SCALIA, J., DISSENTING
HAMDAN V. RUMSFELD
548 U. S. ____ (2006)
SUPREME COURT OF THE UNITED STATES
SALIM AHMED HAMDAN, PETITIONER v. DONALD H. RUMSFELD, SECRETARY OF DEFENSE, et al.
on writ of certiorari to the united states court of appeals for the district of columbia circuit
[June 29, 2006]
Justice Scalia, with whom Justice Thomas and Justice Alito join, dissenting.
On December 30, 2005, Congress enacted the Detainee Treatment Act (DTA). It unambiguously provides that, as of that date, “no court, justice, or judge” shall have jurisdiction to consider the habeas application of a Guantanamo Bay detainee. Notwithstanding this plain directive, the Court today concludes that, on what it calls the statute’s most natural reading, every “court, justice, or judge” before whom such a habeas application was pending on December 30 has jurisdiction to hear, consider, and render judgment on it. This conclusion is patently erroneous. And even if it were not, the jurisdiction supposedly retained should, in an exercise of sound equitable discretion, not be exercised.
The DTA provides: “[N]o court, justice, or judge shall have jurisdiction to hear or consider an application for a writ of habeas corpus filed by or on behalf of an alien detained by the Department of Defense at Guantanamo Bay, Cuba.” §1005(e)(1), 119 Stat. 2742 (internal division omitted). This provision “t[ook] effect on the date of the enactment of this Act,” §1005(h)(1), id., at 2743, which was December 30, 2005. As of that date, then, no court had jurisdiction to “hear or consider” the merits of petitioner’s habeas application. This repeal of jurisdiction is simply not ambiguous as between pending and future cases. It prohibits any exercise of jurisdiction, and it became effective as to all cases last December 30. It is also perfectly clear that the phrase “no court, justice, or judge” includes this Court and its Members, and that by exercising our appellate jurisdiction in this case we are “hear[ing] or consider[ing] … an application for a writ of habeas corpus.”
An ancient and unbroken line of authority attests that statutes ousting jurisdiction unambiguously apply to cases pending at their effective date. For example, in Bruner v. United States, 343 U. S. 112 (1952), we granted certiorari to consider whether the Tucker Act’s provision denying district court jurisdiction over suits by “officers” of the United States barred a suit by an employee of the United States. After we granted certiorari, Congress amended the Tucker Act by adding suits by “ ‘employees’ ” to the provision barring jurisdiction over suits by officers. Id., at 114. This statute narrowing the jurisdiction of the district courts “became effective” while the case was pending before us, ibid., and made no explicit reference to pending cases. Because the statute “did not reserve jurisdiction over pending cases,” id., at 115, we held that it clearly ousted jurisdiction over them. Summarizing centuries of practice, we said: “This rule—that, when a law conferring jurisdiction is repealed without any reservation as to pending cases, all cases fall with the law—has been adhered to consistently by this Court.” Id., at 116–117. See also Landgraf v. USI Film Products, 511 U. S. 244, 274 (1994) (opinion for the Court by Stevens, J.) (“We have regularly applied intervening statutes conferring or ousting jurisdiction, whether or not jurisdiction lay when the underlying conduct occurred or when the suit was filed”).
This venerable rule that statutes ousting jurisdiction terminate jurisdiction in pending cases is not, as today’s opinion for the Court would have it, a judge-made “presumption against jurisdiction,” ante, at 11, that we have invented to resolve an ambiguity in the statutes. It is simple recognition of the reality that the plain import of a statute repealing jurisdiction is to eliminate the power to consider and render judgment—in an already pending case no less than in a case yet to be filed.
“Without jurisdiction the court cannot proceed at all in any cause. Jurisdiction is power to declare the law, and when it ceases to exist, the only function remaining to the court is that of announcing the fact and dismissing the cause. And this is not less clear upon authority than upon principle.” Ex parte McCardle, 7 Wall. 506, 514 (1869) (emphasis added).
To alter this plain meaning, our cases have required an explicit reservation of pending cases in the jurisdiction-repealing statute. For example, Bruner, as mentioned, looked to whether Congress made “any reservation as to pending cases.” 343 U. S., at 116–117; see also id., at 115 (“Congress made no provision for cases pending at the effective date of the Act withdrawing jurisdiction and, for this reason, Courts of Appeals ordered pending cases terminated for want of jurisdiction”). Likewise, in Hallowell v. Commons, 239 U. S. 506 (1916), Justice Holmes relied on the fact that the jurisdiction-ousting provision “made no exception for pending litigation, but purported to be universal,” id., at 508. And in Insurance Co. v. Ritchie, 5 Wall. 541 (1867), we again relied on the fact that the jurisdictional repeal was made “without any saving of such causes as that before us,” id., at 544. As in Bruner, Hallowell, and Ritchie, the DTA’s directive that “no court, justice, or judge shall have jurisdiction,” §1005(e)(1), 119 Stat. 2742, is made “without any reservation as to pending cases” and “purport[s] to be universal.” What we stated in an earlier case remains true here: “[W]hen, if it had been the intention to confine the operation of [the jurisdictional repeal] … to cases not pending, it would have been so easy to have said so, we must presume that Congress meant the language employed should have its usual and ordinary signification, and that the old law should be unconditionally repealed.” Railroad Co. v. Grant, 98 U. S. 398, 403 (1879).
The Court claims that I “rea[d] too much into” the Bruner line of cases, ante, at 12, n. 7, and that “the Bruner rule” has never been “an inflexible trump,” ante, at 19. But the Court sorely misdescribes Bruner—as if it were a kind of early-day Lindh v. Murphy, 521 U. S. 320 (1997), resolving statutory ambiguity by oblique negative inference. On the contrary, as described above, Bruner stated its holding as an unqualified “rule,” which “has been adhered to consistently by this Court.” 343 U. S., at 116–117. Though Bruner referred to an express savings clause elsewhere in the statute, id., at 115, n. 7, it disavowed any reliance on such oblique indicators to vary the plain meaning, quoting Ritchie at length: “ ‘It is quite possible that this effect of the [jurisdiction-stripping statute] was not contemplated by Congress… . [B]ut when terms are unambiguous we may not speculate on probabilities of intention.’ ” 343 U. S., at 116 (quoting 5 Wall., at 544–545).
The Court also attempts to evade the Bruner line of cases by asserting that “the ‘presumption’ [of application to pending cases] that these cases have applied is more accurately viewed as the nonapplication of another presumption—viz., the presumption against retroactivity—in certain limited circumstances.” Ante, at 11. I have already explained that what the Court calls a “presumption” is simply the acknowledgment of the unambiguous meaning of such provisions. But even taking it to be what the Court says, the effect upon the present case would be the same. Prospective applications of a statute are “effective” upon the statute’s effective date; that is what an effective-date provision like §1005(h)(1) means. [Footnote 1] “ ‘[S]hall take effect upon enactment’ is presumed to mean ‘shall have prospective effect upon enactment,’ and that presumption is too strong to be overcome by any negative inference [drawn from other provisions of the statute].” Landgraf, 511 U. S., at 288 (Scalia, J., concurring in judgments). The Court’s “nonapplication of … the presumption against retroactivity” to §1005(e)(1) is thus just another way of stating that the statute takes immediate effect in pending cases.
Though the Court resists the Bruner rule, it cannot cite a single case in the history of Anglo-American law (before today) in which a jurisdiction-stripping provision was denied immediate effect in pending cases, absent an explicit statutory reservation. By contrast, the cases granting such immediate effect are legion, and they repeatedly rely on the plain language of the jurisdictional repeal as an “inflexible trump,” ante, at 19, by requiring an express reservation to save pending cases. See, e.g., Bruner, supra, at 115; Kline v. Burke Constr. Co., 260 U. S. 226, 234 (1922); Hallowell, 239 U. S., at 508; Gwin v. United States, 184 U. S. 669, 675 (1902); Gurnee v. Patrick County, 137 U. S. 141, 144 (1890); Sherman v. Grinnell, 123 U. S. 679, 680 (1887); Railroad Co. v. Grant, supra, at 403, Assessors v. Osbornes, 9 Wall. 567, 575 (1870); Ex parte McCardle, 7 Wall., at 514; Ritchie, supra, at 544; Norris v. Crocker, 13 How. 429, 440 (1852); Yeaton v. United States, 5 Cranch 281 (1809) (Marshall, C.J.), discussed in Gwin, supra, at 675; King v. Justices of the Peace of London, 3 Burr. 1456, 1457, 97 Eng. Rep. 924, 925 (K. B. 1764). Cf. National Exchange Bank of Baltimore v. Peters, 144 U. S. 570, 572 (1892).
Disregarding the plain meaning of §1005(e)(1) and the requirement of explicit exception set forth in the foregoing cases, the Court instead favors “a negative inference … from the exclusion of language from one statutory provision that is included in other provisions of the same statute,” ante, at 13. Specifically, it appeals to the fact that §1005(e)(2) and (e)(3) are explicitly made applicable to pending cases (by §1005(h)(2)). A negative inference of the sort the Court relies upon might clarify the meaning of an ambiguous provision, but since the meaning of §1005(e)(1) is entirely clear, the omitted language in that context would have been redundant.
Even if §1005(e)(1) were at all ambiguous in its application to pending cases, the “negative inference” from §1005(h)(2) touted by the Court would have no force. The numerous cases in the Bruner line would at least create a powerful default “presumption against jurisdiction,” ante, at 11. The negative inference urged by the Court would be a particularly awkward and indirect way of rebutting such a longstanding and consistent practice. This is especially true since the negative inference that might be drawn from §1005(h)(2)’s specification that certain provisions shall apply to pending cases is matched by a negative inference in the opposite direction that might be drawn from §1005(b)(2), which provides that certain provisions shall not apply to pending cases.
The Court’s reliance on our opinion in Lindh v. Murphy, 521 U. S. 320 (1997), is utterly misplaced. Lindh involved two provisions of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA): a set of amendments to chapter 153 of the federal habeas statute that redefined the scope of collateral review by federal habeas courts; and a provision creating a new chapter 154 in the habeas statute specially to govern federal collateral review of state capital cases. See 521 U. S., at 326–327. The latter provision explicitly rendered the new chapter 154 applicable to cases pending at the time of AEDPA’s enactment; the former made no specific reference to pending cases. Id., at 327. In Lindh, we drew a negative inference from chapter 154’s explicit reference to pending cases, to conclude that the chapter 153 amendments did not apply in pending cases. It was essential to our reasoning, however, that both provisions appeared to be identically difficult to classify under our retroactivity cases. First, we noted that, after Landgraf, there was reason for Congress to suppose that an explicit statement was required to render the amendments to chapter 154 applicable in pending cases, because the new chapter 154 “will have substantive as well as purely procedural effects.” 521 U. S., at 327. The next step—and the critical step—in our reasoning was that Congress had identical reason to suppose that an explicit statement would be required to apply the chapter 153 amendments to pending cases, but did not provide it. Id., at 329. The negative inference of Lindh rested on the fact that “[n]othing … but a different intent explain[ed] the different treatment.” Ibid.
Here, by contrast, there is ample reason for the different treatment. The exclusive-review provisions of the DTA, unlike both §1005(e)(1) and the AEDPA amendments in Lindh, confer new jurisdiction (in the D. C. Circuit) where there was none before. For better or for worse, our recent cases have contrasted jurisdiction-creating provisions with jurisdiction-ousting provisions, retaining the venerable rule that the latter are not retroactive even when applied in pending cases, but strongly indicating that the former are typically retroactive. For example, we stated in Hughes Aircraft Co. v. United States ex rel. Schumer, 520 U. S. 939, 951 (1997), that a statute “that creates jurisdiction where none previously existed” is “as much subject to our presumption against retroactivity as any other.” See also Republic of Austria v. Altmann, 541 U. S. 677, 695 (2004) (opinion for the Court by Stevens, J.); id., at 722 (Kennedy, J., dissenting). The Court gives our retroactivity jurisprudence a dazzling clarity in asserting that “subsections (e)(2) and (e)(3) ‘confer’ jurisdiction in a manner that cannot conceivably give rise to retroactivity questions under our precedents.” [Footnote 2] Ante, at 17–18. This statement rises to the level of sarcasm when one considers its author’s description of the governing test of our retroactivity jurisprudence:
“The conclusion that a particular rule operates ‘retroactively’ comes at the end of a process of judgment concerning the nature and extent of the change in the law and the degree of connection between the operation of the new rule and a relevant past event. Any test of retroactivity will leave room for disagreement in hard cases, and is unlikely to classify the enormous variety of legal changes with perfect philosophical clarity. However, retroactivity is a matter on which judges tend to have ‘sound … instinct[s],’ … and familiar considerations of fair notice, reasonable reliance, and settled expectations offer sound guidance.” Landgraf, 511 U. S., at 270 (opinion for the Court by Stevens, J.).
The only “familiar consideration,” “reasonable reliance,” and “settled expectation” I am aware of pertaining to the present case is the rule of Bruner—applicable to §1005(e)(1), but not to §1005(e)(2) and (3)—which the Court stubbornly disregards. It is utterly beyond question that §1005(e)(2)’s and (3)’s application to pending cases (without explicit specification) was not as clear as §1005(e)(1)’s. That is alone enough to explain the difference in treatment.
Another obvious reason for the specification was to stave off any Suspension Clause problems raised by the immediately effective ouster of jurisdiction brought about by subsection (e)(1). That is to say, specification of the immediate effectiveness of subsections (e)(2) and (e)(3) (which, unlike subsection (e)(1), would not fall within the Bruner rule and would not automatically be deemed applicable in pending cases) could reasonably have been thought essential to be sure of replacing the habeas jurisdiction that subsection (e)(1) eliminated in pending cases with an adequate substitute. See infra, at 16–18.
These considerations by no means prove that an explicit statement would be required to render subsections (e)(2) and (e)(3) applicable in pending cases. But they surely gave Congress ample reason to doubt that their application in pending cases would unfold as naturally as the Court glibly assumes. In any event, even if it were true that subsections (e)(2) and (e)(3) “ ‘confer’ jurisdiction in a manner that cannot conceivably give rise to retroactivity questions,” ante, at 17–18, this would merely establish that subsection (h)(2)’s reference to pending cases was wholly superfluous when applied to subsections (e)(2) and (e)(3), just as it would have been for subsection (e)(1). Lindh’s negative inference makes sense only when Congress would have perceived “the wisdom of being explicit” with respect to the immediate application of both of two statutory provisions, 521 U. S., at 328, but chose to be explicit only for one of them—not when it would have perceived no need to be explicit for both, but enacted a redundancy only for one.
In short, it is simply untrue that Congress “ ‘should have been just as concerned about’ ” specifying the application of §1005(e)(1) to pending cases, ante, at 14 (quoting Lindh, 521 U. S., at 329). In fact, the negative-inference approach of Lindh is particularly inappropriate in this case, because the negative inference from §1005(h)(2) would tend to defeat the purpose of the very provisions that are explicitly rendered applicable in pending cases, §1005(e)(2) and (3). Those provisions purport to vest “exclusive” jurisdiction in the D. C. Circuit to consider the claims raised by petitioners here. See infra, at 16–18. By drawing a negative inference À la Lindh, the Court supplants this exclusive-review mechanism with a dual-review mechanism for petitioners who were expeditious enough to file applications challenging the CSRTs or military commissions before December 30, 2005. Whatever the force of Lindh’s negative inference in other cases, it surely should not apply here to defeat the purpose of the very provision from which the negative inference is drawn.
Worst of all is the Court’s reliance on the legislative history of the DTA to buttress its implausible reading of §1005(e)(1). We have repeatedly held that such reliance is impermissible where, as here, the statutory language is unambiguous. But the Court nevertheless relies both on floor statements from the Senate and (quite heavily) on the drafting history of the DTA. To begin with floor statements: The Court urges that some “statements made by Senators preceding passage of the Act lend further support to” the Court’s interpretation, citing excerpts from the floor debate that support its view, ante, 15–16, n. 10. The Court immediately goes on to discount numerous floor statements by the DTA’s sponsors that flatly contradict its view, because “those statements appear to have been inserted into the Congressional Record after the Senate debate.” Ibid. Of course this observation, even if true, makes no difference unless one indulges the fantasy that Senate floor speeches are attended (like the Philippics of Demosthenes) by throngs of eager listeners, instead of being delivered (like Demosthenes’ practice sessions on the beach) alone into a vast emptiness. Whether the floor statements are spoken where no Senator hears, or written where no Senator reads, they represent at most the views of a single Senator. In any event, the Court greatly exaggerates the one-sidedness of the portions of the floor debate that clearly occurred before the DTA’s enactment. Some of the statements of Senator Graham, a sponsor of the bill, only make sense on the assumption that pending cases are covered. [Footnote 3] And at least one opponent of the DTA unmistakably expressed his understanding that it would terminate our jurisdiction in this very case. [Footnote 4] (Of course in its discussion of legislative history the Court wholly ignores the President’s signing statement, which explicitly set forth his understanding that the DTA ousted jurisdiction over pending cases. [Footnote 5])
But selectivity is not the greatest vice in the Court’s use of floor statements to resolve today’s case. These statements were made when Members of Congress were fully aware that our continuing jurisdiction over this very case was at issue. The question was divisive, and floor statements made on both sides were undoubtedly opportunistic and crafted solely for use in the briefs in this very litigation. See, e.g., 151 Cong. Rec. S14257–S14258 (Dec. 21, 2005) (statement of Sen. Levin) (arguing against a reading that would “stri[p] the Federal courts of jurisdiction to consider pending cases, including the Hamdan case now pending in the Supreme Court,” and urging that Lindh requires the same negative inference that the Court indulges today (emphasis added)). The Court’s reliance on such statements cannot avoid the appearance of similar opportunism. In a virtually identical context, the author of today’s opinion has written for the Court that “[t]he legislative history discloses some frankly partisan statements about the meaning of the final effective date language, but those statements cannot plausibly be read as reflecting any general agreement.” Landgraf, 511 U. S., at 262 (opinion for the Court by Stevens, J.). Likewise, the handful of floor statements that the Court treats as authoritative do not “reflec[t] any general agreement.” They reflect the now-common tactic—which the Court once again rewards—of pursuing through floor-speech ipse dixit what could not be achieved through the constitutionally prescribed method of putting language into a bill that a majority of both Houses vote for and the President signs.
With regard to the floor statements, at least the Court shows some semblance of seemly shame, tucking away its reference to them in a half-hearted footnote. Not so for its reliance on the DTA’s drafting history, which is displayed prominently, see ante, at 14–15. I have explained elsewhere that such drafting history is no more legitimate or reliable an indicator of the objective meaning of a statute than any other form of legislative history. This case presents a textbook example of its unreliability. The Court, ante, at 14, trumpets the fact that a bill considered in the Senate included redundant language, not included in the DTA as passed, reconfirming that the abolition of habeas jurisdiction “shall apply to any application or other action that is pending on or after the date of the enactment of this Act.” 151 Cong. Rec. S12655 (Nov. 10, 2005). But this earlier version of the bill also differed from the DTA in other material respects. Most notably, it provided for postdecision review by the D. C. Circuit only of the decisions of CSRTs, not military commissions, ibid.; and it limited that review to whether “the status determination … was consistent with the procedures and standards specified by the Secretary of Defense,” ibid., not whether “the use of such standards and procedures … is consistent with the Constitution and laws of the United States,” DTA §1005(e)(2)(C)(ii), 119 Stat. 2742. To say that what moved Senators to reject this earlier bill was the “action that is pending” provision surpasses the intuitive powers of even this Court’s greatest Justices. [Footnote 6] And to think that the House and the President also had this rejection firmly in mind is absurd. As always—but especially in the context of strident, partisan legislative conflict of the sort that characterized enactment of this legislation—the language of the statute that was actually passed by both Houses of Congress and signed by the President is our only authoritative and only reliable guidepost.
A final but powerful indication of the fact that the Court has made a mess of this statute is the nature of the consequences that ensue. Though this case concerns a habeas application challenging a trial by military commission, DTA §1005(e)(1) strips the courts of jurisdiction to hear or consider any “application for a writ of habeas corpus filed by or on behalf of an alien detained by the Department of Defense at Guantanamo Bay, Cuba.” The vast majority of pending petitions, no doubt, do not relate to military commissions at all, but to more commonly challenged aspects of “detention” such as the terms and conditions of confinement. See Rasul v. Bush, 542 U. S. 466, 498 (2004) (Scalia, J., dissenting). The Solicitor General represents that “[h]abeas petitions have been filed on behalf of a purported 600 [Guantanamo Bay] detainees,” including one that “seek[s] relief on behalf of every Guantanamo detainee who has not already filed an action,” Respondents’ Motion to Dismiss for Lack of Jurisdiction 20, n. 10 (hereinafter Motion to Dismiss). The Court’s interpretation transforms a provision abolishing jurisdiction over all Guantanamo-related habeas petitions into a provision that retains jurisdiction over cases sufficiently numerous to keep the courts busy for years to come.
Because I would hold that §1005(e)(1) unambiguously terminates the jurisdiction of all courts to “hear or consider” pending habeas applications, I must confront petitioner’s arguments that the provision, so interpreted, violates the Suspension Clause. This claim is easily dispatched. We stated in Johnson v. Eisentrager, 339 U. S. 763, 768 (1950):
“We are cited to no instance where a court, in this or any other country where the writ is known, has issued it on behalf of an alien enemy who, at no relevant time and in no stage of his captivity, has been within its territorial jurisdiction. Nothing in the text of the Constitution extends such a right, nor does anything in our statutes.”
Notwithstanding the ill-considered dicta in the Court’s opinion in Rasul, 542 U. S., at 480–481, it is clear that Guantanamo Bay, Cuba, is outside the sovereign “territorial jurisdiction” of the United States. See id., at 500–505 (Scalia, J., dissenting). Petitioner, an enemy alien detained abroad, has no rights under the Suspension Clause.
But even if petitioner were fully protected by the Clause, the DTA would create no suspension problem. This Court has repeatedly acknowledged that “the substitution of a collateral remedy which is neither inadequate nor ineffective to test the legality of a person’s detention does not constitute a suspension of the writ of habeas corpus.” Swain v. Pressley, 430 U. S. 372, 381 (1977); see also INS v. St. Cyr, 533 U. S. 289, 314, n. 38 (2006) (“Congress could, without raising any constitutional questions, provide an adequate substitute through the courts of appeals”).
Petitioner has made no showing that the postdecision exclusive review by the D. C. Circuit provided in §1005(e)(3) is inadequate to test the legality of his trial by military commission. His principal argument is that the exclusive-review provisions are inadequate because they foreclose review of the claims he raises here. Though petitioner’s brief does not parse the statutory language, his argument evidently rests on an erroneously narrow reading of DTA §1005(e)(3)(D)(ii), 119 Stat. 2743. That provision grants the D. C. Circuit authority to review, “to the extent the Constitution and laws of the United States are applicable, whether the use of such standards and procedures to reach the final decision is consistent with the Constitution and laws of the United States.” In the quoted text, the phrase “such standards and procedures” refers to “the standards and procedures specified in the military order referred to in subparagraph (A),” namely “Military Commission Order No. 1, dated August 31, 2005 (or any successor military order).” DTA §1005(e)(3)(D)(i), (e)(3)(A), ibid. This Military Commission Order (Order No. 1) is the Department of Defense’s fundamental implementing order for the President’s order authorizing trials by military commission. Order No. 1 establishes commissions, §2; delineates their jurisdiction, §3; provides for their officers, §4(A); provides for their prosecution and defense counsel, §4(B), (C); lays out all their procedures, both pretrial and trial, §5(A)–(P), §6(A)–(G); and provides for posttrial military review through the Secretary of Defense and the President, §6(H). In short, the “standards and procedures specified in” Order No. 1 include every aspect of the military commissions, including the fact of their existence and every respect in which they differ from courts-martial. Petitioner’s claims that the President lacks legal authority to try him before a military commission constitute claims that “the use of such standards and procedures,” as specified in Order No. 1, is “[in]consistent with the Constitution and laws of the United States,” DTA §1005(e)(3)(D)(ii), 119 Stat. 2743. The D. C. Circuit thus retains jurisdiction to consider these claims on postdecision review, and the Government does not dispute that the DTA leaves unaffected our certiorari jurisdiction under 28 U. S. C. §1254(1) to review the D. C. Circuit’s decisions. Motion to Dismiss 16, n. 8. Thus, the DTA merely defers our jurisdiction to consider petitioner’s claims; it does not eliminate that jurisdiction. It constitutes neither an “inadequate” nor an “ineffective” substitute for petitioner’s pending habeas application. [Footnote 7].
Though it does not squarely address the issue, the Court hints ominously that “the Government’s preferred reading” would “rais[e] grave questions about Congress’ authority to impinge upon this Court’s appellate jurisdiction, particularly in habeas cases.” Ante, at 10–11 (citing Ex parte Yerger, 8 Wall. 85 (1869); Felker v. Turpin, 518 U. S. 651 (1996); Durousseau v. United States, 6 Cranch 307 (1810); United States v. Klein, 13 Wall. 128 (1872); and Ex parte McCardle, 7 Wall. 506). It is not clear how there could be any such lurking questions, in light of the aptly named “Exceptions Clause” of Article III, §2, which, in making our appellate jurisdiction subject to “such Exceptions, and under such Regulations as the Congress shall make,” explicitly permits exactly what Congress has done here. But any doubt our prior cases might have created on this score is surely chimerical in this case. As just noted, the exclusive-review provisions provide a substitute for habeas review adequate to satisfy the Suspension Clause, which forbids the suspension of the writ of habeas corpus. A fortiori they provide a substitute adequate to satisfy any implied substantive limitations, whether real or imaginary, upon the Exceptions Clause, which authorizes such exceptions as §1005(e)(1).
Even if Congress had not clearly and constitutionally eliminated jurisdiction over this case, neither this Court nor the lower courts ought to exercise it. Traditionally, equitable principles govern both the exercise of habeas jurisdiction and the granting of the injunctive relief sought by petitioner. See Schlesinger v. Councilman, 420 U. S. 738, 754 (1975); Weinberger v. Romero-Barcelo, 456 U. S. 305, 311 (1982). In light of Congress’s provision of an alternate avenue for petitioner’s claims in §1005(e)(3), those equitable principles counsel that we abstain from exercising jurisdiction in this case.
In requesting abstention, the Government relies principally on Councilman, in which we abstained from considering a serviceman’s claim that his charge for marijuana possession was not sufficiently “service-connected” to trigger the subject-matter jurisdiction of the military courts-martial. See 420 U. S., at 740, 758. Admittedly, Councilman does not squarely control petitioner’s case, but it provides the closest analogue in our jurisprudence. As the Court describes, ante, at 21, Councilman “identifie[d] two considerations of comity that together favor[ed] abstention pending completion of ongoing court-martial proceedings against service personnel.” But the Court errs in finding these considerations inapplicable to this case. Both of them, and a third consideration not emphasized in Councilman, all cut in favor of abstention here.
First, the Court observes that Councilman rested in part on the fact that “military discipline and, therefore, the efficient operation of the Armed Forces are best served if the military justice system acts without regular interference from civilian courts,” and concludes that “Hamdan is not a member of our Nation’s Armed Forces, so concerns about military discipline do not apply.” Ante, at 22. This is true enough. But for some reason, the Court fails to make any inquiry into whether military commission trials might involve other “military necessities” or “unique military exigencies,” 420 U. S., at 757, comparable in gravity to those at stake in Councilman. To put this in context: The charge against the respondent in Councilman was the off-base possession and sale of marijuana while he was stationed in Fort Sill, Oklahoma, see id., at 739–740. The charge against the petitioner here is joining and actively abetting the murderous conspiracy that slaughtered thousands of innocent American civilians without warning on September 11, 2001. While Councilman held that the prosecution of the former charge involved “military necessities” counseling against our interference, the Court does not even ponder the same question for the latter charge.
The reason for the Court’s “blinkered study” of this question, ante, at 19, is not hard to fathom. The principal opinion on the merits makes clear that it does not believe that the trials by military commission involve any “military necessity” at all: “The charge’s shortcomings … are indicative of a broader inability on the Executive’s part here to satisfy the most basic precondition … for establishment of military commissions: military necessity.” Ante, at 48. This is quite at odds with the views on this subject expressed by our political branches. Because of “military necessity,” a joint session of Congress authorized the President to “use all necessary and appropriate force,” including military commissions, “against those nations, organizations, or persons [such as petitioner] he determines planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001.” Authorization for Use of Military Force, §2(a), 115 Stat. 224, note following 50 U. S. C. §1541 (2000 ed., Supp. III). In keeping with this authority, the President has determined that “[t]o protect the United States and its citizens, and for the effective conduct of military operations and prevention of terrorist attacks, it is necessary for individuals subject to this order … to be detained, and, when tried, to be tried for violations of the laws of war and other applicable laws by military tribunals.” Military Order of Nov. 13, 2001, 3 CFR §918(e) (2002). It is not clear where the Court derives the authority—or the audacity—to contradict this determination. If “military necessities” relating to “duty” and “discipline” required abstention in Councilman, supra, at 757, military necessities relating to the disabling, deterrence, and punishment of the mass-murdering terrorists of September 11 require abstention all the more here.
The Court further seeks to distinguish Councilman on the ground that “the tribunal convened to try Hamdan is not part of the integrated system of military courts, complete with independent review panels, that Congress has established.” Ante, at 22. To be sure, Councilman emphasized that “Congress created an integrated system of military courts and review procedures, a critical element of which is the Court of Military Appeals consisting of civilian judges completely removed from all military influence or persuasion, who would gain over time thorough familiarity with military problems.” 420 U. S., at 758 (internal quotation marks and footnote omitted). The Court contrasts this “integrated system” insulated from military influence with the review scheme established by Order No. 1, which “provides that appeal of a review panel’s decision may be had only to the Secretary of Defense himself, §6(H)(5), and then, finally, to the President, §6(H)(6).” Ante, at 23.
Even if we were to accept the Court’s extraordinary assumption that the President “lack[s] the structural insulation from military influence that characterizes the Court of Appeals for the Armed Forces,” ante, at 23, [Footnote 8] the Court’s description of the review scheme here is anachronistic. As of December 30, 2005, the “fina[l]” review of decisions by military commissions is now conducted by the D. C. Circuit pursuant to §1005(e)(3) of the DTA, and by this Court under 28 U. S. C. §1254(1). This provision for review by Article III courts creates, if anything, a review scheme more insulated from Executive control than that in Councilman. [Footnote 9] At the time we decided Councilman, Congress had not “conferred on any Art[icle] III court jurisdiction directly to review court-martial determinations.” 420 U. S., at 746. The final arbiter of direct appeals was the Court of Military Appeals (now the Court of Appeals for the Armed Forces), an Article I court whose members possessed neither life tenure, nor salary protection, nor the constitutional protection from removal provided to federal judges in Article III, §1. See 10 U. S. C. §867(a)(2) (1970 ed.).
Moreover, a third consideration counsels strongly in favor of abstention in this case. Councilman reasoned that the “considerations of comity, the necessity of respect for coordinate judicial systems” that motivated our decision in Younger v. Harris, 401 U. S. 37 (1971), were inapplicable to courts-martial, because “the particular demands of federalism are not implicated.” 420 U. S., at 756, 757. Though military commissions likewise do not implicate “the particular demands of federalism,” considerations of interbranch comity at the federal level weigh heavily against our exercise of equity jurisdiction in this case. Here, apparently for the first time in history, see Motion to Dismiss 6, a District Court enjoined ongoing military commission proceedings, which had been deemed “necessary” by the President “[t]o protect the United States and its citizens, and for the effective conduct of military operations and prevention of terrorist attacks.” Military Order of Nov. 13, 3 CFR §918(e). Such an order brings the Judicial Branch into direct conflict with the Executive in an area where the Executive’s competence is maximal and ours is virtually nonexistent. We should exercise our equitable discretion to avoid such conflict. Instead, the Court rushes headlong to meet it. Elsewhere, we have deferred exercising habeas jurisdiction until state courts have “the first opportunity to review” a petitioner’s claim, merely to “reduc[e] friction between the state and federal court systems.” O’Sullivan v. Boerckel, 526 U. S. 838, 844, 845 (1999). The “friction” created today between this Court and the Executive Branch is many times more serious.
In the face of such concerns, the Court relies heavily on Ex parte Quirin, 317 U. S. 1 (1942): “Far from abstaining pending the conclusion of military proceedings, which were ongoing, [in Quirin] we convened a special Term to hear the case and expedited our review.” Ante, at 24. It is likely that the Government in Quirin, unlike here, preferred a hasty resolution of the case in this Court, so that it could swiftly execute the sentences imposed, see Hamdi v. Rumsfeld, 542 U. S. 507, 569 (2004) (Scalia, J., dissenting). But the Court’s reliance on Quirin suffers from a more fundamental defect: Once again, it ignores the DTA, which creates an avenue for the consideration of petitioner’s claims that did not exist at the time of Quirin. Collateral application for habeas review was the only vehicle available. And there was no compelling reason to postpone consideration of the Quirin application until the termination of military proceedings, because the only cognizable claims presented were general challenges to the authority of the commissions that would not be affected by the specific proceedings. See supra, at 8–9, n. 2. In the DTA, by contrast, Congress has expanded the scope of Article III review and has channeled it exclusively through a single, postverdict appeal to Article III courts. Because Congress has created a novel unitary scheme of Article III review of military commissions that was absent in 1942, Quirin is no longer governing precedent.
I would abstain from exercising our equity jurisdiction, as the Government requests.
* * *
The Court apparently believes that the effective-date provision means nothing at all. “That paragraph (1), along with paragraphs (2) and (3), is to ‘take effect on the date of enactment,’ DTA §1005(h)(1), 119 Stat. 2743, is not dispositive,” says the Court, ante, at 14, n. 9. The Court’s authority for this conclusion is its quote from INS v. St. Cyr, 533 U. S. 289, 317 (2001), to the effect that “a statement that a statute will become effective on a certain date does not even arguably suggest that it has any application to conduct that occurred at an earlier date.” Ante, at 14, n. 9 (emphasis added, internal quotation marks omitted). But this quote merely restates the obvious: An effective-date provision does not render a statute applicable to “conduct that occurred at an earlier date,” but of course it renders the statute applicable to conduct that occurs on the effective date and all future dates—such as the Court’s exercise of jurisdiction here. The Court seems to suggest that, because the effective-date provision does not authorize retroactive application, it also fails to authorize prospective application (and is thus useless verbiage). This cannot be true.Footnote 2
A comparison with Lindh v. Murphy, 521 U. S. 320 (1997), shows this not to be true. Subsections (e)(2) and (e)(3) of §1005 resemble the provisions of AEDPA at issue in Lindh (whose retroactivity as applied to pending cases the Lindh majority did not rule upon, see 521 U. S., at 326), in that they “g[o] beyond ‘mere’ procedure,” id., at 327. They impose novel and unprecedented disabilities on the Executive Branch in its conduct of military affairs. Subsection (e)(2) imposes judicial review on the Combatant Status Review Tribunals (CSRTs), whose implementing order did not subject them to review by Article III courts. See Memorandum from Deputy Secretary of Defense Paul Wolfowitz re: Order Establishing Com- batant Status Review Tribunals, at 3 §h (July 7, 2004), avail- able at http://www.defenselink.mil/news/Jul2004/d20040707review.pdf (all Internet materials as visited June 27, 2006, and available in Clerk of Court’s case file). Subsection (e)(3) authorizes the D. C. Circuit to review “the validity of any final decision rendered pursuant to Military Commission Order No. 1,” §1005(e)(3)(A), 119 Stat. 2743. Historically, federal courts have never reviewed the validity of the final decision of any military commission; their jurisdiction has been restricted to considering the commission’s “lawful authority to hear, decide and condemn,” In re Yamashita, 327 U. S. 1, 8 (1946) (emphasis added). See also Johnson v. Eisentrager, 339 U. S. 763, 786–787 (1950). Thus, contrary to the Court’s suggestion, ante, at 17, subsections (e)(2) and (e)(3) confer new jurisdiction: They impose judicial oversight on a traditionally unreviewable exercise of military authority by the Commander in Chief. They arguably “spea[k] not just to the power of a particular court but to . . . substantive rights . . . as well,” Hughes Aircraft Co. v. United States ex rel. Shumer, 520 U. S. 939, 951 (1997)—namely, the unreviewable powers of the President. Our recent cases had reiterated that the Executive is protected by the presumption against retroactivity in such comparatively trivial contexts as suits for tax refunds and increased pay, see Landgraf v. USI Film Products, 511 U. S. 244, 271, n. 25 (1994).Footnote 3
“Because I have described how outrageous these claims are—about the exercise regime, the reading materials—most Americans would be highly offended to know that terrorists are suing us in our own courts about what they read.” 151 Cong. Rec. S12756 (Nov. 14, 2005). “Instead of having unlimited habeas corpus opportunities under the Constitution, we give every enemy combatant, all 500, a chance to go to Federal court, the Circuit Court of Appeals for the District of Columbia… . It will be a one-time deal.” Id., at S12754. “This Levin-Graham-Kyl amendment allows every detainee under our control to have their day in court. They are allowed to appeal their convictions.” Id., at S12801 (Nov. 15, 2005); see also id., at S12799 (rejecting the notion that “an enemy combatant terrorist al-Qaida member should be able to have access to our Federal courts under habeas like an American citizen”).Footnote 4
“An earlier part of the amendment provides that no court, justice, or judge shall have jurisdiction to consider the application for writ of habeas corpus… . Under the language of exclusive jurisdiction in the DC Circuit, the U. S. Supreme Court would not have jurisdiction to hear the Hamdan case … .” Id., at S12796 (statement of Sen. Specter).Footnote 5
“[T]he executive branch shall construe section 1005 to preclude the Federal courts from exercising subject matter jurisdiction over any existing or future action, including applications for writs of habeas corpus, described in section 1005.” President’s Statement on Signing of H. R. 2863, the “Department of Defense, Emergency Supple- mental Appropriations to Address Hurricanes in the Gulf of Mexico, and Pandemic Influenza Act, 2006” (Dec. 30, 2005), available at http://www.whitehouse.gov/news/releases/2005/12/print/20051230 8.php.Footnote 6
The Court asserts that “it cannot be said that the changes to subsection (h)(2) were inconsequential,” ante, at 15, n. 10, but the Court’s sole evidence is the self-serving floor statements that it selectively cites.Footnote 7
Petitioner also urges that he could be subject to indefinite delay if military officials and the President are deliberately dilatory in reviewing the decision of his commission. In reviewing the constitutionality of legislation, we generally presume that the Executive will implement its provisions in good faith. And it is unclear in any event that delay would inflict any injury on petitioner, who (after an adverse determination by his CSRT, see 344 F. Supp. 2d 152, 161 (DC 2004)) is already subject to indefinite detention under our decision in Hamdi v. Rumsfeld, 542 U. S. 507 (2004). Moreover, the mere possibility of delay does not render an alternative remedy “inadequate [o]r ineffective to test the legality” of a military commission trial. Swain v. Pressley, 430 U. S. 372, 381 (1977). In an analogous context, we discounted the notion that postponement of relief until postconviction review inflicted any cognizable injury on a serviceman charged before a military court-martial. Schlesinger v. Councilman, 420 U. S. 738, 754–755 (1975); see also Younger v. Harris, 401 U. S. 37, 46 (1971).Footnote 8
The very purpose of Article II’s creation of a civilian Commander in Chief in the President of the United States was to generate “structural insulation from military influence.” See The Federalist No. 28 (A. Hamilton); id., No. 69 (same). We do not live under a military junta. It is a disservice to both those in the Armed Forces and the President to suggest that the President is subject to the undue control of the military.Footnote 9
In rejecting our analysis, the Court observes that appeals to the D. C. Circuit under subsection (e)(3) are discretionary, rather than as of right, when the military commission imposes a sentence less than 10 years’ imprisonment, see ante, at 23, n. 19, 52–53; §1005(e)(3)(B), 119 Stat. 2743. The relevance of this observation to the abstention question is unfathomable. The fact that Article III review is discretionary does not mean that it lacks “structural insulation from military influence,” ante, at 23, and its discretionary nature presents no obstacle to the courts’ future review these cases.
The Court might more cogently have relied on the discretionary nature of review to argue that the statute provides an inadequate substitute for habeas review under the Suspension Clause. See supra, at 16–18. But this argument would have no force, even if all appeals to the D. C. Circuit were discretionary. The exercise of habeas jurisdiction has traditionally been entirely a matter of the court’s equitable discretion, see Withrow v. Williams, 507 U. S. 680, 715–718 (1993) (Scalia, J., concurring in part and dissenting in part), so the fact that habeas jurisdiction is replaced by discretionary appellate review does not render the substitution “inadequate.” Swain, 430 U. S., at 381.