] carries such persuasive force that we have always required a departure from precedent to be supported by some "special justification" ' ").

More important than the 1977 bill that did not become law are the provisions that actually were included in the 1977 revisions. Instead of agreeing with those who sought to withdraw the Corps' jurisdiction over "isolated" waters, Congress opted to exempt several classes of such waters from federal control. §67, 91 Stat. 1601, 33 U.S.C. § 1344(f). For example, the 1977 amendments expressly exclude from the Corps' regulatory power the discharge of fill material "for the purpose of construction or maintenance of farm or stock ponds or irrigation ditches, or the maintenance of drainage ditches," and "for the purpose of construction of temporary sedimentation basins on a construction site which does not include placement of fill material into the navigable waters." Ibid. The specific exemption of these waters from the Corps' jurisdiction indicates that the 1977 Congress recognized that similarly "isolated" waters not covered by the exceptions would fall within the statute's outer limits.

In addition to the enumerated exceptions, the 1977 amendments included a new section, §404(g), which authorized the States to administer their own permit programs over certain nonnavigable waters. Section 404(g)(1) provides, in relevant part:

"The Governor of any State desiring to administer its own individual and general permit program for the discharge of dredged or fill material into the navigable waters (other than those waters which are presently used, or are susceptible to use in their natural condition or by reasonable improvement as a means to transport interstate or foreign commerce . . . , including wetlands adjacent thereto) within its jurisdiction may submit to the Administrator a full and complete description of the program it proposes to establish and administer under State law or under an interstate compact." 33 U.S.C. § 1344(g)(1).

Section 404(g)(1)'s reference to navigable waters "other than those waters which are presently used, or are susceptible to use" for transporting commerce and their adjacent wetlands appears to suggest that Congress viewed (and accepted) the Act's regulations as covering more than navigable waters in the traditional sense. The majority correctly points out that §404(g)(1) is itself ambiguous because it does not indicate precisely how far Congress considered federal jurisdiction to extend. Ante, at 10. But the Court ignores the provision's legislative history, which makes clear that Congress understood §404(g)(1)-and therefore federal jurisdiction-to extend, not only to navigable waters and nonnavigable tributaries, but also to "isolated" waters, such as those at issue in this case.

The Conference Report discussing the 1977 amendments, for example, states that §404(g) "establish[es] a process to allow the Governor of any State to administer an individual and general permit program for the discharge of dredged or fill material into phase 2 and 3 waters after the approval of a program by the Administrator." H. R. Conf. Rep. No. 95_830, p. 101 (1977), reprinted in 3 Legislative History of the Clean Water Act of 1977 (Committee Print Compiled for the Committee on Environment and Public Works by the Library of Congress), Ser. No. 95_14, p. 285 (emphasis added) (hereinafter Leg. Hist. of CWA). Similarly, a Senate Report discussing the 1977 amendments explains that, under §404(g), "the [C]orps will continue to administer the section 404 permit program in all navigable waters for a discharge of dredge or fill material until the approval of a State program for phase 2 and 3 waters." S. Rep. No. 95_370, p. 75 (1977), reprinted in 4 Leg. Hist. of CWA 708 (emphases added).

Of course, as I have already discussed, "phase 1" waters are navigable waters and their contiguous wetlands, "phase 2" waters are the "primary tributaries" of navigable waters and their adjacent wetlands, and "phase 3" waters are all other waters covered by the statute, and can include such "isolated" waters as "intermittent rivers, streams, tributaries, and perched wetlands that are not contiguous or adjacent to navigable waters." The legislative history of the 1977 amendments therefore plainly establishes that, when it enacted §404(g), Congress believed-and desired-the Corps' jurisdiction to extend beyond just navigable waters, their tributaries, and the wetlands adjacent to each.

In dismissing the significance of §404(g)(1), the majority quotes out of context language in the very same 1977 Senate Report that I have quoted above. Ante, at 10, n. 6. It is true that the Report states that "[t]he committee amendment does not redefine navigable waters." S. Rep. No. 95_370, at 75, reprinted in 4 Leg. Hist. of CWA 708 (emphasis added). But the majority fails to point out that the quoted language appears in the course of an explanation of the Senate's refusal to go along with House efforts to narrow the scope of the Corps' CWA jurisdiction to traditionally navigable waters. Thus, the immediately preceding sentence warns that "[t]o limit the jurisdiction of the [FWPCA] with reference to discharges of the pollutants of dredged or fill material would cripple efforts to achieve the act's objectives."14 Ibid. The Court would do well to heed that warning.

The majority also places great weight, ante, at 10, on our statement in Riverside Bayview that §404(g) "does not conclusively determine the construction to be placed on the use of the term `waters' elsewhere in the Act," 474 U.S., at 138, n. 11 (emphasis added). This is simply more selective reading. In that case, we also went on to say with respect to the significance of §404(g) that "the various provisions of the Act should be read in pari materia." Ibid. More-over, our ultimate conclusion in Riverside Bayview was that §404(g) "suggest[s] strongly that the term `waters' as used in the Act" supports the Corps' reading. Ibid.

Although it might have appeared problematic on a "linguistic" level for the Corps to classify "lands" as "waters" in Riverside Bayview, 474 U.S., at 131_132, we squarely held that the agency's construction of the statute that it was charged with enforcing was entitled to deference under Chevron U.S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). Today, however, the majority refuses to extend such deference to the same agency's construction of the same statute, see ante, at 11_13. This refusal is unfaithful to both Riverside Bayview and Chevron. For it is the majority's reading, not the agency's, that does violence to the scheme Congress chose to put into place.

Contrary to the Court's suggestion, the Corps' interpretation of the statute does not "encroac[h]" upon "traditional state power" over land use. Ante, at 12. "Land use planning in essence chooses particular uses for the land; environmental regulation, at its core, does not mandate particular uses of the land but requires only that, however the land is used, damage to the environment is kept within prescribed limits." California Coastal Comm'n v. Granite Rock Co., 480 U.S. 572, 587 (1987). The CWA is not a land-use code; it is a paradigm of environmental regulation. Such regulation is an accepted exercise of federal power. Hodel v. Virginia Surface Mining & Reclamation Assn., Inc., 452 U.S. 264, 282 (1981).

It is particularly ironic for the Court to raise the specter of federalism while construing a statute that makes explicit efforts to foster local control over water regulation. Faced with calls to cut back on federal jurisdiction over water pollution, Congress rejected attempts to narrow the scope of that jurisdiction and, by incorporating §404(g), opted instead for a scheme that encouraged States to supplant federal control with their own regulatory programs. S. Rep. No. 95_370, at 75, reprinted in 4 Leg. Hist. of CWA 708 ("The committee amendment does not redefine navigable waters. Instead, the committee amendment intends to assure continued protection of all the Nation's waters, but allows States to assume the primary responsibility for protecting those lakes, rivers, streams, swamps, marshes, and other portions of the navigable waters outside the [C]orps program in the so-called phase I waters" (emphasis added)). Because Illinois could have taken advantage of the opportunities offered to it through §404(g), the federalism concerns to which the majority adverts are misplaced. The Corps' interpretation of the statute as extending beyond navigable waters, tributaries of navigable waters, and wetlands adjacent to each is manifestly reasonable and therefore entitled to deference.

Because I am convinced that the Court's miserly construction of the statute is incorrect, I shall comment briefly on petitioner's argument that Congress is without power to prohibit it from filling any part of the 31 acres of ponds on its property in Cook County, Illinois. The Corps' exercise of its §404 permitting power over "isolated" waters that serve as habitat for migratory birds falls well within the boundaries set by this Court's Commerce Clause jurisprudence.

In United States v. Lopez, 514 U.S. 549, 558_559 (1995), this Court identified "three broad categories of activity that Congress may regulate under its commerce power": (1) channels of interstate commerce; (2) instrumentalities of interstate commerce, or persons and things in interstate commerce; and (3) activities that "substantially affect" interstate commerce. Ibid. The migratory bird rule at issue here is properly analyzed under the third category. In order to constitute a proper exercise of Congress' power over intrastate activities that "substantially affect" interstate commerce, it is not necessary that each individual instance of the activity substantially affect commerce; it is enough that, taken in the aggregate, the class of activities in question has such an effect. Perez v. United States, 402 U.S. 146 (1971) (noting that it is the "class" of regulated activities, not the individual instance, that is to be considered in the "affects" commerce analysis); see also Hodel, 452 U.S., at 277; Wickard v. Filburn, 317 U.S. 111, 127_128 (1942).

The activity being regulated in this case (and by the Corps' §404 regulations in general) is the discharge of fill material into water. The Corps did not assert jurisdiction over petitioner's land simply because the waters were "used as habitat by migratory birds." It asserted jurisdiction because petitioner planned to discharge fill into waters "used as habitat by migratory birds." Had petitioner intended to engage in some other activity besides discharging fill (i.e., had there been no activity to regulate), or, conversely, had the waters not been habitat for migratory birds (i.e., had there been no basis for federal jurisdiction), the Corps would never have become involved in petitioner's use of its land. There can be no doubt that, unlike the class of activities Congress was attempting to regulate in United States v. Morrison, 529 U.S. 598, 613 (2000) ("[g]ender-motivated crimes"), and Lopez, 514 U.S., at 561 (possession of guns near school property), the discharge of fill material into the Nation's waters is almost always undertaken for economic reasons. See V. Albrecht & B. Goode, Wetland Regulation in the Real World, Exh. 3 (Feb. 1994) (demonstrating that the overwhelming majority of acreage for which §404 permits are sought is intended for commercial, industrial, or other economic use).15

Moreover, no one disputes that the discharge of fill into "isolated" waters that serve as migratory bird habitat will, in the aggregate, adversely affect migratory bird populations. See, e.g., 1 Secretary of the Interior, Report to Congress, The Impact of Federal Programs on Wetlands: The Lower Mississippi Alluvial Plain and the Prairie Pothole Region 79_80 (Oct. 1988) (noting that "isolated," phase 3 waters "are among [the] most important and also the most threatened ecosystems in the United States" because "[t]hey are prime nesting grounds for many species of North American waterfowl _" and provide "[u]p to 50 percent of the [U.S.] production of migratory waterfowl"). Nor does petitioner dispute that the particular waters it seeks to fill are home to many important species of migratory birds, including the second-largest breeding colony of Great Blue Herons in northeastern Illinois, App. to Pet. for Cert. 3a, and several species of waterfowl protected by international treaty and Illinois endangered species laws, Brief for Federal Respondents 7.16

In addition to the intrinsic value of migratory birds, see Missouri v. Holland, 252 U.S. 416, 435 (1920) (noting the importance of migratory birds as "protectors of our forests and our crops" and as "a food supply"), it is undisputed that literally millions of people regularly participate in birdwatching and hunting and that those activities generate a host of commercial activities of great value.17 The causal connection between the filling of wetlands and the decline of commercial activities associated with migratory birds is not "attenuated," Morrison, 529 U.S., at 612; it is direct and concrete. Cf. Gibbs v. Babbitt, 214 F.3d 483, 492_493 (CA4 2000) ("The relationship between red wolf takings and interstate commerce is quite direct-with no red wolves, there will be no red wolf related tourism _").

Finally, the migratory bird rule does not blur the "distinction between what is truly national and what is truly local." Morrison, 529 U.S., at 617_618. Justice Holmes cogently observed in Missouri v. Holland that the protection of migratory birds is a textbook example of a national problem. 252 U.S., at 435 ("It is not sufficient to rely upon the States [to protect migratory birds]. The reliance is vain _"). The destruction of aquatic migratory bird habitat, like so many other environmental problems, is an action in which the benefits (e.g., a new landfill) are disproportionately local, while many of the costs (e.g., fewer migratory birds) are widely dispersed and often borne by citizens living in other States. In such situations, described by economists as involving "externalities," federal regulation is both appropriate and necessary. Revesz, Rehabilitating Interstate Competition: Rethinking the "Race-to-the-Bottom" Rationale for Federal Environmental Regulation, 67 N. Y. U. L. Rev. 1210, 1222 (1992) ("The presence of interstate externalities is a powerful reason for intervention at the federal level"); cf. Hodel, 452 U.S., at 281_282 (deferring to Congress' finding that nationwide standards were "essential" in order to avoid "destructive interstate competition" that might undermine environmental standards). Identifying the Corps' jurisdiction by reference to waters that serve as habitat for birds that migrate over state lines also satisfies this Court's expressed desire for some "jurisdictional element" that limits federal activity to its proper scope. Morrison, 529 U.S., at 612.

The power to regulate commerce among the several States necessarily and properly includes the power to preserve the natural resources that generate such commerce. Cf. Sporhase v. Nebraska ex rel. Douglas, 458 U.S. 941, 953 (1982) (holding water to be an "article of commerce"). Migratory birds, and the waters on which they rely, are such resources. Moreover, the protection of migratory birds is a well-established federal responsibility. As Justice Holmes noted in Missouri v. Holland, the federal interest in protecting these birds is of "the first magnitude." 252 U.S., at 435. Because of their transitory nature, they "can be protected only by national action." Ibid.

Whether it is necessary or appropriate to refuse to allow petitioner to fill those ponds is a question on which we have no voice. Whether the Federal Government has the power to require such permission, however, is a question that is easily answered. If, as it does, the Commerce Clause empowers Congress to regulate particular "activities causing air or water pollution, or other environmental hazards that may have effects in more than one State," Hodel, 452 U.S., at 282, it also empowers Congress to control individual actions that, in the aggregate, would have the same effect. Perez, 402 U.S., at 154; Wickard, 317 U.S., at 127_128.18 There is no merit in petitioner's constitutional argument.

Because I would affirm the judgment of the Court of Appeals, I respectfully dissent.

The issue presented rests upon the initial assumption, which we accept for analytic purposes, that the trial court erred in a Sentencing Guidelines determination after petitioner's conviction of a federal offense. The legal error, petitioner alleges, increased his prison sentence by at least 6 months and perhaps by 21 months. We must decide whether this would be "prejudice" under Strickland v. Washington, 466 U.S. 668 (1984). The Government is not ready to concede error in the sentencing determination but now acknowledges that if an increased prison term did flow from an error the petitioner has established Strickland prejudice. In agreement with the Government and petitioner on this point, we reverse and remand for further proceedings.

* In the 1980's and early 1990's, petitioner Paul Glover was the Vice President and General Counsel of the Chicago Truck Drivers, Helpers, and Warehouse Workers Union (Independent). The evidence showed Glover used his control over the union's investments to enrich himself and his co-conspirators through kickbacks. When the malfeasance was discovered, he was tried in the United States District Court for the Northern District of Illinois. His first trial ended when the jury could not agree, but a second jury convicted him. The presentence investigation report prepared by the probation office recommended that the convictions for labor racketeering, money laundering, and tax evasion be grouped together under United States Sentencing Commission, Guidelines Manual §3D1.2 (Nov. 1994), which allows the grouping of "counts involving substantially the same harm." The Government, insisting that the money laundering counts could not be grouped with the other counts, objected to that recommendation, and the District Court held a hearing on the matter. The money laundering counts, it ruled, should not be grouped with Glover's other offenses. The ruling, as the trial court viewed it, was in conformance with decisions in those Courts of Appeals which had refused to group money laundering counts with other counts for various reasons. See, e.g., United States v. Lombardi, 5 F.3d 568 (CA1 1993); United States v. Porter, 909 F.2d 789 (CA4 1990); United States v. Taylor, 984 F.2d 298 (CA9 1993); United States v. Johnson, 971 F.2d 562 (CA10 1992); United States v. Harper, 972 F.2d 321 (CA11 1992). In the trial court, Glover's attorneys did not submit papers or offer extensive oral arguments contesting the no-grouping argument advanced by the Government. When the District Court decided not to group the money laundering counts with the other counts, Glover's offense level was increased by two levels, yielding a concomitant increase in the sentencing range. Glover was sentenced to 84 months in prison, which was in the middle of the Guidelines range of 78 to 97 months.

On appeal to the Seventh Circuit, Glover's counsel (the same attorneys who represented him in District Court) did not raise the grouping issue; instead, they concentrated on claims that certain testimony from his first trial should not have been admitted at his second trial and that he should not have been assessed a two-level increase for perjury at his first trial. A short time after argument on Glover's appeal, a different panel of the Seventh Circuit held that, under some circumstances, grouping of money laundering offenses with other counts was proper under §3D1.2. United States v. Wilson, 98 F.3d 281 (1996). A month and a half later, the Seventh Circuit rejected both of Glover's arguments and affirmed his conviction and sentence. 101 F.3d 1183 (1996).

Glover filed a pro se motion to correct his sentence under 28 U.S.C. § 2255 (1994 ed. and Supp. III). The failure of his counsel to press the grouping issue, he argued, was ineffective assistance, a position confirmed, in his view, by the Court of Appeals' decision in Wilson. The performance of counsel, he contended, fell below a reasonable standard both at sentencing, when his attorneys did not with any clarity or force contest the Government's argument, and on appeal, when they did not present the issue in their briefs or call the Wilson decision to the panel's attention following the oral argument. He further argued that absent the ineffective assistance, his offense level would have been two levels lower, yielding a Guidelines sentencing range of 63 to 78 months. Under this theory, the 84-month sentence he received was an unlawful increase of anywhere between 6 and 21 months.

The District Court denied Glover's motion, determining that under Seventh Circuit precedent an increase of 6 to 21 months in a defendant's sentence was not significant enough to amount to prejudice for purposes of Strickland v. Washington, supra. As a result, the District Court did not decide the issue whether the performance of Glover's counsel fell below a reasonable standard of competence. On appeal to the Seventh Circuit, the Government argued only that Glover had not suffered prejudice within the meaning of Strickland. See App. to Reply Brief for Petitioner 1a_22a. Citing Durrive v. United States, 4 F.3d 548 (CA7 1993), the Government contended that even were the performance of Glover's counsel ineffective, the resulting additional 6 to 21 months, under the law as established in the Seventh Circuit, would not constitute prejudice. App. to Reply Brief for Petitioner 21a. The Court of Appeals affirmed, relying on that theory. 182 F.3d 921 (1999) (table). We granted Glover's petition for certiorari. 530 U.S. ____ (2000).

The Government no longer puts forth the proposition that a 6 to 21 month prison term increase is not prejudice under Strickland. It now acknowledges that such a rule, without more, would be "inconsistent with this Court's cases and unworkable." Brief for United States 18.

It appears the Seventh Circuit drew the substance of its no-prejudice rule from our opinion in Lockhart v. Fretwell, 506 U.S. 364 (1993). Lockhart holds that in some circumstances a mere difference in outcome will not suffice to establish prejudice. Id., at 369. The Seventh Circuit extracted from this holding the rule at issue here, which denies relief when the increase in sentence is said to be not so significant as to render the outcome of sentencing unreliable or fundamentally unfair. See Durrive, supra, at 550_551. The Court explained last Term that our holding in Lockhart does not supplant the Strickland analysis. See Williams v. Taylor, 529 U.S. 362, 393 (2000) ("Cases such as Nix v. Whiteside, 475 U.S. 157 (1986), and Lockhart v. Fretwell, 506 U.S. 364 (1993), do not justify a departure from a straightforward application of Strickland when the ineffectiveness of counsel does deprive the defendant of a substantive or procedural right to which the law entitles him"); id., at 414 (opinion of O'Connor, J.) ("As I explained in my concurring opinion in [Lockhart], `in the vast majority of cases . . . [t]he determinative question-whether there is "a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different"-remains unchanged' "). The Seventh Circuit was incorrect to rely on Lockhart to deny relief to persons attacking their sentence who might show deficient performance in counsel's failure to object to an error of law affecting the calculation of a sentence because the sentence increase does not meet some baseline standard of prejudice. Authority does not suggest that a minimal amount of additional time in prison cannot constitute prejudice. Quite to the contrary, our jurisprudence suggests that any amount of actual jail time has Sixth Amendment significance. Compare Argersinger v. Hamlin, 407 U.S. 25 (1972) (holding that the assistance of counsel must be provided when a defendant is tried for a crime that results in a sentence of imprisonment), with Scott v. Illinois, 440 U.S. 367 (1979) (holding that a criminal defendant has no Sixth Amendment right to counsel when his trial does not result in a sentence of imprisonment). Our decisions on the right to jury trial in a criminal case do not suggest that there is no prejudice in the circumstances here. Those cases have limited the right to jury trial to offenses where the potential punishment was imprisonment for six months or more. See Argersinger, supra, at 29 (citing Duncan v. Louisiana, 391 U.S. 145 (1968)). But they do not control the question whether a showing of prejudice, in the context of a claim for ineffective assistance of counsel, requires a significant increase in a term of imprisonment.

The Seventh Circuit's rule is not well considered in any event, because there is no obvious dividing line by which to measure how much longer a sentence must be for the increase to constitute substantial prejudice. Indeed, it is not even clear if the relevant increase is to be measured in absolute terms or by some fraction of the total authorized sentence. See Martin v. United States, 109 F.3d 1177, 1183 (CA7 1996) (Rovner, J., dissenting from denial of rehearing en banc). Although the amount by which a defendant's sentence is increased by a particular decision may be a factor to consider in determining whether counsel's performance in failing to argue the point constitutes ineffective assistance, under a determinate system of constrained discretion such as the Sentencing Guidelines it cannot serve as a bar to a showing of prejudice. Compare Spriggs v. Collins, 993 F.2d 85, 88 (CA5 1993) (requiring a showing that a sentence would have been "significantly less harsh" under the Texas discretionary sentencing scheme), with United States v. Phillips, 210 F.3d 345 (CA5 2000) (finding prejudice under the Sentencing Guidelines when an error by counsel led to an increased sentence). We hold that the Seventh Circuit erred in engrafting this additional requirement onto the prejudice branch of the Strickland test. This is not a case where trial strategies, in retrospect, might be criticized for leading to a harsher sentence. Here we consider the sentencing calculation itself, a calculation resulting from a ruling which, if it had been error, would have been correctable on appeal. We express no opinion on the ultimate merits of Glover's claim because the question of deficient performance is not before us, but it is clear that prejudice flowed from the asserted error in sentencing.

The Government makes various arguments for alternative grounds to affirm the Court of Appeals. Among other contentions, the Government suggests that the failure of Glover's counsel to argue for grouping of the money laundering counts was not deficient; that Glover's grouping claim has no legal merit in any event; and that even if Glover had prevailed on his grouping claim, his sentence in fact would have increased as a result. Glover disputes these contentions. We need not describe the arguments in great detail, because despite the fact the parties have joined issue at least in part on these points, they were neither raised in nor passed upon by the Court of Appeals. In the ordinary course we do not decide questions neither raised nor resolved below. See Taylor v. Freeland & Kronz, 503 U.S. 638, 646 (1992). As a general rule, furthermore, we do not decide issues outside the questions presented by the petition for certiorari. This Court's Rule 14.1(a). Whether these issues remain open, and if so whether they have merit, are questions for the Court of Appeals or the District Court to consider and determine in the first instance.

The judgment of the Seventh Circuit is reversed. The case is remanded for further proceedings consistent with this opinion.

It is so ordered.

The Commissioner of Internal Revenue assessed tax deficiencies against petitioners David and Louise Gitlitz and Philip and Eleanor Winn because they used nontaxed discharge of indebtedness to increase their bases in S corporation stock and to deduct suspended losses. In this case we must answer two questions. First, we must decide whether the Internal Revenue Code (Code) permits taxpayers to increase bases in their S corporation stock by the amount of an S corporation's discharge of indebtedness excluded from gross income. And, second, if the Code permits such an increase, we must decide whether the increase occurs before or after taxpayers are required to reduce the S corporation's tax attributes.

* David Gitlitz and Philip Winn1 were shareholders of P. D. W. & A., Inc., a corporation that had elected to be taxed under subchapter S of the Code, 26 U.S.C. § 13611379 (1994 ed. and Supp. III). Subchapter S allows shareholders of qualified corporations to elect a "pass-through" taxation system under which income is subjected to only one level of taxation. See Bufferd v. Commissioner, 506 U.S. 523, 525 (1993). The corporation's profits pass through directly to its shareholders on a pro rata basis and are reported on the shareholders' individual tax returns. See §1366(a)(1)(A).2 To prevent double taxation of income upon distribution from the cor-poration to the shareholders, §1367(a)(1)(A) permits shareholders to increase their corporate bases by items of income identified in §1366(a) (1994 ed. and Supp. III). Corporate losses and deductions are passed through in a similar manner, see §1366(a)(1)(A), and the shareholders' bases in the S corporation's stock and debt are decreased accordingly, see §§1367(a)(2)(B), 1367(b)(2)(A). However, a shareholder cannot take corporate losses and deductions into account on his personal tax return to the extent that such items exceed his basis in the stock and debt of the S corporation. See §1366(d)(1) (Supp. III). If those items exceed the basis, the excess is "suspended" until the shareholder's basis becomes large enough to permit the deduction. See §§1366(d)(1)(2) (1994 ed. and Supp. III).

In 1991, P. D. W. & A. realized $2,021,296 of discharged indebtedness. At the time, the corporation was insolvent in the amount of $2,181,748. Because it was insolvent even after the discharge of indebtedness was added to its balance sheet, P. D. W. & A. excluded the entire discharge of indebtedness amount from gross income under 26 U.S.C. § 108(a) and 108(d)(7)(A). On their tax returns, Gitlitz and Winn increased their bases in P. D. W. & A. stock by their pro rata share (50 percent each) of the amount of the corporation's discharge of indebtedness. Petitioners' theory was that the discharge of indebtedness was an "item of income" subject to pass-through under §1366(a)(1)(A). They used their increased bases to deduct on their personal tax returns corporate losses and deductions, including losses and deductions from previous years that had been suspended under §1366(d). Gitlitz and Winn each had losses (including suspended losses and operating losses) that totaled $1,010,648. With the upward basis adjustments of $1,010,648 each, Gitlitz and Winn were each able to deduct the full amount of their pro rata share of P. D. W. & A.'s losses.

The Commissioner determined that petitioners could not use P. D. W. & A.'s discharge of indebtedness to increase their bases in the stock and denied petitioners' loss deductions. Petitioners petitioned the Tax Court to review the deficiency determinations. The Tax Court, in its initial opinion, granted relief to petitioners and held that the discharge of indebtedness was an "item of income" and therefore could support a basis increase. See Winn v. Commissioner, 73 TCM 3167 (1997), ¶97,286 RIA Memo withdrawn and reissued, 75 TCM 1840 (1998), ¶98,071 RIA Memo TC. In light of the Tax Court's decision in Nelson v. Commissioner, 110 T. C. 114 (1998), aff'd, 182 F.3d 1152 (CA10 1999),3 however, the Tax Court granted the Commissioner's motion for reconsideration and held that shareholders may not use an S corporation's untaxed discharge of indebtedness to increase their bases in corporate stock. See Winn v. Commissioner, 75 TCM 1840 (1998), ¶98,071 RIA Memo TC.

The Court of Appeals affirmed. See 182 F.3d 1143 (CA10 1999). It assumed that excluded discharge of indebtedness is an item of income subject to passthrough to shareholders pursuant to §1366(a)(1)(A), id., at 1148, 1151, n. 7, but held that the discharge of indebtedness amount first had to be used to reduce certain tax attributes of the S corporation under §108(b), and that only the leftover amount could be used to increase basis.4 The Court of Appeals explained that, because the tax attribute to be reduced (in this case the corporation's net operating loss) was equal to the amount of discharged debt, the entire amount of discharged debt was absorbed by the reduction at the corporate level, and nothing remained of the discharge of indebtedness to be passed through to the shareholders under §1366(a)(1)(A). Id., at 1151. Because Courts of Appeals have disagreed on how to treat discharge of indebtedness of an insolvent S corporation, compare Gaudiano v. Commissioner, 216 F.3d 524, 535 (CA6 2000) (holding that tax attributes are reduced before excluded discharged debt income is passed through to shareholders), cert. pending, No. 00_459, Witzel v. Commissioner, 200 F.3d 496, 498 (CA7 2000) (same), cert. pending, No. 99_1693, and 182 F.3d, at 1150 (case below), with United States v. Farley, 202 F.3d 198, 206 (CA3 2000) (holding that excluded discharged debt income is passed through to shareholders before tax attributes are reduced), cert. pending, No. 99_1675; see also Pugh v. Commissioner, 213 F.3d 1324, 1330 (CA11 2000) (holding that excluded discharged debt income is subject to pass-through and can increase basis), cert. pending, No. 00_242, we granted certiorari. 529 U.S. 1097 (2000).

Before we can reach the issue addressed by the Court of Appeals-whether the increase in the taxpayers' corporate bases occurs before or after the taxpayers are required to reduce the S corporation's tax attributes-we must address the argument raised by the Commissioner.5 The Commissioner argues that the discharge of indebtedness of an insolvent S corporation is not an "item of income" and thus never passes through to shareholders. Under a plain reading of the statute, we reject this argument and conclude that excluded discharged debt is indeed an "item of income," which passes through to the shareholders and increases their bases in the stock of the S corporation.

Section 61(a)(12) states that discharge of indebtedness generally is included in gross income. Section 108(a)(1) provides an express exception to this general rule:

"Gross income does not include any amount which (but for this subsection) would be includible in gross income by reason of the discharge _ of indebtedness of the taxpayer if-

. . . . .

"(B) the discharge occurs when the taxpayer is insolvent."

The Commissioner contends that this exclusion from gross income alters the character of the discharge of indebtedness so that it is no longer an "item of income." However, the text and structure of the statute do not support the Commissioner's theory. Section 108(a) simply does not say that discharge of indebtedness ceases to be an item of income when the S corporation is insolvent. Instead it provides only that discharge of indebtedness ceases to be included in gross income. Not all items of income are included in gross income, see §1366(a)(1) (providing that "items of income," including "tax-exempt" income, are passed through to shareholders), so mere exclusion of an amount from gross income does not imply that the amount ceases to be an item of income. Moreover, §§101 through 136 employ the same construction to exclude various items from gross income: "Gross income does not include _ ." The consequence of reading this language in the manner suggested by the Commissioner would be to exempt all items in these sections from pass-through under §1366. However, not even the Commissioner encourages us to reach this sweeping conclusion. Instead the Commissioner asserts that discharge of indebtedness is unique among the types of items excluded from gross income because no economic outlay is required of the taxpayer receiving discharge of indebtedness. But the Commissioner is unable to identify language in the statute that makes this distinction relevant, and we certainly find none.

On the contrary, the statute makes clear that §108(a)'s exclusion does not alter the character of discharge of indebtedness as an item of income. Specifically, §108(e)(1) reads:

"Except as otherwise provided in this section, there shall be no insolvency exception from the general rule that gross income includes income from the discharge of indebtedness."

This provision presumes that discharge of indebtedness is always "income," and that the only question for purposes of §108 is whether it is includible in gross income. If discharge of indebtedness of insolvent entities were not actually "income," there would be no need to provide an exception to its inclusion in gross income; quite simply, if discharge of indebtedness of an insolvent entity were not "income," it would necessarily not be included in gross income.

Notwithstanding the plain language of the statute, the Commissioner argues, generally, that excluded discharge of indebtedness is not income and, specifically, that it is not "tax-exempt income" under §1366(a)(1)(A).6 First, the Commissioner argues that §108 merely codified the "judicial insolvency exception," and that, under this exception, discharge of indebtedness of an insolvent taxpayer was not considered income. The insolvency exception was a rule that the discharge of indebtedness of an insolvent taxpayer was not taxable income. See, e.g., Dallas Transfer & Terminal Warehouse Co. v. Commissioner, 70 F.2d 95 (CA5 1934); Astoria Marine Construction Co. v. Commissioner, 12 T. C. 798 (1949). But the exception has since been limited by §108(e). Section 108(e) precludes us from relying on any understanding of the judicial insolvency exception that was not codified in §108. And as explained above, the language and logic of §108 clearly establish that, although discharge of indebtedness of an insolvent taxpayer is not included in gross income, it is nevertheless income.

The Commissioner also relies on a Treasury Regulation to support his theory that no income is realized from the discharge of the debt of an insolvent:

"Proceedings under Bankruptcy Act.

"(1) Income is not realized by a taxpayer by virtue of the discharge, under section 14 of the Bankruptcy Act (11 U.S.C. 32), of his indebtedness as the result of an adjudication in bankruptcy, or by virtue of an agreement among his creditors not consummated under any provision of the Bankruptcy Act, if immediately thereafter the taxpayer's liabilities exceed the value of his assets." 26 CFR § 1.61_12(b) (2000).

Even if this regulation could be read (countertextually) to apply outside the bankruptcy context, it merely states that "[i]ncome is not realized." The regulation says nothing about whether discharge of indebtedness is income subject to pass-through under §1366.

Second, the Commissioner argues that excluded discharge of indebtedness is not "tax-exempt" income under §1366(a)(1)(A), but rather "tax-deferred" income. According to the Commissioner, because the taxpayer is required to reduce tax attributes that could have provided future tax benefits, the taxpayer will pay taxes on future income that otherwise would have been absorbed by the forfeited tax attributes. Implicit in the Commissioner's labeling of such income as "tax-deferred," however, is the erroneous assumption that §1366(a)(1)(A) does not include "tax-deferred" income. Section 1366 applies to "items of income." This section expressly includes "tax-exempt" income, but this inclusion does not mean that the statute must therefore exclude "tax-deferred" income. The section is worded broadly enough to include any item of income, even tax-deferred income, that "could affect the liability for tax of any shareholder." §1366(a)(1)(A). Thus, none of the Commissioner's contentions alters our conclusion that discharge of indebtedness of an insolvent S corporation is an item of income for purposes of §1366(a)(1)(A).

Having concluded that excluded discharge of indebtedness is an "item of income" and is therefore subject to pass-through to shareholders under §1366, we must resolve the sequencing question addressed by the Court of Appeals-whether pass-through is performed before or after the reduction of the S corporation's tax attributes under §108(b). Section 108(b)(1) provides that "[t]he amount excluded from gross income under [§108(a)] shall be applied to reduce the tax attributes of the taxpayer as provided [in this section]." Section 108(b)(2) then lists the various tax attributes to be reduced in the order of reduction. The first tax attribute to be reduced, and the one at issue in this case, is the net operating loss. See §108(b)(2)(A). Section 108(d)(7)(B) specifies that, for purposes of attribute reduction, the shareholders' suspended losses for the taxable year of discharge are to be treated as the S corporation's net operating loss. If tax attribute reduction is performed before the discharge of indebtedness is passed through to the shareholders (as the Court of Appeals held), the shareholders' losses that exceed basis are treated as the corporation's net operating loss and are then reduced by the amount of the discharged debt. In this case, no suspended losses would remain that would permit petitioners to take deductions.7 If, however, attribute reduction is performed after the discharged debt income is passed through (as petitioners argue), then the shareholders would be able to deduct their losses (up to the amount of the increase in basis caused by the discharged debt). Any suspended losses remaining then will be treated as the S corporation's net operating loss and will be reduced by the amount of the discharged debt. Therefore, the sequence of the steps of pass-through and attribute reduction determines whether petitioners here were deficient when they increased their bases by the discharged debt amount and deducted their losses.

The sequencing question is expressly addressed in the statute. Section 108(b)(4)(A) directs that the attribute reductions "shall be made after the determination of the tax imposed by this chapter for the taxable year of the discharge." (Emphases added.) See also §1017(a) (applying the same sequencing when §108 attribute reduction affects basis of corporate property). In order to determine the "tax imposed," an S corporation shareholder must adjust his basis in his corporate stock and pass through all items of income and loss. See §§1366, 1367 (1994 ed. and Supp III). Consequently, the attribute reduction must be made after the basis adjustment and pass-through. In the case of petitioners, they must pass through the discharged debt, increase corporate bases, and then deduct their losses, all before any attribute reduction could occur. Because their basis increase is equal to their losses, petitioners have no suspended losses remaining. They, therefore, have no net operating losses to reduce.

Although the Commissioner has now abandoned the reasoning of the Court of Appeals below,8 we address the primary arguments made in the Courts of Appeals against petitioners' reading of the sequencing provision. First, one court has expressed the concern that, if the discharge of indebtedness is passed through to the shareholder before the tax attributes are reduced, then there can never be any discharge of indebtedness remaining "at [the] corporate level," §108(d)(7)(A), by which to reduce tax attributes.9 Gaudiano, 216 F.3d, at 533. This concern presumes that tax attributes can be reduced only if the discharge of indebtedness itself remains at the corporate level. The statute, however, does not impose this restriction. Section 108(b)(1) requires only that the tax attributes be reduced by "[t]he amount excluded from gross income," (emphasis added), and that amount is not altered by the mere pass-through of the income to the shareholder.

Second, courts have discussed the policy concern that, if shareholders were permitted to pass through the discharge of indebtedness before reducing any tax attributes, the shareholders would wrongly experience a "double windfall": They would be exempted from paying taxes on the full amount of the discharge of indebtedness, and they would be able to increase basis and deduct their previously suspended losses. See, e.g., 182 F.3d, at 1147_1148. Because the Code's plain text permits the taxpayers here to receive these benefits, we need not address this policy concern.10

* * *

The judgment of the Court of Appeals, accordingly, is reversed.

It is so ordered.

Justice Breyer, dissenting.

I agree with the majority's reasoning with the exception of footnotes 6 and 10. The basic statutory provision before us is 26 U.S.C. § 108-the provision that excludes from the "gross income" of any "insolvent" taxpayer, income that cancellation of a debt (COD) would otherwise generate. As the majority acknowledges, however, ante, at 7, n. 6, §108 contains a subsection that sets forth a special exception. The exception, entitled "Special rules for S corporation," says:

"(A) Certain provisions to be applied at corporate level.

"In the case of an S corporation, subsections (a), (b), (c), and (g) shall be applied at the corporate level." 26 U.S.C. § 108(d)(7)(A).

If one reads this language literally as exclusive, both the COD exclusion (§108(a)) and the tax attribute reduction (§108(b)) would apply only "at the corporate level." Hence the COD income would not flow through to S corporation shareholders. Consequently, the insolvent S corporation's COD income would not increase the shareholder's basis and would not help the shareholder take otherwise unavailable deductions for suspended losses.

The Commissioner argues that we should read the language in this way as preventing the flow-through of the corporation's COD income. Brief for United States 27. He points to the language of a House Committee, which apparently thought, when Congress passed an amendment to §108, that the Commissioner's reading is correct. H. R. Rep. No. 103_111, pp. 624_625 (1993) ("[T]he exclusion and basis reduction are both made at the S corporation level (sec. 108(d)(7)). The shareholders' basis in their stock is not adjusted by the amount of debt discharge income that is excluded at the corporate level"). At least one commentator believes the same. See Loebl, Does the Excluded COD Income of an Insolvent S Corporation Increase the Basis of the Shareholders' Stock?, 52 U. Fla. L. Rev. 957, 981_988 (2000). But see Lockhart & Duffy, Tax Court Rules in Nelson that S Corporation Excluded COD Income Does Not Increase Shareholder Stock Basis, 25 Wm. Mitchell L. Rev. 287 (1999).

The Commissioner finds support for his literal, exclusive reading of §108(d)(7)(A)'s language in the fact that his reading would close a significant tax loophole. That loophole-preserved by the majority-would grant a solvent shareholder of an insolvent S corporation a tax benefit in the form of permission to take an otherwise unavailable deduction, thereby sheltering other, unrelated income from tax. See Witzel v. Commissioner, 200 F.3d 496, 497 (CA7 2000) (Posner, C. J.) ("It is hard to understand the rationale for using a tax exemption to avoid taxation not only on the income covered by the exemption but also on unrelated income that is not tax exempt"). Moreover, the benefit often would increase in value as the amount of COD income increases, a result inconsistent with congressional intent to impose a "price" (attribute reduction), see Lipton, Different Courts Adopt Different Approaches to the Impact of COD Income on S Corporations, 92 J. Tax. 207 (2000), on excluded COD. Further, this deduction-related tax benefit would have very different tax consequences for identically situated taxpayers, depending only upon whether a single debt can be split into segments, each of which is canceled in a different year. For example, under the majority's interpretation, a $1 million debt canceled in one year would permit Taxpayer A to deduct $1 million of suspended losses in that year, thereby permitting A to shelter $1 million of unrelated income in that year. But because §108 reduces tax attributes after the first year, five annual cancellations of $200,000 will not create a $1 million shelter. Timing is all important.

The majority acknowledges some of these policy concerns and confesses ignorance of any "other instance in which §108 directly benefits a solvent entity," but claims that its reading is mandated by the plain text of §108(d)(7)(A) and therefore that the Court may disregard the policy consequences. Ante, at 13, and n. 10. It is difficult, however, to see why we should interpret that language as treating different solvent shareholders differently, given that the words "at the corporate level" were added "[i]n order to treat all shareholders in the same manner." H. R. Rep. No. 98_432, pt. 2, p. 1640 (1984). And it is more difficult to see why, given the fact that the "plain language" admits either interpretation, we should ignore the policy consequences. See Commissioner v. Gillette Motor Transport, Inc., 364 U.S. 130, 134_135 (1960) (abandoning literal meaning of 26 U.S.C. § 1221 (1958 ed.) for a reading more consistent with congressional intent). Accord, Commissioner v. P. G. Lake, Inc., 356 U.S. 260, 264_267 (1958); Corn Products Refining Co. v. Commissioner, 350 U.S. 46, 51_52 (1955); Hort v. Commissioner, 313 U.S. 28, 30_31 (1941).

The arguments from plain text on both sides here produce ambiguity, not certainty. And other things being equal, we should read ambiguous statutes as closing, not maintaining, tax loopholes. Such is an appropriate understanding of Congress' likely intent. Here, other things are equal, for, as far as I am aware, the Commissioner's literal interpretation of §108(d)(7)(A) as exclusive would neither cause any tax-related harm nor create any statutory anomaly. Petitioners argue that it would create a linguistic inconsistency, for they point to a Treasury Regulation that says that the Commissioner will apply hobby loss limitations under §183 "at the corporate level in determining" allowable deductions, while, presumably, nonetheless permitting the deduction so limited to flow through to the shareholder. Treas. Reg. §1.183_1(f), 26 CFR § 1.183_1(f) (2000). But we are concerned here with the "application" of an exclusion, not with "determining" the amount of a deduction. Regardless, the regulation's use of the words "at the corporate level," like the three other appearances of the formulation "applied" or "determined" "at the corporate level" in the Code, occur in contexts that are so very different from this one that nothing we say here need affect their interpretation. See 26 U.S.C. § 49(a)(1)(E)(ii)(I) (determining whether financing is recourse financing); 26 U.S.C. § 264(f)(5)(B) (1994 ed., Supp. IV) (determining how to allocate interest expense to portions of insurance policies); 26 U.S.C. § 302(e)(1)(A) (determining whether a stock distribution shall be treated as a partial liquidation). If there are other arguments militating in favor of the majority's interpretation, I have not found them.

The majority, in footnote 6, says that the words "at the corporate level" in §108(d)(7)(A) apply to the exclusion of COD income from corporate income and to "tax attribute reduction" but do not "suspen[d] the operation of _ ordinary pass-through rules" because §108(d)(7)(A) "does not state or imply that the debt discharge provisions shall apply only `at the corporate level.' " It is the majority, however, that should explain why it reads the provision as nonexclusive (where, as here, its interpretation of the Code results in the "practical equivalent of [a] double deduction," Charles Ilfeld Co. v. Hernandez, 292 U.S. 62, 68 (1934)). See United States v. Skelly Oil Co., 394 U.S. 678, 684 (1969) (requiring "clear declaration of intent by Congress" in such circumstances). I do not contend that §108(d)(7)(A) must be read as having exclusive effect, only that, given the alternative, this interpretation provides the best reading of §108 as a whole. And I can find no "clear declaration of intent by Congress" to support the majority's contrary conclusion regarding §108(d)(7)(A)'s effect. It is that conclusion from which, for the reasons stated, I respectfully dissent

Petitioner, William Fiore, was convicted of violating a Pennsylvania statute prohibiting the operation of a hazardous waste facility without a permit. After Fiore's conviction became final, the Pennsylvania Supreme Court interpreted the statute for the first time, and made clear that Fiore's conduct was not within its scope. However, the Pennsylvania courts refused to grant Fiore collateral relief. We granted certiorari in part to decide when, or whether, the Federal Due Process Clause requires a State to apply a new interpretation of a state criminal statute retroactively to cases on collateral review.

In order to determine if that question was in fact presented, we asked the Pennsylvania Supreme Court whether its decision interpreting the statute not to apply to conduct like Fiore's was a new interpretation, or whether it was, instead, a correct statement of the law when Fiore's conviction became final. The Pennsylvania Supreme Court, responding to our certified question, has now made clear that retroactivity is not at issue. At the same time, that court's interpretation of its statute makes clear that Fiore did not violate the statute. We consequently find that his conviction is not consistent with the demands of the Federal Due Process Clause. See Jackson v. Virginia, 443 U.S. 307, 316 (1979).

* This case, previously described in greater detail in our opinion certifying the state-law question to the Pennsylvania Supreme Court, 528 U.S. 23 (1999), arises out of William Fiore's conviction under a Pennsylvania statute that prohibits "operat[ing] a hazardous waste" facility without a "permit." Pa. Stat. Ann., Tit. 35, §6018.401(a) (Purdon 1993); see Commonwealth v. Fiore, CC No. 8508740 (Ct. Common Pleas, Allegheny Cty., Pa., Jan. 19, 1988), App. 6. The Commonwealth conceded that Fiore in fact had a permit, but argued that Fiore had deviated so dramatically from the permit's terms that he nonetheless had violated the statute. And the Commonwealth's lower courts agreed. See id., at App. 43_44; Commonwealth v. Fiore, No. 00485 PGH 1988 (May 12, 1989), App. 99_100 (affirming Fiore's conviction on the trial court's reasoning).

The Pennsylvania Supreme Court declined to review Fiore's case, Commonwealth v. Fiore, 525 Pa. 577, 575 A. 2d 109 (1990), and his conviction became final. Thereafter, the Pennsylvania Supreme Court agreed to review the conviction of Fiore's co-defendant, David Scarpone, convicted of the same crime at the same time. The Supreme Court reversed Scarpone's conviction on the ground that the statute meant what it said: The statute made it unlawful to operate a facility without a permit; one who deviated from his permit's terms was not a person without a permit; hence, a person who deviated from his permit's terms did not violate the statute. Commonwealth v. Scarpone, 535 Pa. 273, 279, 634 A. 2d 1109, 1112 (1993) (describing the Commonwealth's interpretation as "a bald fiction we cannot endorse").

Fiore, unsuccessful in his subsequent state-court attempts to have his own conviction set aside, see Commonwealth v. Fiore, 445 Pa. Super. 401, 665 A. 2d 1185 (1995), appeal denied, Commonwealth v. Fiore, 544 Pa. 623, 675 A. 2d 1243 (1996), brought a federal habeas corpus action. The District Court granted the writ, but the Court of Appeals for the Third Circuit reversed. 149 F.3d 221 (1998). The Court of Appeals believed that the Pennsylvania Supreme Court, in Scarpone's case, had announced a new rule of law, inapplicable to Fiore's already final conviction. Id., at 227. And, the Court of Appeals said, "state courts are under no [federal] constitutional obligation to apply their decisions retroactively." Id., at 222. We granted certiorari to determine whether Fiore's conviction was inconsistent with the Due Process Clause.

Because we were uncertain whether the Pennsylvania Supreme Court's decision in Scarpone's case represented a change in the law of Pennsylvania, we certified the following question to that court:

"Does the interpretation of Pa. Stat. Ann., Tit. 35, §6018.401(a) (Purdon 1993), set forth in Commonwealth v. Scarpone, 535 Pa. 273, 279, 634 A. 2d 1109, 1112 (1993), state the correct interpretation of the law of Pennsylvania at the date Fiore's conviction became final?" 528 U.S., at 29.

"Scarpone did not announce a new rule of law. Our ruling merely clarified the plain language of the statute. . . . Our interpretation of [§6018.401(a)] in Scarpone furnishes the proper statement of law at the date Fiore's conviction became final." Fiore v. White, 562 Pa. 634, 646, 757 A. 2d 842, 848_849 (2000) (citation omitted).

The Pennsylvania Supreme Court's reply specifies that the interpretation of §6018.401(a) set out in Scarpone "merely clarified" the statute and was the law of Pennsylvania-as properly interpreted-at the time of Fiore's conviction. Because Scarpone was not new law, this case presents no issue of retroactivity. Rather, the question is simply whether Pennsylvania can, consistently with the Federal Due Process Clause, convict Fiore for conduct that its criminal statute, as properly interpreted, does not prohibit.

This Court's precedents make clear that Fiore's conviction and continued incarceration on this charge violate due process. We have held that the Due Process Clause of the Fourteenth Amendment forbids a State to convict a person of a crime without proving the elements of that crime beyond a reasonable doubt. See Jackson, supra, at 316; In re Winship, 397 U.S. 358, 364 (1970). In this case, failure to possess a permit is a basic element of the crime of which Fiore was convicted. Scarpone, supra, at 279, 634 A. 2d, at 1112. And the parties agree that the Commonwealth presented no evidence whatsoever to prove that basic element. To the contrary, the Commonwealth, conceding that Fiore did possess a permit, see Brief for Respondents 1, necessarily concedes that it did not prove he failed to possess one.

The simple, inevitable conclusion is that Fiore's conviction fails to satisfy the Federal Constitution's demands. We therefore reverse the contrary judgment of the Third Circuit and remand this case for proceedings consistent with this opinion.

So ordered.

Congress has provided, in 18 U.S.C. § 3621(e)(2)(B), that the Bureau of Prisons (Bureau or BOP) may reduce by up to one year the prison term of an inmate convicted of a nonviolent felony, if the prisoner successfully completes a substance abuse program. The Bureau's implementing regulation categorically denies early release to prisoners whose current offense is a felony attended by "the carrying, possession, or use of a firearm." 28 CFR § 550.58(a)(1)(vi)(B) (2000). The validity of the Bureau's regulation is the question presented in this case. We hold, in accord with the Court of Appeals for the Eighth Circuit, that the regulation is a permissible exercise of the Bureau's discretion under 18 U.S.C. § 3621(e)(2)(B).

* A

Title 18 U.S.C. § 3621 governs the imprisonment of persons convicted of federal crimes. In 1990, Congress amended the statute to provide that "[t]he Bureau shall _ make available appropriate substance abuse treatment for each prisoner the Bureau determines has a treatable condition of substance addiction or abuse." Pub. L. 101_647, §2903, 104 Stat. 4913. Four years later, Congress again amended §3621, this time to provide incentives for prisoner participation in BOP drug treatment programs. The incentive provision at issue reads: "The period a prisoner convicted of a nonviolent offense remains in custody after successfully completing a treatment program may be reduced by the Bureau of Prisons, but such reduction may not be more than one year from the term the prisoner must otherwise serve." Pub. L. 103_322, §32001, 108 Stat. 1897 (codified at 18 U.S.C. § 3621(e)(2)(B)).

In 1995, the Bureau published a rule to implement the early release incentive. 60 Fed. Reg. 27692_27695; 28 CFR § 550.58. Because the statute explicitly confined the incentive to prisoners convicted of "nonviolent offense[s]," 18 U.S.C. § 3621(e)(2)(B), the BOP ranked ineligible for early release all inmates currently incarcerated for "crime[s] of violence," 60 Fed. Reg. 27692. As explained in the Bureau's program statement, the BOP defined "crimes of violence" to include a drug trafficking conviction under 21 U.S.C. § 841 if the offender received a two-level sentence enhancement under United States Sentencing Commission, Guidelines Manual (USSG) §2D1.1(b)(1), for possessing a dangerous weapon during commission of the drug offense. Bureau of Prisons Program Statement No. 5162.02, §9 (July 24, 1995), reprinted in App. to Brief for Petitioner 17_18.1 "[E]xercising [its] discretion in reducing a sentence," the Bureau also excluded from early release eligibility inmates who had a prior conviction "for homicide, forcible rape, robbery, or aggravated assault." 60 Fed. Reg. 27692 (codified at 28 CFR § 550.58 (1995)).

The Courts of Appeals divided over the validity of the Bureau's definition of crimes of violence to include drug offenses that involved possession of a firearm. A majority of Circuits, including the Eighth, held that §3621(e)(2)(B) required the Bureau to look only to the offense of conviction (drug trafficking), and not to sentencing factors (firearm possession), in determining whether an offender was convicted of a "nonviolent offense," and was therefore eligible under the statute for the early release incentive. Martin v. Gerlinski, 133 F.3d 1076, 1079 (CA8 1998); see also Fristoe v. Thompson, 144 F.3d 627, 631 (CA10 1998); Byrd v. Hasty, 142 F.3d 1395, 1398 (CA11 1998); Roussos v. Menifee, 122 F.3d 159, 164 (CA3 1997); Downey v. Crabtree, 100 F.3d 662, 668 (CA9 1996). The Fourth and Fifth Circuits, however, upheld the Bureau's classification of drug offenses attended by firearm possession as violent crimes. Pelissero v. Thompson, 170 F.3d 442, 447 (CA4 1999); Venegas v. Henman, 126 F.3d 760, 763 (CA5 1997).

This split among the Circuits prompted the Bureau in 1997 to publish the regulation now before the Court. See 62 Fed. Reg. 53690_53691. Like the 1995 rule, the current regulation excludes from early release eligibility offenders who possessed a firearm in connection with their offenses. In contrast to the earlier rule, however, the 1997 regulation does not order this exclusion by defining the statutory term "prisoner convicted of a nonviolent offense" or the cognate term "crimes of violence." Instead, the current regulation relies upon "the discretion allotted to the Director of the Bureau of Prisons in granting a sentence reduction to exclude [enumerated categories of] inmates." Id., at 53690. The regulation, designed to achieve consistent administration of the incentive, now provides:

"(a) Additional early release criteria. (1) As an exercise of the discretion vested in the Director of the Federal Bureau of Prisons, the following categories of inmates are not eligible for early release:

. . . . .

"(iv) Inmates who have a prior felony or misdemeanor conviction for homicide, forcible rape, robbery, or aggravated assault, or child sexual abuse offenses;

. . . . .

"(vi) Inmates whose current offense is a felony:

. . . . .

"(B) That involved the carrying, possession, or use of a firearm or other dangerous weapon _ ." 28 CFR § 550.58(a)(2000).

In sum, the 1995 rule defined the statutory term "prisoner convicted of a nonviolent offense" to exclude categorically an inmate who possessed a firearm in connection with his offense. The current regulation categorically excludes such an inmate, not because §3621(e)(2)(B) so mandates, but pursuant to the Bureau's asserted discretion to prescribe additional early release criteria. Drug traffickers who possess firearms when they engage in crimes are no longer characterized as "violent" offenders within the meaning of the statute. But they are bracketed, for sentence reduction purposes, with persons currently incarcerated for "nonviolent offense[s]" who in the past committed crimes qualifying as violent. The preconviction conduct of both armed offenders and certain redicivists, in the Bureau's view, "suggest[s] that they pose a particular risk to the public." Brief for Respondents 30.

In 1997, petitioner Christopher A. Lopez was convicted of possession with intent to distribute methamphetamine, in violation of 21 U.S.C. § 841. Upon finding that Lopez possessed a firearm in connection with his offense, the District Court enhanced his sentence by two levels pursuant to USSG §2D1.1(b)(1). Lopez is currently scheduled to be released from prison in June 2002.

While incarcerated, Lopez requested substance abuse treatment. The Bureau found him qualified for its residential drug abuse program,2 but categorically ineligible, under 28 CFR § 550.58(a)(1)(vi), for early release. App. 3_7.

When notified that he would not be a candidate for early release, Lopez challenged the BOP's determination by filing a petition for a writ of habeas corpus, under 28 U.S.C. § 2241 in the United States District Court for the District of South Dakota. The District Court granted the petition. In that court's view, the Bureau's 1997 regulation did not correct the infirmity the Eighth Circuit saw in the 1995 rule. See App. 17_18, and n. 4 (citing Martin, 133 F.3d, at 1079). "[I]t is true," the District Court recognized, "that the BOP may exercise a great deal of discretion in determining who among the eligible nonviolent offenders may be released." App. 17. But, the District Court held, the BOP may not categorically count out, "based upon sentencing factors or weapon possession," inmates whose underlying conviction was for a nonviolent crime. Id., at 18. Accordingly, the District Court ordered the BOP "to reconsider Lopez's eligibility for early release." Id., at 19.

The Eighth Circuit reversed. Bellis v. Davis, 186 F.3d 1092 (1999). Section 3621(e)(2)(B), the Court of Appeals observed, "states only that the prison term of an inmate convicted of a nonviolent offense `may be reduced by the Bureau of Prisons.' " Id., at 1094 (quoting 18 U.S.C. § 3621(e)(2)(B)). This discretionary formulation, the Eighth Circuit reasoned, allows the Bureau to devise a regime based on criteria that can be uniformly applied. The statute grants no entitlement to any inmate or class of inmates, the Court of Appeals noted, and it does not instruct the Bureau to make "individual, rather than categorical, assessments of eligibility for inmates convicted of nonviolent offenses." 186 F.3d, at 1094. The court further reasoned that, to the extent Congress left a gap in §3621(e)(2)(B) for the Bureau to fill, deference is owed the BOP's interpretation under Chevron U.S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843_845, 866 (1984), so long as the interpretation is a permissible construction of the statute. 186 F.3d, at 1095. The Bureau had elected to deny early release to certain categories of prisoners, notably recidivists and firearms carriers, whose "conduct indicates that they pose a serious risk to public safety." Ibid. That decision, the Court of Appeals concluded, "represents a manifestly permissible construction of the statute and an appropriate exercise of the BOP's discretion." Ibid.

The Eighth Circuit next explained why its earlier decision in Martin did not control this case, which trains on the BOP's 1997 regulation: Martin addressed only the Bureau's 1995 attempt to interpret the statutory term "nonviolent offense"; the court in that case did not address "whether the BOP may, as an exercise of its discretion, _ look to sentencing factors in deciding which individuals among statutorily eligible inmates are appropriate candidates for early release." 186 F.3d, at 1095. Facing that issue, the Court of Appeals held such an exercise of discretion proper. Ibid.

The Courts of Appeals have again divided, now over the permissibility of the Bureau's current (1997) regulation. The Tenth and Eleventh Circuits, in line with their prior decisions invalidating the 1995 rule, have concluded that §3621(e)(2)(B) permits no categorical exclusions of nonviolent offenders based on sentence enhancements. Ward v. Booker, 202 F.3d 1249, 1256_1257 (CA10 2000); Kilpatrick v. Houston, 197 F.3d 1134, 1135 (CA11 1999). The Ninth Circuit, on the other hand, has agreed with the Eighth Circuit that precedent invalidating the 1995 rule does not control and that, in 1997, the BOP permissibly exercised its discretion under §3621(e)(2)(B) when it categorically excluded from early release consideration inmates who possessed a firearm in connection with their nonviolent offenses. Bowen v. Hood, 202 F.3d 1211, 1218_1220 (2000).

We granted certiorari to resolve this conflict, 529 U.S. 1086 (2000), and now affirm the judgment of the Eighth Circuit.

The statute provides: "The period a prisoner convicted of a nonviolent offense remains in custody after successfully completing a treatment program may be reduced by the Bureau of Prisons _ ." 18 U.S.C. § 3621(e)(2)(B). The measure thus categorically denies early release eligibility to inmates convicted of violent offenses. The question we address is whether the Bureau has discretion to delineate, as an additional category of ineligible inmates, those whose current offense is a felony involving a firearm. 28 CFR § 550.58(a)(1)(vi)(B) (2000).

Lopez urges that the statute is unambiguous. He says that, by identifying a class of inmates ineligible for sentence reductions under §3621(e)(2)(B), i.e., those convicted of a violent offense, Congress has barred the Bureau from identifying further categories of ineligible inmates. "If Congress wanted the BOP to reduce the categories of inmates eligible for the early release incentive (beyond the one identified by Congress), Congress would have specifically placed this grant of authority in the language of the statute." Brief for Petitioner 23. As to the statutory instruction that the Bureau "may" reduce sentences, Lopez initially suggests it is merely a grant of authority to the BOP to reduce a sentence that, prior to the enactment of §3621(e)(2)(B), could not be reduced for successful completion of drug treatment: "The power granted was to give reductions not the power to decide who was eligible to receive reductions." Id., at 21. He alternately contends that the Bureau may take into account only "post-conviction conduct," not "pre-conviction conduct." Reply Brief 4_5. Acting on a case-by-case basis, Lopez asserts, the Bureau may "deny early release to those inmates [who] are statutorily eligible, but who do not deserve early release based on their conduct while in prison." Id., at 5. Under this reading, the Bureau may exercise discretion in denying early release, but only on an individual basis, taking account solely of postconviction conduct.

In the Bureau's view, §3621(e)(2)(B) establishes two prerequisites for sentence reduction: conviction of a nonviolent offense and successful completion of drug treatment. Brief for Respondents 18. If those prerequisites are met, the Bureau "may," but also may not, grant early release. The BOP opposes Lopez's argument that Congress barred the Bureau from imposing limitations categorically or on the basis of preconviction conduct. According to the Bureau, Congress simply "did not address how the Bureau should exercise its discretion within the class of inmates who satisfy the statutory prerequisites for early release." Id., at 23. Because Congress left the question unaddressed, the Bureau maintains, the agency may exclude inmates either categorically or on a case-by-case basis, subject of course to its obligation to interpret the statute reasonably, see Chevron, 467 U.S., at 844, in a manner that is not arbitrary or capricious, see 5 U.S.C. § 706(2)(A). In this instance, the Bureau urges, it has acted reasonably: Its denial of early release to all inmates who possessed a firearm in connection with their current offense rationally reflects the view that such inmates displayed a readiness to endanger another's life; accordingly, in the interest of public safety, they should not be released months in advance of completing their sentences.3

We agree with the Bureau's position. Preliminarily, we note conspicuous anomalies in Lopez's construction. If §3621(e)(2)(B) functions not as a grant of discretion to determine early release eligibility, but both as an authorization and a command to reduce sentences, then Congress' use of the word "may," rather than "shall," has no significance. And if the BOP does have discretion to deny early release to certain inmates, but only based on individualized assessments of postconviction conduct, then the agency cannot categorically deny early release even to recidivists with prior (perhaps multiple) convictions for "homicide, forcible rape _ , or child sexual abuse offenses." 28 CFR § 550.58(a)(1)(iv) (2000). For that provision, as much as the exclusion of inmates imprisoned for offenses involving a firearm, see supra, at 4, entails no individualized determination based on postconviction conduct. Furthermore, Lopez's position would confine the BOP's discretion under §3621(e)(2)(B) to consideration of factors of the kind the Bureau already may consider in granting credit for "satisfactory behavior." See 18 U.S.C. § 3624(b)(1) ("a prisoner [serving a term of more than one year and less than life] may receive credit toward the service of the prisoner's sentence _ subject to determination by the Bureau of Prisons that, during that year, the prisoner has displayed exemplary compliance with such institutional disciplinary regulations").

We turn now to the Bureau's reading of the statutory text, which instructs that the agency "may" reduce the sentence of a nonviolent offender who has successfully completed a drug treatment program. Congress' use of the permissive "may" in §3621(e)(2)(B) contrasts with the legislators' use of a mandatory "shall" in the very same section. Elsewhere in §3621, Congress used "shall" to impose discretionless obligations, including the obligation to provide drug treatment when funds are available. See 18 U.S.C. § 3621(e)(1) ("Bureau of Prisons shall, subject to the availability of appropriations, provide residential substance abuse treatment (and make arrangements for appropriate aftercare)"); see also, e.g., §3621(b) ("The Bureau shall designate the place of the prisoner's imprisonment_ . In designating the place of imprisonment or making transfers under this subsection, there shall be no favoritism given to prisoners of high social or economic status."). Sensibly read, the grant of discretion in §3621(e)(2)(B) to decide whether to reduce a sentence parallels the grant of discretion in §3621(e)(2)(A) to retain a prisoner who successfully completes drug treatment "under such [custodial] conditions as the Bureau deems appropriate." §3621(e)(2)(A). When an eligible prisoner successfully completes drug treatment, the Bureau thus has the authority, but not the duty, both to alter the prisoner's conditions of confinement and to reduce his term of imprisonment.

The constraints Lopez urges-requiring the BOP to make individualized determinations based only on postconviction conduct-are nowhere to be found in §3621(e)(2)(B). Beyond instructing that the Bureau has discretion to reduce the period of imprisonment for a nonviolent offender who successfully completes drug treatment, Congress has not identified any further circumstance in which the Bureau either must grant the reduction, or is forbidden to do so. In this familiar situation, where Congress has enacted a law that does not answer "the precise question at issue," all we must decide is whether the Bureau, the agency empowered to administer the early release program, has filled the statutory gap "in a way that is reasonable in light of the legislature's revealed design." NationsBank of N. C., N. A. v. Variable Annuity Life Ins. Co., 513 U.S. 251, 257 (1995) (citing Chevron, 467 U.S., at 842); see also Reno v. Koray, 515 U.S. 50, 61 (1995) (deferring to BOP's interpretation of statute). We think the agency's interpretation is reasonable both in taking account of preconviction conduct and in making categorical exclusions.

First, as the dissent but not Lopez recognizes, see post, at 4, the Bureau need not blind itself to preconviction conduct that the agency reasonably views as jeopardizing life and limb. By denying eligibility to violent offenders, the statute manifests congressional concern for preconviction behavior-and for the very conduct leading to conviction. The Bureau may reasonably attend to these factors as well. Its regulation in this regard is kin to the Attorney General's order upheld in INS v. Yueh-Shaio Yang, 519 U.S. 26 (1996). That case involved a statute authorizing the Attorney General to waive deportation of aliens deportable for entry fraud. The Attorney General had refused to waive deportation for one alien because of "acts of fraud _ in connection with his entry." Id., at 27. The alien argued that because the statute made aliens who had committed entry fraud eligible for waiver, the Attorney General was precluded from taking such conduct into account "at all" in deciding whether to grant relief. Id., at 30. We rejected this view, stating that the statute "establishes only the alien's eligibility for the waiver. Such eligibility in no way limits the considerations that may guide the Attorney General in exercising her discretion to determine who, among those eligible, will be accorded grace." Id., at 31. Similarly in this case, the statute's restriction of early release eligibility to nonviolent offenders does not cut short the considerations that may guide the Bureau. Just as the Attorney General permissibly considered aspects of entry fraud, even though entry fraud was a criterion of statutory eligibility, so the Bureau may consider aspects of the conduct of conviction, even though the conviction is a criterion of statutory eligibility.4

We also reject Lopez's argument, echoed in part by the dissent, post, at 5, that the agency must not make categorical exclusions, but may rely only on case-by-case assessments.5 "[E]ven if a statutory scheme requires individualized determinations," which this scheme does not, "the decisionmaker has the authority to rely on rulemaking to resolve certain issues of general applicability unless Congress clearly expresses an intent to withhold that authority." American Hospital Assn. v. NLRB, 499 U.S. 606, 612 (1991); accord, Heckler v. Campbell, 461 U.S. 458, 467 (1983). The approach pressed by Lopez-case-by-case decisionmaking in thousands of cases each year, see supra, 12, n. 4-could invite favoritism, disunity, and inconsistency. The Bureau is not required continually to revisit "issues that may be established fairly and efficiently in a single rulemaking proceeding." Heckler, 461 U.S., at 467.6

Having decided that the Bureau may categorically exclude prisoners based on their preconviction conduct, we further hold that the regulation excluding Lopez is permissible. The Bureau reasonably concluded that an inmate's prior involvement with firearms, in connection with the commission of a felony, suggests his readiness to resort to life-endangering violence and therefore appropriately determines the early release decision.7

For the reasons stated, the judgment of the Court of Appeals for the Eighth Circuit is

Affirmed.

Justice Stevens, with whom The Chief Justice and Justice Kennedy join, dissenting.

The question at issue in this case is whether all, or merely some, of the federal prisoners who were convicted of nonviolent offenses and who have successfully completed a Bureau of Prisons (BOP) drug treatment program are eligible for a sentence reduction pursuant to 18 U.S.C. § 3621(e)(2)(B). For the reasons outlined below, I believe that Congress has answered that precise question. The statute expressly states that the sentence of every prisoner in that category "may be reduced." Ibid. The disposition of this case is therefore governed by the first step in the familiar test announced in Chevron U.S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842_843 (1984), for "Congress has directly spoken to the precise question at issue." Id., at 842.

* In drafting the statute in question, Congress was faced with a difficult policy choice: whether the commission of particular crimes made certain categories of offenders so dangerous that the costs of offering them early release in return for the successful completion of a drug treatment program outweighed the rewards. The initial drafts of the bill answered that question in the negative and made all federal prisoners eligible for a sentence reduction of up to one year if they successfully completed a drug treatment program. See, e.g., H. R. Rep. No. 103_320, p. 2 (1993). However, the inclusion of those convicted of violent offenses within the category of those eligible for the inducement soon became a fulcrum of criticism for the larger crime bill within which the statute was embedded.1 Perhaps as a result of these criticisms,2 the statute ultimately adopted limited the inducement to "pris- oner[s] convicted of _ nonviolent offense[s]." 18 U.S.C. § 3621(e)(2)(B).

Both the text of the statute and the aforementioned history demonstrate that Congress directly addressed the "precise question" of what offenses ought to disqualify prisoners from eligibility for a sentence reduction, and that its unambiguous answer was "violent offenses." Under the statute as enacted, those who commit crimes of violence are categorically barred from receiving a sentence reduction while those convicted of nonviolent offenses "may" receive such an inducement.

The BOP regulation challenged here operates to redefine the set of prisoners categorically ineligible for a sentence reduction, a set unambiguously defined in the text of the statute. It does so by taking a group of prisoners whose offenses the Bureau acknowledges are "nonviolent" within the meaning of the statute3 and imposing the same sanction-categorical ineligibility-upon them as the statute imposes upon violent offenders. In so doing, the Bureau ignores Congress's express determination that, when evaluating eligibility for a sentence reduction, the salient distinction is the line between violent and nonviolent offenses. By moving this line, the BOP exceeded its authority and sought to exercise its discretion on an issue with regard to which it has none. See, e.g., Chevron, 467 U.S., at 842_843 ("First, always, is the question whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress"); United States v. Haggar Apparel Co., 526 U.S. 380, 392 (1999) ("In the process of considering a regulation in relation to specific factual situations, a court may conclude the regulation is inconsistent with the statutory language or is an unreasonable implementation of it. In those instances, the regulation will not control").

I fully agree with the majority that federal prisoners do not become entitled to a sentence reduction upon their successful completion of a drug treatment program; the words "may be reduced" do not mean "shall be reduced." Nonetheless, while the statute does not entitle any prisoner to a sentence reduction, it does guarantee nonviolent offenders who successfully complete a drug treatment program consideration for such a reduction.

For every nonviolent offender who participates in a drug treatment program, the BOP may be required to make two individualized determinations: (1) whether he or she has successfully completed that program; and (2), if so, whether his or her preconviction conduct, postconviction conduct, and prospects for rehabilitation justify a sentence reduction. In evaluating whether or not a particular individual is entitled to a sentence reduction, the BOP may give great weight to whichever of these factors it determines to be most relevant. That, however, is a far cry from categorically excluding from consideration prisoners who Congress explicitly intended to obtain such consideration.4

The majority's concern about the risks and burdens associated with case-by-case decisionmaking in a large number of cases is understandable yet ultimately misguided. In order to fulfill the statute's requirements, the BOP must already evaluate every prisoner seeking the sentence reduction on an individual basis to determine whether that prisoner "successfully completed" his or her drug treatment program. Individualized consideration of the second salient question involves consideration of many of the same personalized factors that go into determining whether a prisoner's course of drug treatment has been "successful." To the extent that answering the second question requires consideration of additional factors with a concomitant administrative burden, the costs of such a scheme are, in Congress's judgment, outweighed by the benefits of encouraging drug treatment and of carefully distinguishing between those prisoners who have earned an early return to their communities and those who require further incarceration.

The majority's worry that individualized decisionmaking might lead to "favoritism, disunity, and inconsistency" is similarly misplaced. Ante, at 13. To suggest that decisionmaking must be individualized is not to imply that it must also be standardless. If the Court today invalidated the regulation in question, its decision would not preclude the BOP from adopting a uniform set of criteria for consideration in evaluating applications for sentence reductions. Nor would it necessarily preclude the Bureau from giving dispositive weight to certain postconviction criteria or near-dispositive weight to preconviction criteria. Cf. Heckler v. Campbell, 461 U.S. 458, 467 (1983). The Bureau would remain free to structure its decisionmaking in any way it saw fit as long as in so doing it did not contravene policy decisions explicitly made by the statute's drafters. As Congress has already addressed preincarceration conduct in §3621(e)(2)(B), the Bureau may not categorically exclude a prisoner not convicted of a violent offense from consideration for early release on the basis of such conduct without exceeding the limits of its discretion.

Accordingly, I respectfully dissent.

Washington State's Community Protection Act of 1990 authorizes the civil commitment of "sexually violent predators," persons who suffer from a mental abnormality or personality disorder that makes them likely to engage in predatory acts of sexual violence. Wash. Rev. Code §71.09.010 et seq. (1992). Respondent, Andre Brigham Young, is confined as a sexually violent predator at the Special Commitment Center (Center), for which petitioner is the superintendent. After respondent's challenges to his commitment in state court proved largely unsuccessful, he instituted a habeas action under 28 U.S.C. § 2254 seeking release from confinement. The Washington Supreme Court had already held that the Act is civil, In re Young, 122 Wash. 2d 1, 857 P.2d 989 (1993) (en banc), and this Court held a similar commitment scheme for sexually violent predators in Kansas to be civil on its face, Kansas v. Hendricks, 521 U.S. 346 (1997). The Court of Appeals for the Ninth Circuit nevertheless concluded that respondent could challenge the statute as being punitive "as applied" to him in violation of the Double Jeopardy and Ex Post Facto Clauses, and remanded the case to the District Court for an evidentiary hearing.

* A

Washington State's Community Protection Act of 1990 (Act) was a response to citizens' concerns about laws and procedures regarding sexually violent offenders. One of the Act's provisions authorizes civil commitment of such offenders. Wash. Rev. Code §71.09.010 et seq. (1992 and Supp. 2000). The Act defines a sexually violent predator as someone who has been convicted of, or charged with, a crime of sexual violence and who suffers from a mental abnormality or personality disorder that makes the person likely to engage in predatory acts of sexual violence if not confined in a secure facility. §71.09.020(1) (Supp. 2000). The statute reaches prisoners, juveniles, persons found incompetent to stand trial, persons found not guilty by reason of insanity, and persons at any time convicted of a sexually violent offense who have committed a recent overt act. §71.09.030. Generally, when it appears that a person who has committed a sexually violent offense is about to be released from confinement, the prosecuting attorney files a petition alleging that that person is a sexually violent predator. Ibid. That filing triggers a process for charging and trying the person as a sexually violent predator, during which he is afforded a panoply of protections including counsel and experts (paid for by the State in cases of indigency), a probable cause hearing, and trial by judge or jury at the individual's option. §§71.09.040_71.09.050. At trial, the State bears the burden to prove beyond a reasonable doubt that the person is a sexually violent predator. §71.09.060(1).

Upon the finding that a person is a sexually violent predator, he is committed for control, care, and treatment to the custody of the department of social and health services. Ibid. Once confined, the person has a right to adequate care and individualized treatment. §71.09.080(2). The person is also entitled to an annual examination of his mental condition. §71.09.070. If that examination indicates that the individual's condition is so changed that he is not likely to engage in predatory acts of sexual violence, state officials must authorize the person to petition the court for conditional release or discharge. §71.09.090(1). The person is entitled to a hearing at which the State again bears the burden of proving beyond a reasonable doubt that he is not safe to be at large. Ibid. The person may also independently petition the court for release. §71.09.090(2). At a show cause hearing, if the court finds probable cause to believe that the person is no longer dangerous, a full hearing will be held at which the State again bears the burden of proof. Ibid.

The Act also provides a procedure to petition for conditional release to a less restrictive alternative to confinement. §71.09.090. Before ordering conditional release, the court must find that the person will be treated by a state certified sexual offender treatment provider, that there is a specific course of treatment, that housing exists that will be sufficiently secure to protect the community, and that the person is willing to comply with the treatment and supervision requirements. §71.09.092. Conditional release is subject to annual review until the person is unconditionally released. §§71.09.096, 71.09.098.

Respondent, Andre Brigham Young, was convicted of six rapes over three decades. App. to Pet. for Cert. 33a. Young was scheduled to be released from prison for his most recent conviction in October 1990. One day prior to his scheduled release, the State filed a petition to commit Young as a sexually violent predator. Id., at 32a.

At the commitment hearing, Young's mental health experts testified that there is no mental disorder that makes a person likely to reoffend and that there is no way to predict accurately who will reoffend. The State called an expert who testified, based upon a review of Young's records, that Young suffered from a severe personality disorder not otherwise specified with primarily paranoid and antisocial features, and a severe paraphilia, which would be classified as either paraphilia sexual sadism or paraphilia not otherwise specified (rape). See generally American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders 522_523, 530, 532, 634, 645_646, 673 (4th ed. 1994). In the state expert's opinion, severe paraphilia constituted a mental abnormality under the Act. The State's expert concluded that Young's condition, in combination with the personality disorder, the span of time during which Young committed his crimes, his recidivism, his persistent denial, and his lack of empathy or remorse, made it more likely than not that he would commit further sexually violent acts. The victims of Young's rapes also testified. The jury unanimously concluded that Young was a sexually violent predator.

Young and another individual appealed their commitments in state court, arguing that the Act violated the Double Jeopardy, Ex Post Facto, Due Process, and Equal Protection Clauses of the Federal Constitution. In major respects, the Washington Supreme Court held that the Act is constitutional. In re Young, 122 Wash. 2d 1, 857 P.2d 989 (1993) (en banc). To the extent the court concluded that the Act violated due process and equal protection principles, those rulings are reflected in subsequent amendments to the Act. See Part I_A, supra.

The Washington court reasoned that the claimants' double jeopardy and ex post facto claims hinged on whether the Act is civil or criminal in nature. Following this Court's precedents, the court examined the language of the Act, the legislative history, and the purpose and effect of the statutory scheme. The court found that the legislature clearly intended to create a civil scheme both in the statutory language and legislative history. The court then turned to examine whether the actual impact of the Act is civil or criminal. The Act, the court concluded, is concerned with treating committed persons for a current mental abnormality, and protecting society from the sexually violent acts associated with that abnormality, rather than being concerned with criminal culpability. The court distinguished the goals of incapacitation and treatment from the goal of punishment. The court found that the Washington Act is designed to further legitimate goals of civil confinement and that the claimants had failed to provide proof to the contrary. 122 Wash. 2d, at 18_25, 857 P.2d, at 996_1000.

The Act spawned several other challenges in state and federal court, two of which bear mention. Richard Turay, committed as a sexually violent predator, filed suit in Federal District Court against Center officials under Rev. Stat. §1979, 42 U.S.C. § 1983 alleging unconstitutional conditions of confinement and inadequate treatment at the Center. In 1994, a jury concluded that the Center had failed to provide constitutionally adequate mental health treatment. App. 64_68. The court ordered officials at the Center to bring the institution up to constitutional standards, appointing a Special Master to monitor progress at the Center. The Center currently operates under an injunction. Turay v. Seling, 108 F. Supp. 2d 1148 (WD Wash. 2000). See also Brief for Petitioner 8_9.

Turay also appealed his commitment as a sexually violent predator in state court, claiming, among other things, that the conditions of confinement at the Center rendered the Washington Act punitive "as applied" to him in violation of the Double Jeopardy Clause. The Washington Supreme Court ruled that Turay's commitment was valid. In re Turay, 139 Wash. 2d 379, 986 P.2d 790 (1999) (en banc). The court explained that in Young, it had concluded that the Act is civil. 139 Wash. 2d, at 415, 986 P.2d, at 869. The court also noted that this Court had recently held Kansas' Sexually Violent Predator Act, nearly identical to Washington's Act, to be civil on its face. Ibid. The Washington Supreme Court rejected Turay's theory of double jeopardy, reasoning that the double jeopardy claim must be resolved by asking whether the Act itself is civil. Id., at 416_417, 986 P.2d, at 810 (citing Hudson v. United States, 522 U.S. 93 (1997)). The court concluded that Turay's proper remedy for constitutional violations in conditions of confinement at the Center was his §1983 action for damages and injunctive relief. 139 Wash. 2d, at 420, 986 P.2d, at 812.

That brings us to the action before this Court. In 1994, after unsuccessful challenges to his confinement in state court, Young filed a habeas action under 28 U.S.C. § 2254 against the superintendent of the Center. Young contended that the Act was unconstitutional and that his confinement was illegal. He sought immediate release. The District Court granted the writ, concluding that the Act violated substantive due process, that the Act was criminal rather than civil, and that it violated the double jeopardy and ex post facto guarantees of the Constitution. Young v. Weston, 898 F. Supp. 744 (WD Wash. 1995). The superintendent appealed. While the appeal was pending, this Court decided Kansas v. Hendricks, 521 U.S. 346 (1997), which held that Kansas' Sexually Violent Predator Act, on its face, met substantive due process requirements, was nonpunitive, and thus did not violate the Double Jeopardy and Ex Post Facto Clauses. The Ninth Circuit Court of Appeals remanded Young's case to the District Court for reconsideration in light of Hendricks. 122 F.3d 38 (1997).

On remand, the District Court denied Young's petition. Young appealed and the Ninth Circuit reversed and remanded in part and affirmed in part. 192 F.3d 870 (1999). The Ninth Circuit affirmed the District Court's ruling that Young's confinement did not violate the substantive due process requirement that the State prove mental illness and dangerousness to justify confinement. Id., at 876. The Court of Appeals also left undisturbed the District Court's conclusion that the Act meets procedural due process and equal protection guarantees, and the District Court's rejection of Young's challenges to his commitment proceedings. Id., at 876_877. Young did not seek a petition for a writ of certiorari to the Ninth Circuit for its decision affirming the District Court in these respects, and accordingly, those issues are not before this Court.

The Ninth Circuit reversed the District Court's determination that because the Washington Act is civil, Young's double jeopardy and ex post facto claims must fail. The "linchpin" of Young's claims, the court reasoned, was whether the Act was punitive "as applied" to Young. Id., at 873. The court did not read this Court's decision in Hendricks to preclude the possibility that the Act could be punitive as applied. The court reasoned that actual conditions of confinement could divest a facially valid statute of its civil label upon a showing by the clearest proof that the statutory scheme is punitive in effect. 192 F.3d, at 874.

The Court of Appeals reviewed Young's claims that conditions of confinement at the Center were punitive and did not comport with due process. Id., at 875. Young alleged that for seven years, he had been subject to conditions more restrictive than those placed on true civil commitment detainees, and even state prisoners. The Center, located wholly within the perimeter of a larger Department of Corrections (DOC) facility, relied on the DOC for a host of essential services, including library services, medical care, food, and security. More recently, Young claimed, the role of the DOC had increased to include daily security "walk-throughs." Young contended that the conditions and restrictions at the Center were not reasonably related to a legitimate nonpunitive goal, as residents were abused, confined to their rooms, subjected to random searches of their rooms and units, and placed under excessive security.

Young also contended that conditions at the Center were incompatible with the Act's treatment purpose. The Center had a policy of videotaping therapy sessions and withholding privileges for refusal to submit to treatment. The Center residents were housed in units that, according to the Special Master in the Turay litigation, were clearly inappropriate for persons in a mental health treatment program. The Center still lacked certified sex offender treatment providers. Finally, there was no possibility of release. A court-appointed resident advocate and psychologist concluded in his final report that because the Center had not fundamentally changed over so many years, he had come to suspect that the Center was designed and managed to punish and confine individuals for life without any hope of release to a less restrictive setting. 192 F.3d, at 875. See also Amended Petition for Writ of Habeas Corpus, Supplemental Brief on Remand, and Motion to Alter Judgment 4_5, 8_9, 11_12, 15, 20, 24_26, in No. C94_480C (WD Wash.), Record, Doc. Nos. 57, 155, and 167.

The Ninth Circuit concluded that "[b]y alleging that [the Washington Act] is punitive as applied, Young alleged facts which, if proved, would entitle him to relief." 192 F.3d, at 875. The court remanded the case to the District Court for a hearing to determine whether the conditions at the Center rendered the Act punitive as applied to Young. Id., at 876.

This Court granted the petition for a writ of certiorari, 529 U.S. 1017 (2000), to resolve the conflict between the Ninth Circuit Court of Appeals and the Washington Supreme Court. Compare 192 F.3d 870 (1999), with In re Turay, 139 Wash. 2d 379, 986 P.2d 790 (1999).

As the Washington Supreme Court held and the Ninth Circuit acknowledged, we proceed on the understanding that the Washington Act is civil in nature. The Washington Act is strikingly similar to a commitment scheme we reviewed four Terms ago in Kansas v. Hendricks, 521 U.S. 346 (1997). In fact, Kansas patterned its Act after Washington's. See In re Hendricks, 259 Kan. 246, 249, 912 P.2d 129, 131 (1996). In Hendricks, we explained that the question whether an Act is civil or punitive in nature is initially one of statutory construction. 521 U.S., at 361 (citing Allen v. Illinois, 478 U.S. 364, 368 (1986)). A court must ascertain whether the legislature intended the statute to establish civil proceedings. A court will reject the legislature's manifest intent only where a party challenging the Act provides the clearest proof that the statutory scheme is so punitive in either purpose or effect as to negate the State's intention. 521 U.S., at 361 (citing United States v. Ward, 448 U.S. 242, 248_249 (1980)). We concluded that the confined individual in that case had failed to satisfy his burden with respect to the Kansas Act. We noted several factors: The Act did not implicate retribution or deterrence; prior criminal convictions were used as evidence in the commitment proceedings, but were not a prerequisite to confinement; the Act required no finding of scienter to commit a person; the Act was not intended to function as a deterrent; and although the procedural safeguards were similar to those in the criminal context, they did not alter the character of the scheme. 521 U.S., at 361_365.

We also examined the conditions of confinement provided by the Act. Id., at 363_364. The Court was aware that sexually violent predators in Kansas were to be held in a segregated unit within the prison system. Id., at 368. We explained that the Act called for confinement in a secure facility because the persons confined were dangerous to the community. Id., at 363. We noted, however, that conditions within the unit were essentially the same as conditions for other involuntarily committed persons in mental hospitals. Ibid. Moreover, confinement under the Act was not necessarily indefinite in duration. Id., at 364. Finally, we observed that in addition to protecting the public, the Act also provided treatment for sexually violent predators. Id., at 365_368. We acknowledged that not all mental conditions were treatable. For those individuals with untreatable conditions, however, we explained that there was no federal constitutional bar to their civil confinement, because the State had an interest in protecting the public from dangerous individuals with treatable as well as untreatable conditions. Id., at 366. Our conclusion that the Kansas Act was "nonpunitive thus remove[d] an essential prerequisite for both Hendricks' double jeopardy and ex post facto claims." Id., at 369.

Since deciding Hendricks, this Court has reaffirmed the principle that determining the civil or punitive nature of an Act must begin with reference to its text and legislative history. Hudson v. United States, 522 U.S. 93 (1997). In Hudson, which involved a double jeopardy challenge to monetary penalties and occupational debarment, this Court expressly disapproved of evaluating the civil nature of an Act by reference to the effect that Act has on a single individual. Instead, courts must evaluate the question by reference to a variety of factors " `considered in relation to the statute on its face' "; the clearest proof is required to override legislative intent and conclude that an Act denominated civil is punitive in purpose or effect. Id., at 100 (quoting Kennedy v. Mendoza&nbhyph;Martinez, 372 U.S. 144, 169 (1963)).

With this in mind, we turn to the Court of Appeals' determination that respondent could raise an "as-applied" challenge to the Act on double jeopardy and ex post facto grounds and seek release from confinement. Respondent essentially claims that the conditions of his confinement at the Center are too restrictive, that the conditions are incompatible with treatment, and that the system is designed to result in indefinite confinement. Respondent's claims are in many respects like the claims presented to the Court in Hendricks, where we concluded that the conditions of confinement were largely explained by the State's goal to incapacitate, not to punish. 521 U.S., at 362_368. Nevertheless, we do not deny that some of respondent's allegations are serious. Nor do we express any view as to how his allegations would bear on a court determining in the first instance whether Washington's confinement scheme is civil. Here, we evaluate respondent's allegations as presented in a double jeopardy and ex post facto challenge under the assumption that the Act is civil.

We hold that respondent cannot obtain release through an "as-applied" challenge to the Washington Act on double jeopardy and ex post facto grounds. We agree with petitioner that an "as-applied" analysis would prove unworkable. Such an analysis would never conclusively resolve whether a particular scheme is punitive and would thereby prevent a final determination of the scheme's validity under the Double Jeopardy and Ex Post Facto Clauses. Brief for Petitioner 30; Reply Brief for Petitioner 9. Unlike a fine, confinement is not a fixed event. As petitioner notes, it extends over time under conditions that are subject to change. The particular features of confinement may affect how a confinement scheme is evaluated to determine whether it is civil rather than punitive, but it remains no less true that the query must be answered definitively. The civil nature of a confinement scheme cannot be altered based merely on vagaries in the implementation of the authorizing statute.

Respondent contends that the Ninth Circuit's "as-applied" analysis comports with this Court's precedents. He points out that this Court has considered conditions of confinement in evaluating the validity of confinement schemes in the past. Brief for Respondent 11_16, 29 (citing Hendricks, supra, at 363; Reno v. Flores, 507 U.S. 292, 301_302 (1993); United States v. Salerno, 481 U.S. 739, 747_748 (1987); Allen v. Illinois, 478 U.S. 364, 373_374 (1986); Schall v. Martin, 467 U.S. 253, 269_273 (1984)).

All of those cases, however, presented the question whether the Act at issue was punitive. Permitting respondent's as-applied challenge would invite an end run around the Washington Supreme Court's decision that the Act is civil in circumstances where a direct attack on that decision is not before this Court.

Justice Thomas, concurring in the judgment, takes issue with our view that the question before the Court concerns an as applied challenge to a civil Act. He first contends that respondent's challenge is not a true "as-applied" challenge because respondent does not claim that the statute " `by its own terms' is unconstitutional as applied _ but rather that the statute is not being applied according to its terms at all." Post, at 2. We respectfully disagree. The Act requires "adequate care and individualized treatment," Wash. Rev. Code §71.09.080(2) (Supp. 2000), but the Act is silent with respect to the confinement conditions required at the Center, and that is the source of many of respondent's complaints, see supra, at 7_8. Justice Thomas next contends that we incorrectly assume that the Act is civil, instead of viewing the Act as " `otherwise _ civil,' or civil `on its face.' " Post, at 1 (emphasis added by Thomas, J.). However the Washington Act is described, our analysis in this case turns on the prior finding by the Washington Supreme Court that the Act is civil, and this Court's decision in Hendricks that a nearly identical Act was civil. Petitioner could not have claimed that the Washington Act is "otherwise" or "facially" civil without relying on those prior decisions.

In dissent, Justice Stevens argues that we "incorrectly assum[e]" that the Act is "necessarily civil," post, at 2, but the case has reached this Court under that very assumption. The Court of Appeals recognized that the Act is civil, and treated respondent's claim as an individual, "as-applied" challenge to the Act. The Court of Appeals then remanded the case to the District Court for an evidentiary hearing to determine respondent's conditions of confinement. Contrary to the dissent's characterization of the case, the Court of Appeals did not purport to undermine the validity of the Washington Act as a civil confinement scheme. The court did not conclude that respondent's allegations, if substantiated, would be sufficient to refute the Washington Supreme Court's conclusion that the Act is civil, and to require the release of all those confined under its authority. The Ninth Circuit addressed only respondent's individual case, and we do not decide claims that are not presented by the decision below. Matsushita Elec. Industrial Co. v. Epstein, 516 U.S. 367, 379 (1996). We reject the Ninth Circuit's "as-applied" analysis for double jeopardy and ex post facto claims as fundamentally flawed.

Our decision today does not mean that respondent and others committed as sexually violent predators have no remedy for the alleged conditions and treatment regime at the Center. The text of the Washington Act states that those confined under its authority have the right to adequate care and individualized treatment. Wash. Rev. Code §71.09.080(2) (Supp. 2000); Brief for Petitioner 14. As petitioner acknowledges, if the Center fails to fulfill its statutory duty, those confined may have a state law cause of action. Tr. of Oral Arg. 6, 10_11, 52. It is for the Washington courts to determine whether the Center is operating in accordance with state law and provide a remedy.

State courts, in addition to federal courts, remain competent to adjudicate and remedy challenges to civil confinement schemes arising under the Federal Constitution. As noted above, the Washington Supreme Court has already held that the Washington Act is civil in nature, designed to incapacitate and to treat. In re Young, 122 Wash. 2d, at 18_25, 857 P.2d, at 996_1000. Accordingly, due process requires that the conditions and duration of confinement under the Act bear some reasonable relation to the purpose for which persons are committed. Foucha v. Louisiana, 504 U.S. 71, 79 (1992); Youngberg v. Romeo, 457 U.S. 307, 324 (1982); Jackson v. Indiana, 406 U.S. 715, 738 (1972).

Finally, we note that a §1983 action against the Center is pending in the Western District of Washington. See supra, at 6_7. The Center operates under an injunction that requires it to adopt and implement a plan for training and hiring competent sex offender therapists; to improve relations between residents and treatment providers; to implement a treatment program for residents containing elements required by prevailing professional standards; to develop individual treatment programs; and to provide a psychologist or psychiatrist expert in the diagnosis and treatment of sex offenders to supervise the staff. App. 67. A Special Master has assisted in bringing the Center into compliance with the injunction. In its most recent published opinion on the matter, the District Court noted some progress at the Center in meeting the requirements of the injunction. Turay v. Seling, 108 F. Supp. 2d, at 1154_1155.

This case gives us no occasion to consider how the civil nature of a confinement scheme relates to other constitutional challenges, such as due process, or to consider the extent to which a court may look to actual conditions of confinement and implementation of the statute to determine in the first instance whether a confinement scheme is civil in nature. Justice Scalia, concurring, contends that conditions of confinement are irrelevant to determining whether an Act is civil unless state courts have interpreted the Act as permitting those conditions. By contrast, Justice Stevens would consider conditions of confinement at any time in order to gain "full knowledge of the effects of the statute." Post, at 3.

Whether a confinement scheme is punitive has been the threshold question for some constitutional challenges. See, e.g., Kansas v. Hendricks, 521 U.S. 346 (1997) (double jeopardy and ex post facto); United States v. Salerno, 481 U.S. 739 (1987) (due process); Allen v. Illinois, 478 U.S. 364 (1986) (Fifth Amendment privilege against self-incrimination). Whatever these cases may suggest about the relevance of conditions of confinement, they do not endorse the approach of the dissent, which would render the inquiry into the "effects of the statute," post, at 3, completely open ended. In one case, the Court refused to consider alleged confinement conditions because the parties had entered into a consent decree to improve conditions. Flores, 507 U.S., at 301. The Court presumed that conditions were in compliance with the requirements of the consent decree. Ibid. In another case, the Court found that anecdotal case histories and a statistical study were insufficient to render a regulatory confinement scheme punitive. Martin, 467 U.S., at 272. In such cases, we have decided whether a confinement scheme is punitive notwithstanding the inherent difficulty in ascertaining current conditions and predicting future events.

We have not squarely addressed the relevance of conditions of confinement to a first instance determination, and that question need not be resolved here. An Act, found to be civil, cannot be deemed punitive "as applied" to a single individual in violation of the Double Jeopardy and Ex Post Facto Clauses and provide cause for release.

The judgment of the United States Court of Appeals for the Ninth Circuit is therefore reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Justice Scalia, with whom Justice Souter joins, concurring.

I agree with the Court's holding that a statute, "found to be civil in nature, cannot be deemed punitive" or criminal "as applied" for purposes of the Ex Post Facto and Double Jeopardy Clauses. Ante, at 15. The Court accurately observes that this holding gives us "no occasion to consider the extent to which a court may look to actual conditions of confinement and implementation of the statute to determine in the first instance whether a confinement scheme is civil in nature." Ante, at 14. I write separately to dissociate myself from any implication that this reserved point may be an open question. I do not regard it as such since, three years ago, we rejected a similar double jeopardy challenge (based upon the statute's implementation "as applied" to the petitioner), where the statute had not yet been determined to be civil in nature, and where we were making that determination "in the first instance." See Hudson v. United States, 522 U.S. 93 (1997). To be consistent with the most narrow holding of that case (which, unlike this one, did not involve imposition of confinement), any consideration of subsequent implementation in the course of making a "first instance" determination cannot extend to all subsequent implementation, but must be limited to implementation of confinement, and of other impositions that are "not a fixed event," ante, at 11. That, however, would be a peculiar limitation, since even "fixed events" such as the imposition of a fine can, in their implementation, acquire penal aspects-exemplified in Hudson by the allegedly punitive size of the fines, and by the availability of reduction for "good-faith" violations, see 522 U.S., at 97_98, 104. Moreover, the language and the reasoning of Hudson leave no room for such a peculiar limitation.

In that case, the petitioners contended that the punitive nature of the statute that had been applied to them could be assessed by considering the aforementioned features of the fines. We flatly rejected that contention, which found support in our prior decision in United States v. Halper, 490 U.S. 435 (1989). Halper, we said, had erroneously made a "significant departure" from our prior jurisprudence, in deciding "to `asses[s] the character of the actual sanctions imposed,' 490 U.S., at 447, rather than, as Kennedy [v. Mendoza-Martinez, 372 U.S. 144 (1963),] demanded, evaluating the `statute on its face' to determine whether it provided for what amounted to a criminal sanction, [id.], at 169." 522 U.S., at 101. The Kennedy factors, we said, " `must be considered in relation to the statute on its face,' " 522 U.S., at 100, quoting from Kennedy v. Mendoza-Martinez, 372 U.S. 144, 169 (1963). We held that "[t]he fact that petitioners' `good faith' was considered in determining the amount of the penalty to be imposed in this case [a circumstance that would normally indicate the assessment is punitive] is irrelevant, as we look only to `the statute on its face' to determine whether a penalty is criminal in nature." Hudson, supra, at 104, quoting Kennedy, supra, at 169. We repeated, to be sure, the principle that the statutory scheme would be criminal if it was sufficiently punitive " `either in purpose or effect,' " Hudson, supra, at 99 (emphasis added), quoting United States v. Ward, 448 U.S. 242, 248_249 (1980), but it was clear from the opinion that this referred to effects apparent upon the face of the statute.

The short of the matter is that, for Double Jeopardy and Ex Post Facto Clause purposes, the question of criminal penalty vel non depends upon the intent of the legislature;1 and harsh executive implementation cannot "transfor[m] what was clearly intended as a civil remedy into a criminal penalty," Rex Trailer Co. v. United States, 350 U.S. 148, 154 (1956), any more than compassionate executive implementation can transform a criminal penalty into a civil remedy. This is not to say that there is no relief from a system that administers a facially civil statute in a fashion that would render it criminal. The remedy, however, is not to invalidate the legislature's handiwork under the Double Jeopardy Clause, but to eliminate whatever excess in administration contradicts the statute's civil character. When, as here, a state statute is at issue, the remedy for implementation that does not comport with the civil nature of the statute is resort to the traditional state proceedings that challenge unlawful executive action; if those proceedings fail, and the state courts authoritatively interpret the state statute as permitting impositions that are indeed punitive, then and only then can federal courts pronounce a statute that on its face is civil to be criminal. Such an approach protects federal courts from becoming enmeshed in the sort of intrusive inquiry into local conditions at state institutions that are best left to the State's own judiciary, at least in the first instance. And it avoids federal invalidation of state statutes on the basis of executive implementation that the state courts themselves, given the opportunity, would find to be ultra vires. Only this approach, it seems to me, is in accord with our sound and traditional reluctance to be the initial interpreter of state law. See Railroad Comm'n of Tex. v. Pullman Co., 312 U.S. 496, 500_501 (1941).

With this clarification, I join the opinion of the Court.

Justice Thomas, concurring in the judgment.

We granted certiorari to decide whether "an otherwise valid civil statute can be divested of its civil nature" simply because of an administrative agency's failure to implement the statute according to its terms. Pet. for Cert. i (emphasis added). The majority declines to answer this question. Instead, it assumes that the statute at issue is civil-rather than "otherwise _ civil," or civil "on its face." Young v. Weston, 122 F.3d 38 (CA9 1997). And then it merely holds that a statute that is civil cannot be deemed the opposite of civil-"punitive," as the majority puts it-as applied to a single individual. Ante, at 15. In explaining this conclusion, the majority expressly reserves judgment on whether the manner of implementation should affect a court's assessment of a statute as civil in the "first instance." Ante, at 11, 15. I write separately to express my view, first, that a statute which is civil on its face cannot be divested of its civil nature simply because of the manner in which it is implemented, and second, that the distinction between a challenge in the "first instance" and a subsequent challenge is one without a difference.

Before proceeding, it is important to clarify the issue in this case. The majority adopts the Ninth Circuit's nomenclature and refers to respondent's claim as an "as applied" challenge, see, e.g., ante, at 12, but that label is at best misleading. Typically an "as applied" challenge is a claim that a statute, "by its own terms, infringe[s] constitutional freedoms in the circumstances of [a] particular case." United States v. Christian Echoes Nat. Ministry, Inc., 404 U.S. 561, 565 (1972) (per curiam) (emphasis added). In contrast, respondent's claim is not that Washington's Community Protection Act of 1990 (Washington Act), Wash. Rev. Code §71.09.010 et seq. (1992), "by its own terms" is unconstitutional as applied to him,1 but rather that the statute is not being applied according to its terms at all.2 Respondent essentially contends that the actual conditions of confinement, notwithstanding the text of the statute, are punitive and incompatible with the Act's treatment purpose. See ante, at 7_8.

A challenge, such as this one, to the implementation of a facially civil statute is not only "unworkable," as the majority puts it, ante, at 11, but also prohibited by our decision in Hudson v. United States, 522 U.S. 93 (1997). In Hudson, we held that, when determining whether a statute is civil or criminal, a court must examine the "statute on its face." Id., at 101, quoting Kennedy v. Mendoza-Martinez, 372 U.S. 144, 169 (1963) (internal quotation marks omitted). In so holding, we expressly disavowed the approach used in United States v. Halper, 490 U.S. 435, 448 (1989), which evaluated the "actual sanctions imposed." 522 U.S., at 101, quoting Halper, supra, at 447 (internal quotation marks omitted). Respondent's claim is flatly inconsistent with the holding of Hudson because respondent asks us to look beyond the face of the Washington Act and to examine instead the actual sanctions imposed on him, that is, the actual conditions of confinement. Respondent argues, and the Ninth Circuit held, that Hudson's reach is limited to the particular sanctions involved in that case-monetary penalties and occupational disbarment-and does not apply here, where the sanction is confinement. Hudson, however, contains no indication whatsoever that its holding is limited to the specific sanctions at issue. To the contrary, as we explained in Hudson, a court may not elevate to dispositive status any of the factors that it may consider in determining whether a sanction is criminal.3 522 U.S., at 101. One of these nondispositive factors is confinement. Id., at 99 (stating that one of the factors is "[w]hether the sanction involves an affirmative disability or restraint," quoting Mendoza-Martinez, supra, at 168 (internal quotation marks omitted)). Yet elevating confinement to dispositive status is exactly what respondent asks us to do when he advances his distinction between confinement and other sanctions. Because Hudson rejects such an argument, respondent's claim fails.

An implementation-based challenge to a facially civil statute would be as inappropriate in reviewing the statute in the "first instance," ante, at 11, 15 (majority opinion), as it is here. In the first instance, as here, there is no place for such a challenge in the governing jurisprudence. Hudson, which requires courts to look at the face of the statute, precludes implementation-based challenges at any time. Moreover, the implementation-based claim would be as "unworkable," ante, at 11 (majority opinion), in the first instance as in later challenges. Because the actual conditions of confinement may change over time and may vary from facility to facility, an implementation-based challenge, if successful, would serve to invalidate a statute that may be implemented without any constitutional infirmities at a future time or in a separate facility. To use the majority's words, the validity of a statute should not be "based merely on vagaries in the implementation of the authorizing statute." Ibid.

And yet the majority suggests that courts may be able to consider conditions of confinement in determining whether a statute is punitive. Ante, at 12, 15. To the extent that the conditions are actually provided for on the face of the statute, I of course agree. Cf. Hudson, supra, at 101 (directing courts to look at " `the statute on its face' "). However, to the extent that the conditions result from the fact that the statute is not being applied according to its terms, the conditions are not the effect of the statute, ante, at 13, but rather the effect of its improper implementation.4 A suit based on these conditions cannot prevail.

* * *

The Washington Act does not provide on its face for punitive conditions of confinement, and the actual conditions under which the Act is implemented are of no concern to our inquiry. I therefore concur in the judgment of the Court.

Justice Stevens, dissenting.

A sexual predator may be imprisoned for violating the law, and, if he is mentally ill, he may be committed to an institution until he is cured. Whether a specific statute authorizing the detention of such a person is properly viewed as "criminal" or "civil" in the context of federal constitutional issues is often a question of considerable difficulty. See Kansas v. Hendricks, 521 U.S. 346 (1997) (reversing, by a 5 to 4 vote, a decision of the Kansas Supreme Court invalidating Kansas' Sexually Violent Predator Act); Allen v. Illinois, 478 U.S. 364 (1986) (upholding, by a 5 to 4 vote, Illinois' Sexually Dangerous Persons Act); In re Young, 122 Wash. 2d 1, 857 P.2d 989 (1993) (en banc) (upholding, by a 5 to 4 vote, the provisions of Washington's Community Protection Act of 1990 dealing with sexually violent predators).

It is settled, however, that the question whether a state statute is civil or criminal in nature for purposes of complying with the demands of the Federal Constitution is a question of federal law. If a detainee comes forward with " `the clearest proof' that `the statutory scheme [is] so punitive either in purpose or effect as to negate [the State's] intention' that the proceeding be civil, it must be considered criminal." Allen, 478 U.S., at 369 (quoting United States v. Ward, 448 U.S. 242, 248_249 (1980)) (emphasis added). See also Hudson v. United States, 522 U.S. 93, 100, 105 (1997). Accordingly, we have consistently looked to the conditions of confinement as evidence of both the legislative purpose behind the statute and its actual effect. See Hendricks, 521 U.S., at 361, 367_369; Schall v. Martin, 467 U.S. 253, 269_271 (1984); Allen, 478 U.S., at 369, 373_374. As we have acknowledged in those cases, the question whether a statute is in fact punitive cannot always be answered solely by reference to the text of the statute.

The majority in this case, however, incorrectly assumes that the Act at issue is necessarily civil. The issue the majority purports to resolve is whether an Act that is otherwise civil in nature can be deemed criminal in a specific instance based on evidence of its application to a particular prisoner. However, respondent Young's petition did not present that issue. Rather, consistent with our case law, Young sought to introduce evidence of the conditions of confinement as evidence of the punitive purpose and effect of the Washington statute. See Amended Pet. for Writ of Habeas Corpus 6 and Supp. Brief on Remand 2, 6, 10_11, in No. C94_480C (WD Wash.), Record, Doc. Nos. 57, 155. As a result, Young in no way runs afoul of Hudson v. United States, 522 U.S. 93 (1997). Properly read, Hudson acknowledges that resolving whether an Act is civil or criminal in nature can take into account whether the statutory scheme has a punitive effect.1 Id., at 99. What Hudson rejects is an approach not taken by respondent-one that bypasses this threshold question in favor of a dispositive focus on the sanction actually imposed on the specific individual.2 Id., at 101_102.

To be sure, the question whether an Act is civil or punitive in nature "is initially one of statutory construction." Ante, at 9 (majority opinion). However, under the majority's analysis, there is no inquiry beyond that of statutory construction. Ante, at 11. In essence, the majority argues that because the constitutional query must be answered definitively and because confinement is not a "fixed event," conditions of confinement should not be considered at all, except in the first challenge to a statute, when, as a practical matter, the evidence of such conditions is most likely not to constitute the requisite "clearest proof." This seems to me quite wrong. If conditions of confinement are such that a detainee has been punished twice in violation of the Double Jeopardy Clause, it is irrelevant that the scheme has been previously labeled as civil without full knowledge of the effects of the statute.3

In this case, Young has made detailed allegations concerning both the absence of treatment for his alleged mental illness and the starkly punitive character of the conditions of his confinement. If proved, those allegations establish not just that those detained pursuant to the statute are treated like those imprisoned for violations of Washington's criminal laws, but that, in many respects, they receive significantly worse treatment.4 If those allegations are correct, the statute in question should be characterized as a criminal law for federal constitutional purposes. I therefore agree with the Court of Appeals' conclusion that respondent should be given the opportunity to come forward with the "clearest proof" that his allegations are true.

Accordingly, I respectfully dissent.

In Freedman v. Maryland, 380 U.S. 51 (1965), a case involving a state motion-picture censorship scheme, the Court announced procedural requirements necessary to guard against unconstitutional prior restraint of expression. Those requirements included assurance of "a prompt final judicial decision, to minimize the deterrent effect of an interim and possibly erroneous denial of a license." Id., at 59. Twenty-five years later, in FW/PBS, Inc. v. Dallas, 493 U.S. 215 (1990), the Court applied some of the Freedman standards to a municipal ordinance conditioning the operation of sexually oriented businesses on receipt of a license. Unsuccessful applicants for an adult-business license, the opinion announcing the judgment stated, must be accorded "an avenue for prompt judicial review." Id., at 229.

Courts have divided over the meaning of FW/PBS's "prompt judicial review" requirement. Some have held that the unsuccessful applicant for an adult business license must be assured a prompt judicial determination on the merits of the permit denial. See, e.g., Baby Tam & Co. v. Las Vegas, 154 F.3d 1097, 1101_1102 (CA9 1998); 11126 Baltimore Blvd., Inc. v. Prince George's County, 58 F.3d 988, 999_1000 (CA4 1995) (en banc). Others, like the Court of Appeals of Wisconsin whose judgment is before us, 231 Wis. 2d 93, 115_116, 604 N. W. 2d 870, 882 (1999), have held that prompt access to court review suffices. See, e.g., Boss Capital, Inc. v. City of Casselberry, 187 F.3d 1251, 1256_1257 (CA11 1999); TK's Video, Inc. v. Denton County, 24 F.3d 705, 709 (CA5 1994). We granted certiorari to resolve the conflict. 530 U.S. 1249 (2000). We now find, however, that the issue stemming from Freedman is not genuinely presented to us in this case. We therefore dismiss the petition and leave the judgment of the Wisconsin court undisturbed.

* The City of Waukesha, Wisconsin (City), requires sellers of sexually explicit materials to obtain and annually renew adult business licenses. See Waukesha Municipal Code §§8.195(2), (7) (1995), reprinted in App. to Pet. for Cert. 101, 104. Petitioner City News and Novelty, Inc. (City News), pursuant to a City license first obtained in 1989, owned and operated an adult-oriented shop in downtown Waukesha. In November 1995, City News applied for a renewal of its license, then due to expire in two months. In December 1995, Waukesha's Common Council (Council) denied the application, finding that City News had violated the City's ordinance by permitting minors to loiter on the premises, failing to maintain an unobstructed view of booths in the store, and allowing patrons to engage in sexual activity inside the booths. Waukesha's refusal to renew City News's license was upheld in administrative proceedings and on judicial review in the state courts.

Petitioning for certiorari, City News raised three questions. First, City News asserted that the persuasion burden had been improperly assigned to it. Second, City News urged that Waukesha's ordinance unconstitutionally accorded City officials unbridled discretion to vary punishments for ordinance violations. Third, City News asked us to "resolve the conflict among the circuits concerning whether the guarantee of prompt judicial review that must accompany [an adult business] licensing scheme means a prompt judicial determination or simply the right to promptly file for judicial review." Pet. for Cert. 13. We granted the petition only on the third question. Accordingly, City News cannot now contend that any of the substantive requirements governing adult business licenses in Waukesha conflict with the First Amendment. Nor does City News contend that the evidence failed to substantiate the charged violations. We now explain why City News is not properly situated to raise the question on which we granted review.

In letters sent to Waukesha two months after petitioning for review in this Court, City News gave notice that it would withdraw its renewal application and close its business upon the City's grant of a license to another corporation, B. J. B., Inc., "a larger and more modern business" with which City News felt "it could not effectively compete." Letters from Jeff Scott Olson to Vince Moschella (June 12 and 19, 2000), Respondent's Lodging, Vol. 1, Tab No. 14. Waukesha granted B. J. B.'s license application on June 20. It is undisputed that City News has ceased to operate as an adult business and no longer seeks to renew its license. Tr. of Oral Arg. 14_15.

Observing that City News neither now pursues nor currently expresses an intent to pursue a license under Waukesha law, Waukesha asserts that the case has become moot, for City News no longer has "a legally cognizable interest in the outcome." County of Los Angeles v. Davis, 440 U.S. 625, 631 (1979) (citing Powell v. McCormack, 395 U.S. 486, 496 (1969)). We agree that the case no longer qualifies for judicial review. Urging that the case remains fit for adjudication, City News tenders two points. We find neither persuasive.

Noting that it "has never promised not to apply for a license" in the future, Reply Brief 1, City News first contends that, notwithstanding the voluntary termination of its license renewal effort, a live controversy remains under the Court's reasoning in Erie v. Pap's A. M., 529 U.S. 277 (2000). In our view, Erie differs critically from this case. In Erie, we similarly granted a petition to review a state-court judgment addressing an adult business' First Amendment challenge to a city ordinance. We concluded that the controversy persisted, even though the adult business had shut down. We reached that conclusion, it is true, in part because the business "could again decide to operate." Id., at 287. That speculation standing alone, however, did not shield the case from a mootness determination. Another factor figured prominently. The nude dancing entrepreneur in Erie sought "to have the case declared moot" after the business had "prevailed below," obtaining a judgment that invalidated Erie's ordinance. Id., at 288. Had we accepted the entrepreneur's plea, then consistent with our practice when a case becomes moot on review from a state court, we would have dismissed the petition, leaving intact the judgment below. See ASARCO Inc. v. Kadish, 490 U.S. 605, 621, n. 1 (1989); Erie, 529 U.S., at 305 (Scalia, J., concurring in judgment). Thus, had we declared Erie moot, the defendant municipality would have been saddled with an "ongoing injury," i.e., the judgment striking its law. Erie, 529 U.S., at 288. And the plaintiff arguably would have prevailed in an "attemp[t] to manipulate the Court's jurisdiction to insulate a favorable decision from review." Ibid.

In this case, we confront no parallel circumstance. The adult enterprise before us left the fray as a loser, not a winner. Our dismissal here does not keep Waukesha under the weight of an adverse judgment, or deprive Waukesha of its victory in state court. Nor does a mootness dismissal reward an arguable manipulation of our jurisdiction, for plaintiff City News, unlike the nude dancing entrepreneur in Erie, opposes a declaration of mootness.1

City News also urges that it experiences ongoing injury because it is conclusively barred by Waukesha's ordinance from reopening as an adult business until 2005. It is far from clear, however, whether City News actually suffers that disability.2 And as our prior discussion suggests, supra, at 4, a live controversy is not maintained by speculation that City News might be temporarily disabled from reentering a business that City News has left and currently asserts no plan to reenter. See Spencer v. Kemna, 523 U.S. 1, 15_16 (1998).

City News's contention that it remains a qualified complainant also fails for a separate reason. Full briefing and argument have revealed that the question City News tendered, and which we took up for review, is not now and never was accurately reflective of City News's grievance. Unlike the initial license applicant whose expression cannot begin prepermission (the situation of the complainant in Freedman), City News was already licensed to conduct an adult business and sought to fend off a stop order. Swift judicial review is the remedy needed by those held back from speaking. We do not doubt that an ongoing adult enterprise facing loss of its license to do business may allege First Amendment injuries. Such an establishment's typical concern, however, is not the speed of court proceedings, but the availability of a stay of adverse action during the pendency of judicial review, however long that review takes.

Unlike the petitioner in Freedman, who sought, through swift court review, an end to the status quo of silence, City News sought to maintain, pendente lite, the status quo of speech (or expressive conduct). Brief for Petitioner 43_44. We venture no view on the merits of an argument urging preservation of speech (or expressive conduct) as the status quo pending administrative and judicial review proceedings. It suffices to point out that the question is not the one on which the courts have divided or on which we granted certiorari.

For the reasons stated, the writ of certiorari is

Dismissed.

The issue is whether a statewide association incorporated to regulate interscholastic athletic competition among public and private secondary schools may be regarded as engaging in state action when it enforces a rule against a member school. The association in question here includes most public schools located within the State, acts through their representatives, draws its officers from them, is largely funded by their dues and income received in their stead, and has historically been seen to regulate in lieu of the State Board of Education's exercise of its own authority. We hold that the association's regulatory activity may and should be treated as state action owing to the pervasive entwinement of state school officials in the structure of the association, there being no offsetting reason to see the association's acts in any other way.

* Respondent Tennessee Secondary School Athletic Association (Association) is a not-for-profit membership corporation organized to regulate interscholastic sport among the public and private high schools in Tennessee that belong to it. No school is forced to join, but without any other authority actually regulating interscholastic athletics, it enjoys the memberships of almost all the State's public high schools (some 290 of them or 84% of the Association's voting membership), far outnumbering the 55 private schools that belong. A member school's team may play or scrimmage only against the team of another member, absent a dispensation.

The Association's rulemaking arm is its legislative council, while its board of control tends to administration. The voting membership of each of these nine-person committees is limited under the Association's bylaws to high school principals, assistant principals, and superintendents elected by the member schools, and the public school administrators who so serve typically attend meetings during regular school hours. Although the Association's staff members are not paid by the State, they are eligible to join the State's public retirement system for its employees. Member schools pay dues to the Association, though the bulk of its revenue is gate receipts at member teams' football and basketball tournaments, many of them held in public arenas rented by the Association.

The constitution, bylaws, and rules of the Association set standards of school membership and the eligibility of students to play in interscholastic games. Each school, for example, is regulated in awarding financial aid, most coaches must have a Tennessee state teaching license, and players must meet minimum academic standards and hew to limits on student employment. Under the bylaws, "in all matters pertaining to the athletic relations of his school," App. 138, the principal is responsible to the Association, which has the power "to suspend, to fine, or otherwise penalize any member school for the violation of any of the rules of the Association or for other just cause," id., at 100.

Ever since the Association was incorporated in 1925, Tennessee's State Board of Education (State Board) has (to use its own words) acknowledged the corporation's functions "in providing standards, rules and regulations for interscholastic competition in the public schools of Tennessee," id., at 211. More recently, the State Board cited its statutory authority, Tenn. Code Ann. 49-1-302 (App. 220), when it adopted language expressing the relationship between the Association and the Board. Specifically, in 1972, it went so far as to adopt a rule expressly "designat[ing]" the Association as "the organization to supervise and regulate the athletic activities in which the public junior and senior high schools in Tennessee participate on an interscholastic basis." Tennessee State Board of Education, Administrative Rules and Regulations, Rule 0520-1-2-.26 (1972) (later moved to Rule 0520-1-2-.08). The Rule provided that "the authority granted herein shall remain in effect until revoked" and instructed the State Board's chairman to "designate a person or persons to serve in an ex-officio capacity on the [Association's governing bodies]." App. 211. That same year, the State Board specifically approved the Association's rules and regulations, while reserving the right to review future changes. Thus, on several occasions over the next 20 years, the State Board reviewed, approved, or reaffirmed its approval of the recruiting Rule at issue in this case. In 1996, however, the State Board dropped the original Rule 0520-1-2-.08 expressly designating the Association as regulator; it substituted a statement "recogniz[ing] the value of participation in interscholastic athletics and the role of [the Association] in coordinating interscholastic athletic competition," while "authoriz[ing] the public schools of the state to voluntarily maintain membership in [the Association]." Id., at 220.

The action before us responds to a 1997 regulatory enforcement proceeding brought against petitioner, Brentwood Academy, a private parochial high school member of the Association. The Association's board of control found that Brentwood violated a rule prohibiting "undue influence" in recruiting athletes, when it wrote to incoming students and their parents about spring football practice. The Association accordingly placed Brentwood's athletic program on probation for four years, declared its football and boys' basketball teams ineligible to compete in playoffs for two years, and imposed a $3,000 fine. When these penalties were imposed, all the voting members of the board of control and legislative council were public school administrators.

Brentwood sued the Association and its executive director in federal court under Rev. Stat. 1979, 42 U.S.C. 1983 claiming that enforcement of the Rule was state action and a violation of the First and Fourteenth Amendments. The District Court entered summary judgment for Brentwood and enjoined the Association from enforcing the Rule. 13 F. Supp. 2d 670 (MD Tenn. 1998). In holding the Association to be a state actor under 1983 and the Fourteenth Amendment, the District Court found that the State had delegated authority over high school athletics to the Association, characterized the relationship between the Association and its public school members as symbiotic, and emphasized the predominantly public character of the Association's membership and leadership. The court relied on language in National Collegiate Athletic Assn. v. Tarkanian, 488 U.S. 179, 193, n. 13 (1988), suggesting that statewide interscholastic athletic associations are state actors, and on other federal cases in which such organizations had uniformly been held to be acting under color of state law.

The United States Court of Appeals for the Sixth Circuit reversed. 180 F.3d 758 (1999). It recognized that there is no single test to identify state actions and state actors but applied three criteria derived from Blum v. Yaretsky, 457 U.S. 991 (1982), Lugar v. Edmondson Oil Co., 457 U.S. 922 (1982), and Rendell-Baker v. Kohn, 457 U.S. 830 (1982), and found no state action under any of them. It said the District Court was mistaken in seeing a symbiotic relationship between the State and the Association, it emphasized that the Association was neither engaging in a traditional and exclusive public function nor responding to state compulsion, and it gave short shrift to the language from Tarkanian on which the District Court relied. Rehearing en banc was later denied over the dissent of two judges, who criticized the panel decision for creating a conflict among state and federal courts, for being inconsistent with Tarkanian, and for lacking support in the "functional" analysis of private activity required by West v. Atkins, 487 U.S. 42 (1988), for assessing the significance of cooperation between public officials and a private actor. 190 F.3d 705 (CA6 1999) (Merritt, J., dissenting from denial of rehearing en banc).

We granted certiorari, 528 U.S. 1153 (2000), to resolve the conflict1 and now reverse.

* Our cases try to plot a line between state action subject to Fourteenth Amendment scrutiny and private conduct (however exceptionable) that is not. Tarkanian, supra, at 191; Jackson v. Metropolitan Edison Co., 419 U.S. 345, 349 (1974). The judicial obligation is not only to " 'preserv[e] an area of individual freedom by limiting the reach of federal law' and avoi[d] the imposition of responsibility on a State for conduct it could not control," Tarkanian, supra, at 191 (quoting Lugar, supra, at 936-937), but also to assure that constitutional standards are invoked "when it can be said that the State is responsible for the specific conduct of which the plaintiff complains," Blum, supra, at 1004 (emphasis in original). If the Fourteenth Amendment is not to be displaced, therefore, its ambit cannot be a simple line between States and people operating outside formally governmental organizations, and the deed of an ostensibly private organization or individual is to be treated sometimes as if a State had caused it to be performed. Thus, we say that state action may be found if, though only if, there is such a "close nexus between the State and the challenged action" that seemingly private behavior "may be fairly treated as that of the State itself." Jackson, supra, at 351.2

What is fairly attributable is a matter of normative judgment, and the criteria lack rigid simplicity. From the range of circumstances that could point toward the State behind an individual face, no one fact can function as a necessary condition across the board for finding state action; nor is any set of circumstances absolutely sufficient, for there may be some countervailing reason against attributing activity to the government. See Tarkanian, 488 U.S., at 193, 196; Polk County v. Dodson, 454 U.S. 312 (1981).

Our cases have identified a host of facts that can bear on the fairness of such an attribution. We have, for example, held that a challenged activity may be state action when it results from the State's exercise of "coercive power," Blum, 457 U.S., at 1004, when the State provides "significant encouragement, either overt or covert," ibid., or when a private actor operates as a "willful participant in joint activity with the State or its agents," Lugar, supra, at 941 (internal quotation marks omitted). We have treated a nominally private entity as a state actor when it is controlled by an "agency of the State," Pennsylvania v. Board of Directors of City Trusts of Philadelphia, 353 U.S. 230, 231 (1957) (per curiam), when it has been delegated a public function by the State, cf., e.g., West v. Atkins, supra, at 56; Edmonson v. Leesville Concrete Co., 500 U.S. 614, 627-628 (1991), when it is "entwined with governmental policies" or when government is "entwined in [its] management or control," Evans v. Newton, 382 U.S. 296, 299, 301 (1966).

Amidst such variety, examples may be the best teachers, and examples from our cases are unequivocal in showing that the character of a legal entity is determined neither by its expressly private characterization in statutory law, nor by the failure of the law to acknowledge the entity's inseparability from recognized government officials or agencies. Lebron v. National Railroad Passenger Corporation, 513 U.S. 374 (1995), held that Amtrak was the Government for constitutional purposes, regardless of its congressional designation as private; it was organized under federal law to attain governmental objectives and was directed and controlled by federal appointees. Pennsylvania v. Board of Directors of City Trusts of Philadelphia, supra, held the privately endowed Gerard College to be a state actor and enforcement of its private founder's limitation of admission to whites attributable to the State, because, consistent with the terms of the settlor's gift, the college's board of directors was a state agency established by state law. Ostensibly the converse situation occurred in Evans v. Newton, supra, which held that private trustees to whom a city had transferred a park were nonetheless state actors barred from enforcing racial segregation, since the park served the public purpose of providing community recreation, and "the municipality remain[ed] entwined in [its] management [and] control," id., at 301.

These examples of public entwinement in the management and control of ostensibly separate trusts or corporations foreshadow this case, as this Court itself anticipated in Tarkanian, supra. Tarkanian arose when an undoubtedly state actor, the University of Nevada, suspended its basketball coach, Tarkanian, in order to comply with rules and recommendations of the National Collegiate Athletic Association (NCAA). The coach charged the NCAA with state action, arguing that the state university had delegated its own functions to the NCAA, clothing the latter with authority to make and apply the university's rules, the result being joint action making the NCAA a state actor.

To be sure, it is not the strict holding in Tarkanian that points to our view of this case, for we found no state action on the part of the NCAA. We could see, on the one hand, that the university had some part in setting the NCAA's rules, and the Supreme Court of Nevada had gone so far as to hold that the NCAA had been delegated the university's traditionally exclusive public authority over personnel. Id., at 190. But on the other side, the NCAA's policies were shaped not by the University of Nevada alone, but by several hundred member institutions, most of them having no connection with Nevada, and exhibiting no color of Nevada law. Id., at 193. Since it was difficult to see the NCAA, not as a collective membership, but as surrogate for the one State, we held the organization's connection with Nevada too insubstantial to ground a state action claim. Id., at 193, 196.

But dictum in Tarkanian pointed to a contrary result on facts like ours, with an organization whose member public schools are all within a single State. "The situation would, of course, be different if the [Association's] membership consisted entirely of institutions located within the same State, many of them public institutions created by the same sovereign." Id., at 193, n. 13. To support our surmise, we approvingly cited two cases: Clark v. Arizona Interscholastic Assn., 695 F.2d 1126 (CA9 1982), cert. denied, 464 U.S. 818 (1983), a challenge to a state high school athletic association that kept boys from playing on girls' interscholastic volleyball teams in Arizona; and Louisiana High School Athletic Assn. v. St. Augustine High School, 396 F.2d 224 (CA5 1968), a parochial school's attack on the racially segregated system of interscholastic high school athletics maintained by the athletic association. In each instance, the Court of Appeals treated the athletic association as a state actor.

Just as we foresaw in Tarkanian, the "necessarily fact-bound inquiry," Lugar, 457 U.S., at 939, leads to the conclusion of state action here. The nominally private character of the Association is overborne by the pervasive entwinement of public institutions and public officials in its composition and workings, and there is no substantial reason to claim unfairness in applying constitutional standards to it.

The Association is not an organization of natural persons acting on their own, but of schools, and of public schools to the extent of 84% of the total. Under the Association's bylaws, each member school is represented by its principal or a faculty member, who has a vote in selecting members of the governing legislative council and board of control from eligible principals, assistant principals and superintendents.

Although the findings and prior opinions in this case include no express conclusion of law that public school officials act within the scope of their duties when they represent their institutions, no other view would be rational, the official nature of their involvement being shown in any number of ways. Interscholastic athletics obviously play an integral part in the public education of Tennessee, where nearly every public high school spends money on competitions among schools. Since a pickup system of interscholastic games would not do, these public teams need some mechanism to produce rules and regulate competition. The mechanism is an organization overwhelmingly composed of public school officials who select representatives (all of them public officials at the time in question here), who in turn adopt and enforce the rules that make the system work. Thus, by giving these jobs to the Association, the 290 public schools of Tennessee belonging to it can sensibly be seen as exercising their own authority to meet their own responsibilities. Unsurprisingly, then, the record indicates that half the council or board meetings documented here were held during official school hours, and that public schools have largely provided for the Association's financial support. A small portion of the Association's revenue comes from membership dues paid by the schools, and the principal part from gate receipts at tournaments among the member schools. Unlike mere public buyers of contract services, whose payments for services rendered do not convert the service providers into public actors, see Rendell-Baker, 457 U.S., at 839-843, the schools here obtain membership in the service organization and give up sources of their own income to their collective association. The Association thus exercises the authority of the predominantly public schools to charge for admission to their games; the Association does not receive this money from the schools, but enjoys the schools' moneymaking capacity as its own.

In sum, to the extent of 84% of its membership, the Association is an organization of public schools represented by their officials acting in their official capacity to provide an integral element of secondary public schooling. There would be no recognizable Association, legal or tangible, without the public school officials, who do not merely control but overwhelmingly perform all but the purely ministerial acts by which the Association exists and functions in practical terms. Only the 16% minority of private school memberships prevents this entwinement of the Association and the public school system from being total and their identities totally indistinguishable.

To complement the entwinement of public school officials with the Association from the bottom up, the State of Tennessee has provided for entwinement from top down. State Board members are assigned ex officio to serve as members of the board of control and legislative council, and the Association's ministerial employees are treated as state employees to the extent of being eligible for membership in the state retirement system.

It is, of course, true that the time is long past when the close relationship between the surrogate association and its public members and public officials acting as such was attested frankly. As mentioned, the terms of the State Board's Rule expressly designating the Association as regulator of interscholastic athletics in public schools was deleted in 1996, the year after a Federal District Court held that the Association was a state actor because its rules were "caused, directed and controlled by the Tennessee Board of Education," Graham v. TSSAA, No. 1:95-CV-044, 1995 WL 115890, *5 (ED Tenn., Feb. 20, 1995).3

But the removal of the designation language from Rule 0520-1-2-.08 affected nothing but words. Today the State Board's member-designees continue to sit on the Association's committees as nonvoting members, and the State continues to welcome Association employees in its retirement scheme. The close relationship is confirmed by the Association's enforcement of the same preamendment rules and regulations reviewed and approved by the State Board (including the recruiting Rule challenged by Brentwood), and by the State Board's continued willingness to allow students to satisfy its physical education requirement by taking part in interscholastic athletics sponsored by the Association. The most one can say on the evidence is that the State Board once freely acknowledged the Association's official character but now does it by winks and nods.4 The amendment to the Rule in 1996 affected candor but not the "momentum" of the Association's prior involvement with the State Board. Evans v. Newton, 382 U.S., at 301. The District Court spoke to this point in finding that because of "custom and practice," "the conduct of the parties has not materially changed" since 1996, "the connections between TSSAA and the State [being] still pervasive and entwined." 13 F. Supp. 2d, at 681.

The entwinement down from the State Board is therefore unmistakable, just as the entwinement up from the member public schools is overwhelming. Entwinement will support a conclusion that an ostensibly private organization ought to be charged with a public character and judged by constitutional standards; entwinement to the degree shown here requires it.

Entwinement is also the answer to the Association's several arguments offered to persuade us that the facts would not support a finding of state action under various criteria applied in other cases. These arguments are beside the point, simply because the facts justify a conclusion of state action under the criterion of entwinement, a conclusion in no sense unsettled merely because other criteria of state action may not be satisfied by the same facts.

The Association places great stress, for example, on the application of a public function test, as exemplified in Rendell-Baker v. Kohn, 457 U.S. 830 (1982). There, an apparently private school provided education for students whose special needs made it difficult for them to finish high school. The record, however, failed to show any tradition of providing public special education to students unable to cope with a regular school, who had historically been cared for (or ignored) according to private choice. It was true that various public school districts had adopted the practice of referring students to the school and paying their tuition, and no one disputed that providing the instruction aimed at a proper public objective and conferred a public benefit. But we held that the performance of such a public function did not permit a finding of state action on the part of the school unless the function performed was exclusively and traditionally public, as it was not in that case. The Association argues that application of the public function criterion would produce the same result here, and we will assume, arguendo, that it would. But this case does not turn on a public function test, any more than Rendell-Baker had anything to do with entwinement of public officials in the special school.

For the same reason, it avails the Association nothing to stress that the State neither coerced nor encouraged the actions complained of. "Coercion" and "encouragement" are like "entwinement" in referring to kinds of facts that can justify characterizing an ostensibly private action as public instead. Facts that address any of these criteria are significant, but no one criterion must necessarily be applied. When, therefore, the relevant facts show pervasive entwinement to the point of largely overlapping identity, the implication of state action is not affected by pointing out that the facts might not loom large under a different test.

This is not to say that all of the Association's arguments are rendered beside the point by the public officials' involvement in the Association, for after application of the entwinement criterion, or any other, there is a further potential issue, and the Association raises it. Even facts that suffice to show public action (or, standing alone, would require such a finding) may be outweighed in the name of some value at odds with finding public accountability in the circumstances. In Polk County, 454 U.S., at 322, a defense lawyer's actions were deemed private even though she was employed by the county and was acting within the scope of her duty as a public defender. Full-time public employment would be conclusive of state action for some purposes, see West v. Atkins, 487 U.S., at 50, accord, Lugar, 457 U.S., at 935, n. 18, but not when the employee is doing a defense lawyer's primary job; then, the public defender does "not ac[t] on behalf of the State; he is the State's adversary." Polk County, supra, at 323, n. 13. The state-action doctrine does not convert opponents into virtual agents.

The assertion of such a countervailing value is the nub of each of the Association's two remaining arguments, neither of which, however, persuades us. The Association suggests, first, that reversing the judgment here will somehow trigger an epidemic of unprecedented federal litigation. Brief for Respondents 35. Even if that might be counted as a good reason for a Polk County decision to call the Association's action private, the record raises no reason for alarm here. Save for the Sixth Circuit, every Court of Appeals to consider a statewide athletic association like the one here has found it a state actor. This majority view began taking shape even before Tarkanian, which cited two such decisions approvingly, see supra, at 9, (and this was six years after Blum, Rendell-Baker, and Lugar, on which the Sixth Circuit relied here). No one, however, has pointed to any explosion of 1983 cases against interscholastic athletic associations in the affected jurisdictions. Not to put too fine a point on it, two District Courts in Tennessee have previously held the Association itself to be a state actor, see Graham, 1995 WL 115890, at *5; Crocker v. Tennessee Secondary School Athletic Assn., 735 F. Supp. 753 (MD Tenn. 1990), affirmance order, 908 F.2d 972, 973 (CA6 1990), but there is no evident wave of litigation working its way across the State. A reversal of the judgment here portends nothing more than the harmony of an outlying Circuit with precedent otherwise uniform.

Nor do we think there is anything to be said for the Association's contention that there is no need to treat it as a state actor since any public school applying the Association's rules is itself subject to suit under 1983 or Title IX of the Education Amendments of 1972, 86 Stat. 373, 20 U.S.C. 1681-1688. Brief for Respondents 30. If Brentwood's claim were pushing at the edge of the class of possible defendant state actors, an argument about the social utility of expanding that class would at least be on point, but because we are nowhere near the margin in this case, the Association is really asking for nothing less than a dispensation for itself. Its position boils down to saying that the Association should not be dressed in state clothes because other, concededly public actors are; that Brentwood should be kept out of court because a different plaintiff raising a different claim in a different case may find the courthouse open. Pleas for special treatment are hard to sell, although saying that does not, of course, imply anything about the merits of Brentwood's complaint; the issue here is merely whether Brentwood properly names the Association as a 1983 defendant, not whether it should win on its claim.

The judgment of the Court of Appeals for the Sixth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

Justice Thomas, with whom The Chief Justice, Justice Scalia, and Justice Kennedy join, dissenting.

We have never found state action based upon mere "entwinement." Until today, we have found a private organization's acts to constitute state action only when the organization performed a public function; was created, coerced, or encouraged by the government; or acted in a symbiotic relationship with the government. The majority's holding-that the Tennessee Secondary School Athletic Association's (TSSAA) enforcement of its recruiting rule is state action-not only extends state-action doctrine beyond its permissible limits but also encroaches upon the realm of individual freedom that the doctrine was meant to protect. I respectfully dissent.

* Like the state-action requirement of the Fourteenth Amendment, the state-action element of 42 U.S.C. 1983 excludes from its coverage "merely private conduct, however discriminatory or wrongful." American Mfrs. Mut. Ins. Co. v. Sullivan, 526 U.S. 40, 50 (1999) (internal quotation marks omitted). "Careful adherence to the 'state action' requirement" thus "preserves an area of individual freedom by limiting the reach of federal law and federal judicial power." Lugar v. Edmondson Oil Co., 457 U.S. 922, 936 (1982). The state-action doctrine also promotes important values of federalism, "avoid[ing] the imposition of responsibility on a State for conduct it could not control." National Collegiate Athletic Assn. v. Tarkanian, 488 U.S. 179, 191 (1988). Although we have used many different tests to identify state action, they all have a common purpose. Our goal in every case is to determine whether an action "can fairly be attributed to the State." Blum v. Yaretsky, 457 U.S. 991, 1004 (1982); American Mfrs., supra, at 52.

* Regardless of these various tests for state action, common sense dictates that the TSSAA's actions cannot fairly be attributed to the State, and thus cannot constitute state action. The TSSAA was formed in 1925 as a private corporation to organize interscholastic athletics and to sponsor tournaments among its member schools. Any private or public secondary school may join the TSSAA by signing a contract agreeing to comply with its rules and decisions. Although public schools currently compose 84% of the TSSAA's membership, the TSSAA does not require that public schools constitute a set percentage of its membership, and, indeed, no public school need join the TSSAA. The TSSAA's rules are enforced not by a state agency but by its own board of control, which comprises high school principals, assistant principals, and superintendents, none of whom must work at a public school. Of course, at the time the recruiting rule was enforced in this case, all of the board members happened to be public school officials. However, each board member acts in a representative capacity on behalf of all the private and public schools in his region of Tennessee, and not simply his individual school.

The State of Tennessee did not create the TSSAA. The State does not fund the TSSAA and does not pay its employees.1 In fact, only 4% of the TSSAA's revenue comes from the dues paid by member schools; the bulk of its operating budget is derived from gate receipts at tournaments it sponsors. The State does not permit the TSSAA to use state-owned facilities for a discounted fee, and it does not exempt the TSSAA from state taxation. No Tennessee law authorizes the State to coordinate interscholastic athletics or empowers another entity to organize interscholastic athletics on behalf of the State.2 The only state pronouncement acknowledging the TSSAA's existence is a rule providing that the State Board of Education permits public schools to maintain membership in the TSSAA if they so choose.3 Moreover, the State of Tennessee has never had any involvement in the particular action taken by the TSSAA in this case: the enforcement of the TSSAA's recruiting rule prohibiting members from using "undue influence" on students or their parents or guardians "to secure or to retain a student for athletic purposes." App. 115. There is no indication that the State has ever had any interest in how schools choose to regulate recruiting.4 In fact, the TSSAA's authority to enforce its recruiting rule arises solely from the voluntary membership contract that each member school signs, agreeing to conduct its athletics in accordance with the rules and decisions of the TSSAA.

Even approaching the issue in terms of any of the Court's specific state-action tests, the conclusion is the same: The TSSAA's enforcement of its recruiting rule against Brentwood Academy is not state action. In applying these tests, courts of course must place the burden of persuasion on the plaintiff, not the defendant, because state action is an element of a 1983 claim. American Mfrs., 526 U.S., at 49-50; West v. Atkins, 487 U.S. 42, 48 (1988).

The TSSAA has not performed a function that has been "traditionally exclusively reserved to the State." Jackson v. Metropolitan Edison Co., 419 U.S. 345, 352 (1974). The organization of interscholastic sports is neither a traditional nor an exclusive public function of the States. Widespread organization and administration of interscholastic contests by schools did not begin until the 20th century. See M. Lee, A History of Physical Education and Sports in the U.S. A. 73 (1983) (explaining that what little interscholastic athletics there was in the 19th century "came almost entirely in the closing decade of the century and was largely pupil inspired, pupil controlled, and pupil coached"); id., at 68, 146 (stating that no control of high school sports occurred until 1896, when a group of teachers in Wisconsin set up a committee to control such contests, and pointing out that "[i]t was several years before the idea caught on in other states"). Certainly, in Tennessee, the State did not even show an interest in interscholastic athletics until 47 years after the TSSAA had been in existence and had been orchestrating athletic contests throughout the State. Even then, the State Board of Education merely acquiesced in the TSSAA's actions and did not assume the role of regulating interscholastic athletics. Cf. Blum, 457 U.S., at 1004-1005 ("Mere approval of or acquiescence in the initiatives of a private party is not sufficient to justify holding the State responsible for those initiatives . . ."); see also Flagg Bros., Inc. v. Brooks, 436 U.S. 149, 164-165 (1978). The TSSAA no doubt serves the public, particularly the public schools, but the mere provision of a service to the public does not render such provision a traditional and exclusive public function. See Rendell-Baker v. Kohn, 457 U.S. 830, 842 (1982).

It is also obvious that the TSSAA is not an entity created and controlled by the government for the purpose of fulfilling a government objective, as was Amtrak in Lebron v. National Railroad Passenger Corporation, 513 U.S. 374, 394 (1995). See also Pennsylvania v. Board of Directors of City Trusts of Philadelphia, 353 U.S. 230 (1957) (per curiam) (holding that a state agency created under state law was a state actor). Indeed, no one claims that the State of Tennessee played any role in the creation of the TSSAA as a private corporation in 1925. The TSSAA was designed to fulfill an objective-the organization of interscholastic athletic tournaments-that the government had not contemplated, much less pursued. And although the board of control currently is composed of public school officials, and although public schools currently account for the majority of the TSSAA's membership, this is not required by the TSSAA's constitution.

In addition, the State of Tennessee has not "exercised coercive power or . . . provided such significant encouragement [to the TSSAA], either overt or covert," Blum, 457 U.S., at 1004, that the TSSAA's regulatory activities must in law be deemed to be those of the State. The State has not promulgated any regulations of interscholastic sports, and nothing in the record suggests that the State has encouraged or coerced the TSSAA in enforcing its recruiting rule. To be sure, public schools do provide a small portion of the TSSAA's funding through their membership dues, but no one argues that these dues are somehow conditioned on the TSSAA's enactment and enforcement of recruiting rules.5 Likewise, even if the TSSAA were dependent on state funding to the extent of 90%, as was the case in Blum, instead of less than 4%, mere financial dependence on the State does not convert the TSSAA's actions into acts of the State. See Blum, supra, at 1011; Rendell-Baker, supra, at 840; see also Moose Lodge No. 107 v. Irvis, 407 U.S. 163, 173 (1972) ("The Court has never held, of course, that discrimination by an otherwise private entity would be violative of the Equal Protection Clause if the private entity receives any sort of benefit or service at all from the State . . ."). Furthermore, there is no evidence of "joint participation," Lugar, 457 U.S., at 941-942, between the State and the TSSAA in the TSSAA's enforcement of its recruiting rule. The TSSAA's board of control enforces its recruiting rule solely in accordance with the authority granted to it under the contract that each member signs.

Finally, there is no "symbiotic relationship" between the State and the TSSAA. Moose Lodge, supra, at 175; cf. Burton v. Wilmington Parking Authority, 365 U.S. 715 (1961). Contrary to the majority's assertion, see ante, at 10-11, the TSSAA's "fiscal relationship with the State is not different from that of many contractors performing services for the government." Rendell-Baker, supra, at 843. The TSSAA provides a service-the organization of athletic tournaments-in exchange for membership dues and gate fees, just as a vendor could contract with public schools to sell refreshments at school events. Certainly the public school could sell its own refreshments, yet the existence of that option does not transform the service performed by the contractor into a state action. Also, there is no suggestion in this case that, as was the case in Burton, the State profits from the TSSAA's decision to enforce its recruiting rule.

Because I do not believe that the TSSAA's action of enforcing its recruiting rule is fairly attributable to the State of Tennessee, I would affirm.

Although the TSSAA's enforcement activities cannot be considered state action as a matter of common sense or under any of this Court's existing theories of state action, the majority presents a new theory. Under this theory, the majority holds that the combination of factors it identifies evidences "entwinement" of the State with the TSSAA, and that such entwinement converts private action into state action. Ante, at 7-8. The majority does not define "entwinement," and the meaning of the term is not altogether clear. But whatever this new "entwinement" theory may entail, it lacks any support in our state-action jurisprudence. Although the majority asserts that there are three examples of entwinement analysis in our cases, there is no case in which we have rested a finding of state action on entwinement alone.

Two of the cases on which the majority relies do not even use the word "entwinement." See Lebron, supra, at 374; City Trusts, supra, at 230. Lebron concerned the status of Amtrak, a corporation that Congress created and placed under Government control for the specific purpose of achieving a governmental objective (namely to avert the threatened extinction of passenger train service in the United States). 513 U.S., at 383, 386. Without discussing any notion of entwinement, we simply held that, when "the Government creates a corporation by special law, for the furtherance of governmental objectives, and retains for itself permanent authority to appoint a majority of the directors of that corporation, the corporation is part of the Government for purposes of the First Amendment." Id., at 400. Similarly, in City Trusts, we did not consider entwinement when we addressed the question whether an agency established by state law was a state actor. See 353 U.S., at 231. In that case, the Pennsylvania legislature passed a law creating a board of directors to operate a racially segregated school for orphans. Ibid. Without mentioning "entwinement," we held that, because the board was a state agency, its actions were attributable to the State. Ibid.

The majority's third example, Evans v. Newton, 382 U.S. 296 (1966), lends no more support to an "entwinement" theory than do Lebron and City Trusts. Although Evans at least uses the word "entwined," 382 U.S., at 299 ("Conduct that is formally 'private' may become so entwined with governmental policies or so impregnated with a governmental character as to become subject to the constitutional limitations placed upon state action"), we did not discuss entwinement as a distinct concept, let alone one sufficient to transform a private entity into a state actor when traditional theories of state action do not. On the contrary, our analysis rested on the recognition that the subject of the dispute, a park, served a "public function," much like a fire department or a police department. Id., at 302. A park, we noted, is a "public facility" that "serves the community." Id., at 301-302. Even if the city severed all ties to the park and placed its operation in private hands, the park still would be "municipal in nature," analogous to other public facilities that have given rise to a finding of state action: the streets of a company town in Marsh v. Alabama, 326 U.S. 501 (1946), the elective process in Terry v. Adams, 345 U.S. 461 (1953), and the transit system in Public Utilities Comm'n of D. C. v. Pollak, 343 U.S. 451 (1952). 382 U.S., at 301-302. Because the park served public functions, the private trustees operating the park were considered to be state actors.6

These cases, therefore, cannot support the majority's "entwinement" theory. Only Evans speaks of entwinement at all, and it does not do so in the same broad sense as does the majority.7 Moreover, these cases do not suggest that the TSSAA's activities can be considered state action, whether the label for the state-action theory is "entwinement" or anything else.

* * *

Because the majority never defines "entwinement," the scope of its holding is unclear. If we are fortunate, the majority's fact-specific analysis will have little bearing beyond this case. But if the majority's new entwinement test develops in future years, it could affect many organizations that foster activities, enforce rules, and sponsor extracurricular competition among high schools-not just in athletics, but in such diverse areas as agriculture, mathematics, music, marching bands, forensics, and cheerleading. Indeed, this entwinement test may extend to other organizations that are composed of, or controlled by, public officials or public entities, such as firefighters, policemen, teachers, cities, or counties. I am not prepared to say that any private organization that permits public entities and public officials to participate acts as the State in anything or everything it does, and our state-action jurisprudence has never reached that far. The state-action doctrine was developed to reach only those actions that are truly attributable to the State, not to subject private citizens to the control of federal courts hearing 1983 actions.

I respectfully dissent.

In this case we are asked to decide whether the National Bank for Cooperatives, which Congress has designated as a federally chartered instrumentality of the United States, is exempt from state income taxation. We hold that it is not.

* In the Farm Credit Act of 1933, 48 Stat. 257, as amended, 12 U.S.C. 2001 et seq., Congress created various lending institutions within the Farm Credit System to meet the specific credit needs of farmers. Among these institutions were banks for cooperatives, one in each of 12 farm credit districts, and a Central Bank for Cooperatives. These banks were designed to make loans to cooperative associations engaged in marketing farm products, purchasing farm supplies, or furnishing farm services.

Today, the Farm Credit System includes banks for cooperatives, production credit associations, farm credit banks, and federal land bank associations. 2002(a). By statute, each of these institutions is designated as a "federally chartered instrumentalit[y] of the United States." 2121 (banks for cooperatives and Central Bank for Cooperatives); 2141 (National Bank for Cooperatives); 2071(a) and (b)(7) (production credit associations); 2011(a) (farm credit banks); 2091(a) and (b)(4) (federal land bank associations). The Farm Credit Act also addresses the taxation of these institutions. The provision applicable to a bank for cooperatives, the institution at issue in this case, states:

"Each bank for cooperatives and its obligations are instrumentalities of the United States and as such any and all notes, debentures, and other obligations issued by such bank shall be exempt, both as to principal and interest from all taxation (except surtaxes, estate, inheritance, and gift taxes) now or hereafter imposed by the United States or any State, territorial, or local taxing authority, except that interest on such obligations shall be subject to Federal income taxation in the hands of the holder." 2134.

Respondent CoBank ACB is the successor to all rights and obligations of the National Bank for Cooperatives, which had been formed in 1989 through the consolidation of 10 district banks for cooperatives and the Central Bank for Cooperatives.1 The National Bank for Cooperatives filed Missouri corporate income tax returns for the years 1991 through 1994 and paid the taxes shown on those returns. In March 1996, CoBank filed amended returns on behalf of the National Bank for Cooperatives, requesting an exemption from all state income taxes and refunds on the taxes paid-erroneously, it alleged-for 1991 through 1994. Relying on the doctrine of implied tax immunity that originated in McCulloch v. Maryland, 4 Wheat. 316 (1819), CoBank asserted that the Supremacy Clause of the Constitution accords federal instrumentalities immunity from state taxation unless Congress has expressly waived this immunity. CoBank argued that, because the current version of the Farm Credit Act does not expressly waive this immunity, banks for cooperatives are exempt from Missouri's corporate income tax. The Director of Revenue of Missouri denied the request.

On appeal, the Administrative Hearing Commission upheld the Director of Revenue's assessment of corporate income tax, because the National Bank for Cooperatives had not established that it was a federal instrumentality statutorily exempt from state taxation of its income. The commission determined that Congress did not provide expressly that banks for cooperatives, in contrast to farm credit banks and federal land bank associations, would have immunity from state income taxation. The commission reasoned that had Congress intended to confer upon banks for cooperatives the same immunity that was provided to farm credit banks and federal land bank associations, it would have done so expressly. For jurisdictional reasons, the commission did not decide CoBank's constitutional claim.

The Missouri Supreme Court reversed the commission's decision and held that banks for cooperatives are exempt from state income taxation.2 Production Credit Assn. of Southeastern Mo. v. Director of Revenue, 10 S. W. 3d 142, 143 (2000). The Missouri Supreme Court held that the Supremacy Clause of the Constitution provides federal instrumentalities immunity from state taxation unless Congress has expressly waived this immunity. According to the Missouri Supreme Court, because the current version of the Farm Credit Act is silent as to such institutions' immunity from state taxation, Congress cannot be said to have expressly consented to state income taxation and, thus, the institutions are exempt from state income taxes. The Missouri Supreme Court noted that other courts that had addressed the issue of state taxation of member institutions of the Farm Credit System also had concluded that the States could not tax such institutions. Id., at 143-144 (citing Farm Credit Servs. of Central Ark., PCA v. Arkansas, 76 F.3d 961, 964 (CA8 1996), rev'd on other grounds, 520 U.S. 821 (1997); State v. Farm Credit Servs. of Central Ark., 338 Ark. 322, 327, 994 S. W. 2d 453, 456 (1999), cert. denied, 529 U.S. 1036 (2000); Northwest La. Production Credit Assn. v. State, 98-1995 (La. App. 11/5/99), 746 So. 2d 280).

The New Mexico Court of Appeals and the Indiana Supreme Court have reached the opposite conclusion with respect to state taxation of production credit associations. See Production Credit Assn. of Eastern N. M. v. Taxation and Revenue Dept., 2000 NMCA-021 26, 999 P.2d 1031, 1038, cert. denied, 997 P.2d 820 (N. M. 2000); Indiana Dept. of State Revenue v. Farm Credit Servs. of Mid-America, ACA, 734 N. E. 2d 551, 560 (Ind. 2000). Since the statutory history and provisions regarding the taxation of production credit associations and banks for cooperatives are virtually identical, compare 12 U.S.C. 2077 with 2134; compare Farm Credit Act of 1971, 2.17, 85 Stat. 602, with 3.13, 85 Stat. 608; compare Farm Credit Amendments Act of 1985, 205(d)(16), 99 Stat. 1705, with 205(e)(10), 99 Stat. 1705,3 we granted certiorari to resolve this conflict. 530 U.S. 1260 (2000).

Congress has expressly designated banks for cooperatives as "instrumentalities of the United States." 12 U.S.C. 2121. We have held, in addressing state taxation of contractors conducting business with the United States, that an instrumentality is entitled to implied tax immunity only when it is "so closely connected to the Government that the two cannot realistically be viewed as separate entities." United States v. New Mexico, 455 U.S. 720, 735 (1982). Relying on New Mexico, the Director of Revenue argues that banks for cooperatives are not "so closely connected" to the United States as to be indistinguishable from the United States, and that banks for cooperatives thus are not entitled to immunity from state taxation. CoBank disagrees with this characterization and asks us to conclude that banks for cooperatives are indeed virtual arms of the United States, worthy of implied tax immunity under McCulloch.

We need not, however, reach this implied immunity question. Implied immunity becomes an issue only when Congress has failed to indicate whether an instrumentality is subject to state taxation. In this case, Congress has provided that banks for cooperatives are subject to state taxation. To be sure, Congress did not include an express statement in the current version of 2134. However, nothing in the statute indicates a repeal of the previous express approval of state taxation, and the structure of the Farm Credit Act indicates by negative implication that banks for cooperatives are not entitled to immunity.

* Upon their creation in 1933, banks for cooperatives were subject to state income taxation except during periods when the United States held stock in the banks. Farm Credit Act of 1933, 63, 48 Stat. 267 ("Such banks, ... and their income, shall be exempt from all taxation now or hereafter imposed by the United States or by any State, Territorial, or local taxing authority ... . The exemption provided herein shall not apply ... with respect to ... any ... Bank for Cooperatives, or its property or income after the stock held in it by the United States has been retired"). Under this statute, as soon as governmental investment in a bank for cooperatives was repaid (as it was for all such banks by 1968), the bank had to pay state income taxes because the exemption from such taxation no longer applied.

When Congress amended the Farm Credit Act in 1971, it did not change the rule that banks for cooperatives are subject to state taxation unless the United States holds stock in the banks. Farm Credit Act of 1971, 3.13, 85 Stat. 608. Although all banks for cooperatives were at the time privately owned, Congress provided that the Governor of the Farm Credit Administration had the authority on behalf of the United States to purchase stock in banks for cooperatives "as a temporary investment in the stock of the institution to help one or several of the banks ... to meet emergency credit needs of borrowers." 4.0, 85 Stat. 609. The 1971 version of 2134 therefore provided, in relevant part:

"Such banks ... and their income shall be exempt from all taxation now or hereafter imposed by the United States or by any State, territorial, or local taxing authority ... . The exemption provided in the preceding sentence shall apply only for any year or part thereof in which stock in the bank for cooperatives is held by the Governor of the Farm Credit Administration."

3.13, 85 Stat. 608-609.

In 1985, Congress enacted various amendments to the Act. Among other things, these amendments eliminated the position of Governor of the Farm Credit Administration, discontinued the Farm Credit Administration's authority to own stock in banks for cooperatives, and included numerous "Technical and Conforming Amendments." Farm Credit Amendments Act of 1985, 201, 99 Stat. 1688; 101, 99 Stat. 1678; 205, 99 Stat. 1703-1707. One of these technical and conforming amendments was the deletion of the two sentences within 2134 that, first, exempted a bank for cooperatives from state taxation and, second, limited that exemption to periods when the Governor held stock in the bank. 205(e)(10), 99 Stat. 1705, as amended, 12 U.S.C. 2134.

CoBank argues that the deletion of these two sentences altered the States' ability to tax the income of banks for cooperatives. According to CoBank, because the deletion eliminated the express statutory authorization for such taxation, Congress intended banks for cooperatives to be immune from state taxation under McCulloch's implied immunity doctrine. We do not share CoBank's interpretation as to the effect of this amendment, because there is no indication that Congress intended to change the taxation of banks for cooperatives with the 1985 amendments. Since 1933, the States could collect revenue from banks for cooperatives. Nothing in the 1985 amendments expressly changes this. And, it would be surprising, indeed, if Congress had eliminated this important fact sub silentio.

CoBank's interpretation would mean that Congress made a radical-but entirely implicit-change in the taxation of banks for cooperatives with the 1985 amendment to 2134. The amendment to 2134 was merely one of numerous "technical and conforming amendments" to the Farm Credit Act. Farm Credit Amendments Act of 1985, 205, 99 Stat. 1703-1707 (section entitled "Technical and Conforming Amendments"). In fact, the deletion of the sentence within 2134 referring to the Governor was one of more than 30 deletions of references to the Governor, a position eliminated by the 1985 amendments to the Act. Ibid. The more logical interpretation of this amendment to 2134 is that Congress merely deleted language that had become superfluous once the United States no longer owned, and no longer could own, stock in banks for cooperatives. This explanation accords with the more than 50-year history of the Farm Credit Act, permitting the States to tax banks for cooperatives except when there was governmental investment in the banks. Had Congress simply deleted the final sentence of 2134 that limited the exemption while retaining the sentence granting the exemption, we would have no trouble concluding that Congress had eliminated the States' ability to tax banks for cooperatives. Short of this act, however, we find Congress' silence insufficient to disrupt the 50-year history of state taxation of banks for cooperatives.

In addition, the structure of the Farm Credit Act confirms that banks for cooperatives are subject to state taxation. With respect to each lending institution in the Farm Credit System, the Act contains a taxation provision that specifically delineates the immunity from taxation enjoyed by that entity. For example, farm credit banks and federal land bank associations receive the type of immunity from state taxation that the Missouri Supreme Court held to be implied here for banks for cooperatives. See 12 U.S.C. 2023 ("The Farm Credit Banks and the capital, reserves, and surplus thereof, and the income derived therefrom, shall be exempt from Federal, State, municipal, and local taxation ..."); 2098 ("Each Federal land bank association and the capital, reserves, and surplus thereof, and the income derived therefrom, shall be exempt from Federal, State, municipal, and local taxation ...").

By contrast, since their creation in 1933, banks for cooperatives have been granted only limited exemptions from taxation. Had Congress intended to confer upon banks for cooperatives the more comprehensive exemption from taxation that it had provided to farm credit banks and federal land bank associations, it would have done so expressly as it had done elsewhere in the Farm Credit Act. Thus, in light of the structure of the Farm Credit Act-and the explicit grant of immunity to other institutions within the Farm Credit System-Congress' silence with respect to banks for cooperatives indicates that banks for cooperatives are subject to state taxation.

***

The judgment of the Missouri Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered.