Appellant James Dwight Thomas, a state prisoner, appeals the district court's denial of his petition for a writ of habeas corpus. We affirm.

In 1973, Appellant pled nolo contendere in Florida state court to second-degree murder and was sentenced to 85 years' imprisonment. He subsequently escaped from a Florida prison and fled to New York, where he remained a fugitive for close to 17 years. In 1993, Appellant was extradited to Florida to serve the remainder of his original sentence.

The Florida Parole Commission (FPC) considered Appellant's case in 1994 and 1997, and ultimately set a presumptive parole release date of October 20, 2017. Appellant challenged the proceedings before the FPC. Of particular relevance to this case, Appellant sought a writ of mandamus in Leon County Circuit Court and then a writ of habeas corpus in Calhoun County Circuit Court. The Leon County Circuit Court denied Appellant's petition for a writ of mandamus. Appellant did not appeal that decision. The Calhoun County Circuit Court denied Appellant's petition for a writ of habeas corpus, concluding Appellant was procedurally barred from relitigating issues that had already been raised in his mandamus petition and denying Appellant's remaining claims on the merits. Appellant unsuccessfully appealed that decision.

In February 1999, Appellant sought post-conviction relief in federal court by filing pro se a document captioned "FORM TO BE USED BY FEDERAL PRISONERS FILING A PETITION FOR WRIT OF HABEAS CORPUS PURSUANT TO 28 U.S.C. § 2241." The district court referred the petition to a magistrate judge, who noted that "[a]lthough petitioner styled his case as under 28 U.S.C. § 2241, since he alleges he is in custody pursuant to a state judgment in violation of his constitutional rights, his appropriate avenue for relief is under 28 U.S.C. § 2254."

The magistrate judge recommended that two of Appellant's four claims be denied under exhaustion and procedural bar principles and that Appellant's remaining two claims be denied on the merits. The district court adopted the magistrate judge's report and recommendation and denied Appellant's petition for habeas relief.

On July 5, 2001, we issued a certificate of appealability under 28 U.S.C. § 2253 on the following issue:

Whether the district court erred in finding that appellant's first two claims that he was denied due process and equal protection under the law by the Florida Parole Commission in the setting of his presumptive parole release date and in refusing to set an effective parole release date were procedurally barred in light of O'Sullivan v. Boerckel, 526 U.S. 838, 119 S. Ct. 1728, 144 L. Ed. 2d 1 (1999).

After this issue was briefed, however, we sua sponte vacated the July 5 certificate of appealability and issued a new one on the following issue only:

Whether the district court erred in converting petitioner's application for a writ of habeas corpus pursuant to 28 U.S.C. § 2241 into an application for a writ of habeas corpus pursuant to 28 U.S.C. § 2254 in that petitioner's application did not collaterally attack a state court conviction. Instead, the application attacked a decision of the Florida Parole Commission concerning petitioner's parole eligibility.

We now hold Appellant's petition for a writ of habeas corpus is governed by both 28 U.S.C. § 2241 and 28 U.S.C. § 2254.

Before addressing the merits of the issue set out in the second COA, we briefly address the State's contention that the issue is not properly before the Court because Appellant never raised it.

Notably, there is no question we have jurisdiction in this case. "A COA is usually a jurisdictional prerequisite to an appeal in a post-conviction relief proceeding following a state or federal court conviction." Pagan v. United States, 353 F.3d 1343, 1344-45 (11th Cir.2003); see also 28 U.S.C. § 2253(c)(1) ("Unless a circuit justice or judge issues a certificate of appealability, an appeal may not be taken to the court of appeals...."). In this case, the first COA gave us jurisdiction over the appeal. It is of no consequence that we subsequently asked the parties to proceed on a different issue. Even assuming we did err by sua sponte issuing the second COA, any such error was not jurisdictional. See Young v. United States, 124 F.3d 794, 799 (7th Cir.1997) (refusing to treat an erroneously issued COA the same as the lack of a COA).

Moreover, we will not now consider nonjurisdictional challenges to the propriety of our issuance of the second COA. The entire point of § 2253's COA requirement is to eliminate those appeals that have little or no merit, thereby preserving judicial resources. See Miller-El v. Cockrell, 537 U.S. 322, 337, 123 S. Ct. 1029, 1040, 154 L. Ed. 2d 931 (2003) ("By enacting [the AEDPA COA requirement] ... Congress confirmed the necessity and the requirement of differential treatment for those appeals deserving of attention from those that plainly do not."). Once the parties have briefed and argued the issue set out in a COA and we have reached the point of considering an appeal on the merits, the time for scrutinizing the COA has long since passed. Were we to entertain belated challenges to COAs, we effectively would add another layer of complexity to the already complicated regime for post-conviction relief. See Young, 124 F.3d at 799 ("An obligation to determine whether a certificate should have been issued, even if the parties do not present this issue for decision ... would increase the complexity of appeals in collateral attacks and the judicial effort required to resolve them, the opposite of the legislative plan."). This we will not do.

The State's primary complaint regarding the second COA is that Appellant did not raise the issue set out therein and therefore has waived it. As we rule in the State's favor on the merits based on prior Circuit precedent, we need not resolve the waiver issue.

The issue set out in the second COA — whether the district court erred by treating the petition filed by Appellant, a state prisoner in custody pursuant to the judgment of a state court, as one under 28 U.S.C. § 2254 rather than one under 28 U.S.C. § 2241 — has been resolved by our intervening decision in Medberry v. Crosby, 351 F.3d 1049 (11th Cir.2003), cert. denied, ___ U.S. ___, 124 S. Ct. 2098, 158 L. Ed. 2d 714 (2004).

In Medberry, we held that a habeas petition filed by a state prisoner in custody pursuant to the judgment of a state court is subject both to § 2241 and to § 2254, with its attendant restrictions.1 In deciding Medberry, we conducted a searching review of habeas and post-conviction relief law. Although there is no need to repeat here all that we said in Medberry, we will for the sake of completeness repeat some of the more significant points from that decision.

"[T]here are two distinct means of securing post-conviction relief in the federal courts: an application for a writ of habeas corpus (governed by, inter alia, §§ 2241 and 2254) and a motion to vacate a sentence (governed by § 2255)." Medberry, 351 F.3d at 1058. With respect to the writ of habeas corpus proper, we stated in Medberry:

[T]he writ of habeas corpus is a single post-conviction remedy principally governed by two different statutes....

The difference between the statutes lies in the breadth of the situations to which they apply. Section 2241 provides that a writ of habeas corpus may issue to a prisoner in the following five situations:

(1) He is in custody under or by color of the authority of the United States or is committed for trial before some court thereof; or

(2) He is in custody for an act done or omitted in pursuance of an Act of Congress, or an order, process, judgment or decree of a court or judge of the United States; or

(3) He is in custody in violation of the Constitution or laws or treaties of the United States; or

(4) He, being a citizen of a foreign state and domiciled therein is in custody for an act done or omitted under any alleged right, title, authority, privilege, protection, or exemption claimed under the commission, order or sanction of any foreign state, or under color thereof, the validity and effect of which depend upon the law of nations; or

(5) It is necessary to bring him into court to testify or for trial.

28 U.S.C. § 2241(c). Section 2254, on the other hand, applies to a subset of those to whom § 2241(c)(3) applies — it applies to "a person in custody pursuant to the judgment of a State court" who is "in custody in violation of the Constitution or laws or treaties of the United States." 28 U.S.C. § 2254(a) (emphasis added).

Id. at 1059. Accordingly, we concluded in Medberry that where a prisoner is in custody pursuant to the judgment of a state court, his petition is subject to both § 2241 and § 2254. Id. at 1064.

In so ruling, we pointed out that our conclusion was supported by the canon of statutory construction that the more specific statute takes precedence over the more general statute. Id. at 1060. "`[B]oth Sections 2241 and 2254 authorize [petitioner's] challenge to the legality of his continued state custody,' but ... allowing him to file his `petition in federal court pursuant to Section 2241 without reliance on Section 2254 would ... thwart Congressional intent.'" Id. (quoting Coady v. Vaughn, 251 F.3d 480, 484-85 (3d Cir.2001)). We next pointed out that our conclusion was further supported by the canon of statutory construction against reading any provision (even any word) of a statute so as to make it superfluous. Id. We stated:

Our reading of §§ 2241 and 2254 as governing a single post-conviction remedy, with the § 2254 requirements applying to petitions brought by a state prisoner in custody pursuant to the judgment of a State court, gives meaning to § 2254 without rendering § 2241(c)(3) superfluous. Under our reading, there remain some state prisoners to whom § 2254 does not apply. Section 2254 is limited to state prisoners "in custody pursuant to the judgment of a State court." 28 U.S.C. § 2254(a). State pre-trial detention, for example, might violate the Constitution or the laws or treaties of the United States. Yet a person held in such pre-trial detention would not be "in custody pursuant to the judgment of a State court." Such a prisoner would file an application for a writ of habeas corpus governed by § 2241 only. To read §§ 2241 and 2254 other than as we do would effectively render § 2254 meaningless because state prisoners could bypass its requirements by proceeding under § 2241.

If § 2254 were not a restriction on § 2241's authority to grant the writ of habeas corpus, and were instead a freestanding, alternative post-conviction remedy, then § 2254 would serve no function at all. It would be a complete dead letter, because no state prisoner would choose to run the gauntlet of § 2254 restrictions when he could avoid those limitations simply by writing "§ 2241" on his petition for federal post-conviction relief. All of Congress's time and effort in enacting § 2254, amending it in 1966, and further amending it in 1996 with AEDPA would have been a complete waste. Section 2254 would never be used or applied, and all of the thousands of decisions over the past half-century from the Supreme Court and other federal courts interpreting and applying the provisions of § 2254 would have been pointless. Section 2254 would be a great irrelevancy because a state prisoner could simply opt out of its operation by choosing a different label for his petition.

Id. at 1060-61.

Our conclusion in Medberry was also supported by our extensive review of the history of the post-conviction relief statutes:

Neither does the history of § 2254 suggest that the section is anything more than a limitation on the preexisting authority under § 2241(c)(3) to grant the writ of habeas corpus to state prisoners. The 1948 codification which created § 2254 merely codified judge-made restrictions on issuing the writ of habeas corpus as authorized under § 2241.... Because it was merely declarative of judicial limitations imposed on habeas relief under § 2241, § 2254 could not possibly have created a new post-conviction remedy. Section 2254(a) merely specifies the class of state prisoners to which the additional restrictions of § 2254 apply.

Id. at 1060.

[A] state prisoner seeking post-conviction relief from a federal court has but one remedy: an application for a writ of habeas corpus. All applications for writs of habeas corpus are governed by § 2241, which generally authorizes federal courts to grant the writ — to both federal and state prisoners. Most state prisoners' applications for writs of habeas corpus are subject also to the additional restrictions of § 2254. That is, if a state prisoner is "in custody pursuant to the judgment of a State court," his petition is subject to § 2254. If, however, a prisoner is in prison pursuant to something other than a judgment of a state court, e.g., a pre-trial bond order, then his petition is not subject to § 2254.

Id. at 1062. Insofar as we held a state prisoner in custody pursuant to the judgment of a state court is subject to § 2254 and its various restrictions, we were in agreement with the Second, Third, Seventh, and Eighth Circuits. Id. (citing Cook v. New York State Div. of Parole, 321 F.3d 274, 278-79 (2d Cir.2003); Crouch v. Norris, 251 F.3d 720, 723 (8th Cir.2001); Coady, 251 F.3d 480, 484-85; Walker v. O'Brien, 216 F.3d 626, 633 (7th Cir.2000)).

There is no merit to Appellant's argument that § 2254 does not apply where a state prisoner challenges parole decisions rather than court rulings. Section 2254 is triggered where a prisoner is "in custody pursuant to the judgment of a State court." 28 U.S.C. § 2254(a). Appellant is in custody pursuant to the judgment of the Florida court. Therefore § 2254 applies to Appellant's petition.

A state prisoner cannot evade the procedural requirements of § 2254 by filing something purporting to be a § 2241 petition. If the terms of § 2254 apply to a state habeas petitioner — i.e., if he is "in custody pursuant to the judgment of a State court" — then we must apply its requirements to him. Appellant's case is hardly unusual. His petition for a writ of habeas corpus is governed by both § 2241 and § 2254. The district court did not err in applying § 2254 to Appellant's petition.

Appellant's use of the § 2241 form for federal prisoners raised an issue of whether he could somehow proceed under § 2241 separate and apart from § 2254. We apply our decision in Medberry v. Crosby, 351 F.3d 1049 (11th Cir.2003), and hold that Appellant's petition is subject to both § 2241 and § 2254. The district court therefore did not err in applying § 2254 to Appellant's petition.

AFFIRMED.

This case presents the important question of whether state prisoners may bring habeas corpus petitions under 28 U.S.C. § 2241 in order to avoid the restrictions of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), Pub.L. No. 104-132, 110 Stat. 1214 (1996), that unquestionably apply to petitions brought under 28 U.S.C. § 2254. The majority gives this issue extremely short shrift. Instead of addressing the question on its merits — and responding to the substantive critiques I have outlined in this special concurrence — the majority rests its conclusion entirely on one case, Medberry v. Crosby, 351 F.3d 1049 (11th Cir.2003). This appeal also raises significant questions about the power of circuit courts to consider matters sua sponte, both in general and in the specific context of habeas proceedings.

Because I do not believe the majority has set forth the facts of this case in sufficient detail to consider this appeal adequately, I will begin in Section A by offering an independent statement of facts that traces Thomas's journey through the Florida and New York State criminal justice systems. Section B turns to the procedural history of this case, explaining the gravamen of his federal habeas corpus petition, which the district court denied, and discussing our sua sponte grant of a revised certificate of appealability ("COA").

The facts of this case are, for the most part, uncontested. On May 11, 1973, petitioner James Thomas pled nolo contendere to second-degree murder before the Broward County, Florida, Circuit Court and was sentenced to eighty-five years in state prison. A few years later, in 1976, Thomas escaped from prison and fled to New York, where he hid for close to seventeen years. During this time, he was convicted in a New York state court of grand larceny, attempted unauthorized use of a motor vehicle, and resisting arrest while on escape status.1 In late 1993, Thomas was again arrested in New York and extradited to Florida to serve out his original second-degree murder sentence.2

On May 11, 1994, at Thomas's first parole hearing, the Florida Parole Commission ("FPC") set Thomas's presumptive parole release date ("PPRD") as October 20, 2019.3 The FPC reached this result by initially setting Thomas's PPRD 120 months away from his effective incarceration date,4 as required by Florida's standard parole determination matrix, then moving that date back an additional 240 months due to the "aggravat[ing]" factor of Thomas's New York convictions.5 On December 19, 1994, Thomas petitioned the FPC to review its PPRD determination, but the FPC rejected his application as untimely.6

In 1996, still dissatisfied with his PPRD, Thomas petitioned the Leon County Circuit Court for a writ of mandamus directed to the FPC. First, he contended that the FPC should have set an "effective parole release date" ("EPRD") for him, rather than merely a PPRD.7 Second, he claimed that the FPC erred in pushing back his PPRD 240 months based on his New York convictions.8 He based these claims on Florida law and did not expressly allege any federal constitutional violations.

At Thomas's second parole hearing, on March 5, 1997, while his mandamus petition was pending, the FPC again considered his PPRD and decided to move it up two years to October 20, 2017. A few weeks later, the circuit court denied Thomas's petition, summarily concluding without explanation that "[Thomas's] PPRD was properly calculated" and that the FPC "properly assessed a 240 month aggravation ... based on [his] New York convictions." Thomas v. Fla. Parole Comm'n, No. 96-02827, at *1 (Fla. Leon County Cir. Ct. May 6, 1997). The court further held that Thomas "failed to demonstrate any error or abuse of discretion on behalf of the [FPC]." Id. Thomas did not appeal this ruling.

On October 1, 1997, the FPC held a third parole hearing for Thomas at which a representative of the Broward County State Attorney's Office testified against further reductions in Thomas's parole. Following this hearing, the FPC declined to accelerate Thomas's PPRD. Citing a statute that had been enacted after Thomas's initial conviction, the FPC also decided that his next parole interview should not be held for five years.

A month later, Thomas petitioned the Leon County Circuit Court for a writ of habeas corpus. Citing the Fourteenth Amendment's Due Process Clause, he recast in federal constitutional terms the allegations raised in his mandamus petition concerning the calculation of his PPRD at his first parole hearing. Specifically, he argued that Florida law required the FPC to establish an EPRD instead of merely a PPRD. Moreover, he maintained that Florida law expressly prevented the FPC from considering his New York convictions, since none of them resulted in his incarceration for more than 60 days,9 and, in the alternative, because he had lived in New York for a period of ten years without any convictions.10 Thomas claimed that by ignoring the clear mandates of its own regulations, the FPC denied him due process of law and treated him in an arbitrary and capricious manner.

The habeas petition also raised several new constitutional claims concerning Thomas's third parole hearing. He argued that the FPC violated the Ex Post Facto Clause in scheduling his next interview for 2002 under the new parole statute, instead of in 1999, as it would have been under the statute as it existed at the time of his conviction. Moreover, he claimed that the FPC denied him "full and fair consideration" by permitting the representative from the State Attorney's Office to testify.

Because Thomas was incarcerated in Calhoun County, the Leon County Circuit Court transferred his habeas petition to the Calhoun County Circuit Court, which rejected the petition. Thomas v. Fla. Parole Comm'n, No. 97-0385 (Fla. Calhoun County Cir. Ct. Jan. 27, 1998). The court found that Thomas was procedurally barred from relitigating the issues that had originally been raised in his mandamus petition concerning the FPC's calculation of his PPRD at his first parole hearing. Id. at *2. It essentially ignored his Ex Post Facto argument arising from his third parole hearing, stating only that the FPC is "authorize[d] ... to use its discretion in determining how subsequent interviews will be scheduled." Id. The court concluded by briefly noting that it was "proper" for the representative from the State Attorney's Office to testify at his third parole hearing. Id.

Thomas thereafter petitioned the Florida District Court of Appeal for a writ of certiorari, challenging the circuit court's conclusion that his claims regarding the FPC's calculation of his PPRD at his first parole hearing were procedurally barred. On October 28, 1998, the district court of appeal denied the writ per curiam, without opinion. Thomas filed a petition for review to the Florida Supreme Court, which was likewise rejected.

On February 23, 1999, Thomas filed the complaint now before us, a petition for a writ of habeas corpus under 28 U.S.C. § 2241. The petition asserted four claims. First, the FPC denied Thomas due process and equal protection of the laws under the Fourteenth Amendment's Due Process Clause through the manner in which it calculated his PPRD at his first parole hearing. Second, the FPC violated these same constitutional provisions in failing to establish an EPRD for him at that hearing. Third, the FPC violated the Ex Post Facto Clause at his third parole hearing by scheduling his next hearing for 2002, rather than 1999. Finally, the FPC violated the Fourteenth Amendment in using false information in establishing his PPRD and permitting the representative from the State Attorney's Office to testify at his third hearing.

The district court referred Thomas's petition to a magistrate judge, who sua sponte considered the petition as seeking a writ of habeas under 28 U.S.C. § 2254, applied § 2254's standards, and rejected Thomas's claims. Thomas's arguments about his PPRD, as well as his claim about the FPC's alleged use of false information about him, were held to be procedurally barred under 28 U.S.C. § 2254(b)(1)(A) because the Leon County Circuit Court had rejected them in denying his 1996 petition for a writ of mandamus, and he had failed to appeal its adverse ruling. The state court's rulings concerning the Ex Post Facto Clause and the testimony of the State Attorney's representative at his third hearing were upheld because they did not involve "unreasonable application[s] of clearly established federal law." The district court adopted the magistrate judge's recommendations and rejected Thomas's claims.

On July 5, 2001, this court issued a COA under 28 U.S.C. § 2253 on the following issues:

Whether the district court erred in finding that appellant's first two claims that he was denied due process and equal protection of the law by the Florida Parole Commission in the setting of his presumptive parole release date and in refusing to set an effective parole release date were procedurally barred in light of O'Sullivan v. Boerckel, 526 U.S. 838, 119 S. Ct. 1728, 144 L. Ed. 2d 1 (1999).

After this issue was briefed, this court sua sponte vacated this COA and issued a new one on the issue of

[w]hether the district court erred in converting petitioner's application for a writ of habeas corpus pursuant to 28 U.S.C. § 2241 into an application for a writ of habeas corpus pursuant to 28 U.S.C. § 2254 in that petitioner's application did not collaterally attack a state court conviction. Instead, the application attacked a decision of the Florida Parole Commission concerning petitioner's parole eligibility.

Part II of this opinion addresses the FPC's contention that we acted improperly in sua sponte issuing a COA on this issue, in light of the fact that neither side raised the issue either before the district court or on appeal. Part III turns to the substance of the §§ 2241/2254 issue, explaining the relationship between §§ 2241 and 2254 and concluding that Thomas had the right to have his petition assessed under the former rather than the latter statute. Part IV concludes that while the standards for habeas petitions under §§ 2241 and 2254 differ in many respects, the exhaustion requirement crafted through the common law for § 2241 is similar to § 2254's statutorily created exhaustion requirement (28 U.S.C. § 2254(b)(1)(A)), and so Thomas's claim was properly dismissed as procedurally barred.

As stated earlier, the COA we originally granted Thomas was replaced by a COA presenting the issue of

[w]hether the district court erred in converting petitioner's application for a writ of habeas corpus pursuant to 28 U.S.C. § 2241 into an application for a writ of habeas corpus pursuant to 28 U.S.C. § 2254 in that petitioner's application did not collaterally attack a state court conviction. Instead, the application attacked a decision of the Florida Parole Commission concerning petitioner's parole eligibility.

This issue was thereafter briefed by both sides and heard in oral argument. In its brief, however, FPC challenges this new COA on two grounds. First, FPC points out that "[a]ppellant [Thomas] never raised th[is] issue ... in his initial round of briefs in this case ... and although the issue may be one of interest to bench and bar, as a matter of law the issue is not properly before this Court.... [C]laims not presented to the Court of Appeals are considered waived." Appellee's Supplemental Answer Br. at 28-29. I address this point in Subpart A.

In Subpart B, I respond to FPC's contention that we lacked the power to consider the §§ 2241/2254 issue because Thomas failed to raise it in the district court. Having established in these first two Subparts our power to grant a COA sua sponte in a habeas case on issues a petitioner failed to raise either in the district court or in his request for a COA, Subpart C explains why it was appropriate for us to exercise that power in this case. Subpart D briefly concludes. Before delving into this discussion, however, it is worth noting that many circuits would not even entertain FPC's attack on the second COA. As Judge Easterbrook writes, for example, "[O]nce the briefs have been written and the case heard, there is little point in scrutinizing the certificate of appealability. An obligation to determine whether a certificate should have been issued ... would increase the complexity of appeals in collateral attacks and the judicial effort required to resolve them...." Young v. United States, 124 F.3d 794, 799 (7th Cir.1997); see also Tiedeman v. Benson, 122 F.3d 518, 522 (8th Cir.1997) ("[W]e are fully informed about the merits, and it would make no sense to go through the unnecessary step of remanding to the District Court with the request that an issue or issues be specified [in a COA], when we already know, having fully considered the case, what we think the result ought to be.").

Even if this court had erred in issuing the second COA, it is not clear that we should consequently ignore the substantive legal issue it raises. The consensus among the circuits seems to be that "an erroneously issued COA ... in a federal habeas proceeding is different from the absence of one and thus does not disable the court of appeals from proceeding to the merits once the briefs have been written." Gatlin v. Madding, 189 F.3d 882, 886 (9th Cir.1999); see also United States v. Talk, 158 F.3d 1064, 1068 (10th Cir.1998) ("[A]n erroneously-issued certificate does not deprive us of jurisdiction to hear a certified appeal."). I am reluctant to follow this approach because COAs are jurisdictional, and I do not think an error in granting a COA can allow us to expand the scope of our jurisdiction. On the other hand, I do not wish to authorize meta-litigation in habeas cases over whether a COA was properly issued, particularly because the question in reviewing a grant of a COA — whether there has been a substantial showing that a prisoner has been denied a constitutional right — is so similar to the substantive question in the underlying habeas appeal — whether a prisoner has been denied a constitutional right. In this particular case, in an effort to guide future judges in handling habeas cases, I will demonstrate why we acted properly in sua sponte issuing the second COA before turning to the substance of Thomas's claims.

The FPC's first major claim is that we lack the power to grant sua sponte a COA in a habeas case concerning an issue for which the petitioner did not request one. Put another way, the FPC argues that, in the habeas context, a petitioner's failure to raise an issue on appeal precludes us from considering it. I discuss in Section 1 the general inherent power of federal circuit courts of appeals to consider matters sua sponte whether or not presented to them by the parties. In Section 2, I explain how this power is not curtailed by any of AEDPA's limitations on appeals in habeas cases. Section 3 shows that district courts have widely been acknowledged to have the power to grant COAs sua sponte, and demonstrates that there is no reason to deny circuit courts a similar power. Section 4 discusses how my conclusion is consistent with that of the majority of circuits to have considered this issue.

FPC argues, essentially, that we should not consider the §§ 2241/2254 issue because Thomas waived it by not raising it on appeal. Waiver, however, applies to the right of a litigant to have his claim heard. As the cases cited by the FPC demonstrate, a litigant in a habeas case can waive his right to have a matter considered by this court by failing to raise it before us. See, e.g., Johnson Enters. of Jacksonville v. FPL Group, Inc., 162 F.3d 1290, 1308 n. 43 (11th Cir.1998) ("[Petitioner] challenged other rulings of the district court in its notice of appeal; however, it briefed only those issues listed above. Claims not presented to the court of appeals are considered waived."). While such cases address the general inability of an appellant to ask a court of appeals to consider issues that were not raised in a notice of appeal or brief, they do not address the power of appellate courts to raise important issues sua sponte, as happened here.

The scope of a petitioner's rights has no bearing on this court's power. It is beyond dispute that, in general, we have the power to consider issues that a party fails to raise on appeal, even though the petitioner does not have the right to demand such consideration. See United States v. Boyd, 208 F.3d 638, 652 (7th Cir.2000) ("As we did [in previous cases], and as the Ninth Circuit did in [a previous case], so too have other circuits acknowledged that they may, when justice requires it, raise critical issues of law sua sponte."). While this power is most frequently exercised in the plain error context, see, e.g., United States v. Ardley, 273 F.3d 991, 1003-04 (2001) (en banc) (Tjoflat, J., dissenting), it extends beyond such situations. Satterfield v. Sigmon, No. 95-2411, 91 F.3d 131, 1996 WL 383922, at *3, 1996 U.S.App. LEXIS 16419, at *7 (4th Cir. Mar. 7 1996) ("Generally, federal appellate courts do not consider issues that the parties have not raised either below or on appeal. Yet in `exceptional circumstances,' to prevent injustice, we may raise issues sua sponte."). Thus, this case ultimately involves a question of our power, not of Thomas's rights or whether he waived them.

The conditions under which a court may consider a nonjurisdictional matter sua sponte have never adequately been considered and have caused a great deal of confusion among jurists. For example, Justice Scalia once pointed out that "the refusal to consider arguments not raised is a sound prudential practice, rather than a statutory or constitutional mandate, and there are times when prudence dictates the contrary." Davis v. United States, 512 U.S. 452, 464, 114 S. Ct. 2350, 2358, 129 L. Ed. 2d 362 (1994) (Scalia, J., concurring). Barely two years earlier, however, Justice Scalia had emphasized, "The rule that points not argued will not be considered is more than just a prudential rule of convenience; its observance, at least in the vast majority of cases, distinguishes our adversary system of justice from the inquisitorial one." United States v. Burke, 504 U.S. 229, 246, 112 S. Ct. 1867, 1877, 119 L. Ed. 2d 34 (1992) (Scalia, J., concurring). In Trest v. Cain, 522 U.S. 87, 89-90, 118 S. Ct. 478, 480, 139 L. Ed. 2d 444 (1997), one of the only Supreme Court cases to address this issue directly, the Court declined to decide whether a court of appeals in a habeas case may sua sponte consider the issue of a defendant's procedural default in state court, and instead held only that circuit courts are not required to do so.

The Supreme Court's clearest statement on this matter came in Singleton v. Wulff, where it held, "The matter of what questions may be taken up and resolved for the first time on appeal is one left primarily to the discretion of the courts of appeals, to be exercised on the facts of individual cases." 428 U.S. 106, 121, 96 S. Ct. 2868, 2877, 49 L. Ed. 2d 826 (1976); see also Cruz v. Melecio, 204 F.3d 14, 22 n. 7 (1st Cir.2000) ("Notwithstanding that the parties did not raise the issues that impel us to this course either to the district court or on this appeal, we have the power to do so sua sponte."); United States v. Heater, 63 F.3d 311, 332 (4th Cir.1995) ("Although [petitioner] did not raise the Ex Post Facto argument himself, we find it within our discretion to consider this constitutional concern sua sponte.").

Consequently, the fact that Thomas did not mention the §§ 2241/2254 issue in his COA petition does not bar us from exercising our discretion to rule sua sponte on this critical threshold issue.

Having established the general power of federal circuit courts of appeal to consider matters sua sponte, it is now necessary to determine whether there is something peculiar about the specific context of habeas corpus that prevents us from exercising that power in such appeals. Neither AEDPA nor the Federal Rules of Appellate Procedure can responsibly be read as stripping us of this discretion. Subsection (a) offers a plain-meaning interpretation of the pertinent statutes and rules. Subsection (b) comes to the same conclusion based on substantive canons of statutory construction. Subsection (c) discusses how this conclusion is entirely consistent with both the precedents and practices of this circuit.

a.

28 U.S.C. § 2253(c)(1) states that a circuit court may not hear an appeal in a habeas case from a state prisoner "[u]nless a circuit justice or judge issues a certificate of appealability." The law goes on to explain that a COA may be granted "only if the applicant has made a substantial showing of the denial of a constitutional right." Id. § 2253(c)(2). This statute concludes by stating that the COA must specify the issues that meet this standard. Id. § 2253(c)(3).

One can argue that because Thomas failed to raise the §§ 2241/2254 issue for which the second COA was issued, he never made a "substantial showing of the denial of a constitutional right" as required by § 2253(c)(2). I decline to read the statute so narrowly. The main focus of § 2253, in my opinion, is to deter prisoners from bringing numerous frivolous claims and to limit appeals to cases involving possible constitutional violations. See Joint Explanatory Statement of the Comm. of Conference, H. Rep. No. 104-518, at 111 (1996), reprinted in 1996 U.S.C.C.A.N. 944, 944 ("This title incorporates reforms to curb the abuse of the statutory writ of habeas corpus...."). Nothing in the text or legislative history of the statute suggests that Congress was concerned with whether it was the petitioner or the court who detected a potential constitutional error. Consequently, I do not read § 2253(c)(2) as prohibiting the court of appeals from issuing COAs based on potential constitutional problems it notices on its own. In the rare case like this one, where an appellate court sua sponte questions the district court's resolution of an issue that touches, even indirectly, upon constitutional rights, § 2253(c)(2)'s requirement that a "substantial showing of the denial of a constitutional right" be made before a COA may issue is met.

The Federal Rules of Appellate Procedure are more problematic. Rule 22(b) states, "If an applicant files a notice of appeal, the district judge who rendered the judgment must either issue a certificate of appealability or state why a certificate should not issue.... If the district judge has denied the certificate, the applicant may request a circuit judge to issue the certificate." Fed. R.App. P. 22(b)(1). The logical import of this provision seems to be that a circuit judge may not issue a COA unless and until a district judge has denied it. See United States v. Mitchell, 216 F.3d 1126, 1130 (D.C.Cir.2000) ("Rule 22(b) requires initial application in the district court for a COA before the court of appeals acts on a COA request."). In Hunter v. United States, we unanimously interpreted this Rule as follows:

Only if the district judge who rendered the judgment has declined to issue the certificate does a circuit judge come into the picture. Under the plain language of the rule, an applicant for the writ gets two bites at the appeal certificate apple: one before the district judge, and if that one is unsuccessful, he gets a second one before a circuit judge.

101 F.3d 1565, 1575 (11th Cir.1996) (en banc).

Of course, we may suspend the rules of appellate procedure for "good cause." Fed. R.App. P. 2. Nevertheless, it is unnecessary to resort to such an extreme measure. I read Rule 22 as requiring a court of appeals to reject a petitioner's application for a COA concerning issues that were not presented in the petitioner's application to the district court. I do not, however, view this rule as restricting the discretion of the court of appeals to grant a certificate based on issues it identifies on its own.

b.

Even moving beyond a plain-text meaning of § 2253(c) to apply substantive canons of statutory interpretation, we arrive at the same conclusion — AEDPA does not reduce the otherwise broad power of a circuit court to entertain matters sua sponte that a petitioner failed to raise on appeal. In interpreting a statute, we may not presume that Congress implicitly intends to strip a coordinate branch of Government of one of its longstanding inherent powers. Chambers v. NASCO, Inc., 501 U.S. 32, 47, 111 S. Ct. 2123, 2134, 115 L. Ed. 2d 27 (1991) ("`[W]e do not lightly assume that Congress has intended to depart from established principles' such as the scope of a court's inherent power." (quoting Weinberger v. Romero-Barcelo, 456 U.S. 305, 313, 102 S. Ct. 1798, 1803, 72 L. Ed. 2d 91 (1982))).

Section 2253(c)(2) states, "A certificate of appealability may issue ... only if the applicant has made a substantial showing of the denial of a constitutional right." As discussed above, federal courts have the inherent discretionary power to raise issues sua sponte that the parties neglected to raise on appeal. See supra Section I.A.1. Section 2253(c)'s language is insufficiently clear or explicit to divest the courts of appeals of this traditional power. Indeed, if Congress had intended to divest us of this power, it did so surprisingly indirectly. Cf. Williams v. United States, 458 U.S. 279, 287, 102 S. Ct. 3088, 3093, 73 L. Ed. 2d 767, 774 (1982) ("Yet, if Congress really set out to enact a national bad check law in § 1014, it did so with a peculiar choice of language and in an unusually backhanded manner."). In the absence of a clear statement (or even any real indication at all) that Congress intended to curtail our ability to consider issues on appeal sua sponte, the canons of statutory construction counsel against construing § 2253(c) as absolutely restricting appellate courts to issues mentioned by a prisoner in his COA petition.

c.

Even putting aside the above analysis, our cases establish the power of our court to add issues to a COA sua sponte. At the time Congress amended § 2253 to require appellants in habeas cases to obtain COAs instead of "certificates of probable cause" ("CPCs"), as the statute previously required, we were faced with several pending appeals that were based on CPCs. These CPCs did not meet the new requirements of the amended § 2253(c)(3) because they did not list the specific constitutional violations for which the prisoners were seeking review.

We recognized in Peoples v. Haley that an appellate panel, faced with a CPC, had the authority to specify the issues on which appeal may be taken, thereby transforming the deficient CPC into a COA that met the requirements of § 2253(c)(3). 227 F.3d 1342, 1346 (11th Cir.2000) ("[I]t is within the discretion of the court of appeals whether to apply the COA standards itself, or remand to the district court."); see also Putman v. Head, 268 F.3d 1223, 1228 (11th Cir.2001) ("[I]n exercising our discretion, we may either remand to the district court with instructions to enumerate the issues, or we may rule which issues raised by the petitioner warrant a COA.").

Exercising this discretion, we have most frequently chosen simply to amend a CPC so that it conforms to § 2253's new requirements, rather than dismissing the appeal or remanding the CPC to the district court. See, e.g., Eagle v. Linahan, 279 F.3d 926, 936 (11th Cir.2001) ("In so `fixing' Eagle's CPC, we necessarily limit the claims properly before us to those that we deem satisfy the AEDPA standards for appeal..."); Franklin v. Hightower, 215 F.3d 1196, 1199 (11th Cir.2000) ("By applying AEDPA's standards to this appeal and issuing a proper COA (if warranted), this panel may `fix' the inadequacies of the present CPC."); Putman, 268 F.3d at 1228 ("In this case, we choose to decide ourselves which issues, if any, are worthy of a COA.").11 Thus, although a court of appeals is necessarily "limited to the issues specified" in the COA, Murray v. United States, 145 F.3d 1249, 1250 (11th Cir.1998), there is no reason why we may not amend the certificate to add issues sua sponte.

Moving beyond the specific issue of sua sponte amending COAs, this circuit has also shown a tremendous willingness to consider sua sponte arguments favorable to the Government in habeas appeals that were not raised in the Government's briefs. In Horsley v. Alabama, 45 F.3d 1486, 1492 n. 10 (11th Cir.1995), we stated, "[E]ven if the State has abandoned this argument we have the discretion to overlook a failure to argue harmlessness and to undertake sua sponte the task of considering harmlessness." Similarly, in Housel v. Head, 238 F.3d 1289, 1297 (11th Cir.2001) (internal citations omitted), we held, "This nonretroactivity rule, born in Teague v. Lane, is a threshold issue, and one that we have discretion to raise sua sponte." See also Ochran v. United States, 117 F.3d 495, 503 (11th Cir.1997) (recognizing the power of appellate panels to consider sua sponte whether a prisoner has waived certain claims, despite the fact that the State did not advance an argument based on waiver); Acosta v. Artuz, 221 F.3d 117, 121 (2d Cir.2000) (recognizing "the authority of courts to raise sua sponte affirmative defenses where the defense implicates values beyond the interests of the parties"). Since we consider sua sponte nonjurisdictional arguments favorable to the State, it would seem grossly unfair to deny prisoners similar consideration.

Another reason to believe that appellate courts may sua sponte amend COAs is that district courts are widely recognized to have this power. See, e.g., United States v. Brown, 305 F.3d 304, 305-06 (5th Cir.2002) ("The district court nevertheless determined that if Apprendi was held applicable on collateral review, Brown's maximum sentence would be capped at five years ... and it therefore sua sponte granted Brown a certificate of appealability (COA) on the issue [of] whether Teague barred consideration of Apprendi claims in an initial § 2255 motion."); Clay v. United States, 30 Fed.Appx. 607, 608 (7th Cir.2002) ("The district court sua sponte issued a certificate of appealability as to whether [the defendant] was denied his Sixth Amendment right to the effective assistance of counsel."), rev'd on other grounds 537 U.S. 522, 123 S. Ct. 1072, 155 L. Ed. 2d 88 (2003); Dunn v. Colleran, 247 F.3d 450, 456 (3d Cir.2001) ("Finding, however, that [its] conclusion was `by no means free of doubt,' the District Court sua sponte granted a certificate of appealability."); Deleon v. Strack, 234 F.3d 84, 87 (2d Cir.2000) (upholding district court's decision to issue a COA sua sponte).

A district court's power to grant a COA sua sponte seems to be implied by its power to sua sponte deny one. See, e.g., Alexander v. Johnson, 211 F.3d 895, 898 (5th Cir.2000) ("It is perfectly lawful for district court's [sic] to deny COA [sic] sua sponte."). It would seem grossly unfair to conclude that a district court may preemptively reject a COA petition of its own volition, but may never grant one on its own. Such a conclusion would tilt the scales of justice too far in favor of the state and disrupt the delicate equipoise in which parties stand before the bar of justice.

Admittedly, there are important differences between district and circuit courts. See id. at 898 ("Arguably, the district court ... is in the best position to determine whether the petitioner has made a substantial showing of a denial of a constitutional right on the issues before that court. Further briefing and argument on the very issues the court has just ruled on would be repetitious."). A district court may sua sponte grant or deny a COA at the same time it rules on the merits of a habeas petition or rejects it on procedural grounds. This is arguably the best time for a district judge to decide this matter because the issues are still fresh in his mind.

Nevertheless, the same federal statute authorizes both district and circuit courts to issue COAs. See 28 U.S.C. § 2253(c)(1) ("Unless a circuit justice or judge issues a certificate of appealability, an appeal may not be taken to the court of appeals from — the final order in a habeas corpus proceeding in which the detention complained of arises out of process issued by a State court...."). It would be odd to interpret the statute as implicitly allowing one set of courts authorized to issue COAs to do so sua sponte, but prohibiting the other set of courts from doing so.

The only case I have been able to find that questions the power of a district court to grant sua sponte a COA, Berthoff v. United States, 308 F.3d 124 (1st Cir.2002), is procedurally bizarre. Berthoff was tried and convicted of a drug conspiracy and several other related charges. The district court sentenced him to twenty-one years in prison. One of Berthoff's co-conspirators, who the district court described as "the individual next to Berthoff most culpable in this conspiracy," pled guilty, cooperated with the government, and received a sentence of only three years. Berthoff, 308 F.3d at 126.

Berthoff sought a writ of habeas corpus claiming ineffective assistance of counsel because his attorney never informed him of the plea "overtures" made by the Government. The district court denied the petition, but then sua sponte granted Berthoff a COA concerning the disparity in sentencing between him and the defendants who pled guilty. Apparently, however, the parties nevertheless briefed the ineffective assistance of counsel claim for the court of appeals. The First Circuit vacated the COA and remanded the case to the district court with instructions to "clarify the issue warranting a COA." Id. at 126.

While on remand, Berthoff sought to amend his habeas petition to add an Apprendi claim. The district court declined to issue a COA on either the ineffective assistance of counsel issue or the Apprendi issue. It did, however, issue another COA on the sentencing disparity issue, allowing Berthoff to argue that the 700% differential between his sentence and that of the defendants who pled guilty unconstitutionally burdened the free exercise of his right to a jury trial. Strangely enough, when it granted this COA, the district court declared:

[I]n an appropriate case, where the government has engaged in illegal fact bargaining with one defendant, I would not hesitate to hold that a defendant similarly situated in all material respects could take advantage of the fact bargain in order freely to exercise the right to trial by jury guaranteed by the Sixth Amendment. This is not such a case.

Id. at 128 (emphasis added). Thus, the district court sua sponte granted a COA on an issue that neither party had raised and that the court itself declared was irrelevant to the facts of the underlying case. As the First Circuit emphasized, "[T]his case simply is an inappropriate vehicle for the district court's concerns.... Accordingly, we do not reach the merits of the constitutional issue at this time." Id. at 129.

Thus, Berthoff does not undermine the power of district courts to issue COAs sua sponte on questions a petitioners fails to raise; it instead prevents district courts from issuing COAs on issues that are simply not present in a particular case. Having established the generally recognized authority of district courts to issue COAs sua sponte, we would be hard pressed to decline to recognize the power of circuit courts of appeals to do so as well.

The majority of circuit courts to have addressed this issue agree with my conclusion that circuit courts may grant COAs sua sponte. In Mack v. Holt, the Sixth Circuit noted, for example, that it had

granted [the petitioner's] motion for a certificate of appealability as to whether the trial court properly determined that the waiver of his right to counsel was knowing and intelligent. This court also sua sponte granted a certificate of appealability as to whether [the petitioner's] § 2254 petition to the district court was timely filed.

62 Fed.Appx. 577, 578 (6th Cir.2003). The Eighth Circuit, in United States v. Morgan, 244 F.3d 674, 675 (8th Cir.2001), went even further, holding that its precedents "should not be read as foreclosing the right of an Eighth Circuit hearing panel to exercise its discretion to consider sua sponte issues beyond those specified in a certificate of appealability, whether the certificate was issued by a district court or by an administrative panel of this court." Judge Bye's separate opinion in that case fully endorses this sentiment, stating, "Today's order explains that [prior Eighth Circuit cases] do not prevent a hearing panel from exercising its discretion to consider uncertified issues sua sponte. I fully agree with this approach...." Id. (Bye, J., concurring in part and dissenting in part).

Both Morgan and the result I advocate in the instant case achieve the same substantive result: review by a court of appeals of an issue in a habeas case that had not been raised by the prisoner in his application for a COA. Nevertheless, I believe my approach is preferable because it respects § 2253(c)'s requirement that issues be contained in a COA before being considered on appeal. Under my approach, an appellate panel goes through the formalism of amending the COA before considering the merits of an issue the prisoner failed to raise. Nevertheless, the Eighth Circuit's holding lends support to the conclusion that a circuit panel may raise issues sua sponte in habeas appeals.

The Third Circuit appears to be the only other appellate court to have addressed this issue. That court, in dicta, offered an interpretation of § 2253(c) that, at first, appears to contradict my conclusion. In Hubley v. Superintendent, the court stated,

A "certificate of appealability in a case brought under § 2253(c)(2) may issue, in the literal language of the statute, only if the applicant has made a substantial showing of the denial of a constitutional right."... Where an applicant fails to make such a showing, "we do not have jurisdiction to review the merits of Appellant's case."

57 Fed.Appx. 927, 929 (3d Cir.2003) (quoting United States v. Cepero, 224 F.3d 256, 267, 268 (3d Cir.2000); other internal citations omitted).

The Hubley court did not base its ruling on this observation. Moreover, the case from which it quoted this proposition, Cepero, put no special emphasis whatsoever on the fact that it was specifically the prisoner, rather than the court, who raised the issue of the possible constitutional violation. Thus, not only is this isolated fragment from Hubley dicta, it is built around a quoted passage that has been taken entirely out of context. Moreover, the Hubley dicta is arguably inconsistent with the Third Circuit's previous recognition of the power of district courts to grant COAs sua sponte. See, e.g., Robinson v. Johnson, 313 F.3d 128, 133 n. 2 (3d Cir.2002) ("Although Robinson moved in the Court of Appeals for a certificate of appealability, it was the District Court which granted the request. This is not a problem since the District Court may grant sua sponte a certificate of appealability.").

Finally, Hubley is easily distinguishable from the instant case. In Hubley, the prisoner received a COA on certain issues, yet briefed additional issues, asking the Third Circuit to grant a COA concerning those additional issues and rule on their merits. The Hubley court properly rejected this invitation, stating,

Having complied with the Rules of this Court, [the State] would therefore be prejudiced by not having briefed the merits of this new issue. An eleventh hour certificate of appealability is therefore not appropriate. Contrary to petitioner's argument, it is not the stated "practice" of this Court to grant, sua sponte, certificates of appealability whenever a habeas petitioner wishes to raise a new argument on appeal.

57 Fed.Appx. at 931. The Hubley court came to the correct conclusion. To allow prisoners to decide for themselves to include arguments in their briefs other than those specified in their COAs would eviscerate § 2253(c). A clear ban on this practice prevents the State from needlessly having to brief responses to issues not included in the COA.

The above holding declares only that the Third Circuit will not amend a COA "whenever a habeas petitioner wishes to raise a new argument on appeal." Id. (emphasis added). This declaration does not address the issue of whether the panel has the power to amend a COA sua sponte prior to the parties' briefing of the appeal because the panel felt a particular issue should be addressed. A sua sponte amendment to a COA by the court of appeals, particularly when the parties are given the opportunity to brief the new issue, would not result in the unfairness that troubled the Hubley court. Thus, persuasive precedent from other circuits indicates that we have the power to amend a COA sua sponte to consider issues that the prisoner did not raise in his COA petition.

In light of these persuasive and wide-ranging arguments, I conclude that this court has the power, even in habeas cases, to issue or amend a COA sua sponte to address constitutional issues that a petitioner fails to present in his petition for a COA.

FPC's second waiver-related argument is that we should not have sua sponte issued a COA on the §§ 2241/2254 issue because Thomas failed to present it to the district court. As with issues a petitioner fails to raise on appeal, all this means is that Thomas waived his right to have us consider it; his failure does not affect the discretionary power of this court to choose to do so.

This point is best made in First Ala. Bank of Montgomery, N.A. v. First State Insurance Co., 899 F.2d 1045, 1060 n. 8 (11th Cir.1990), which states that the "[appellant's] failure to press the argument before the district court foreclosed its right to present it on appeal." However, in the same footnote, we immediately added, "[O]ur rule foreclosing review of issues not presented below is not a jurisdictional limitation but is instead a rule which may be abrogated in our sound discretion.... The case law in this area indicates that we hear issues otherwise waived only in instances where strict application of the rule would result in patently unjust results." Id. (citations omitted). Similarly, while Noritake Co. v. M/V Hellenic Champion states that "[a]s a general principle of appellate review, this court will not consider a legal issue or theory that was not presented to the trial court," it immediately goes on to stress that this rule does not apply if a "miscarriage of justice" will occur. 627 F.2d 724, 732 (5th Cir.1980)12; see also United States v. Godoy, 821 F.2d 1498, 1504 (11th Cir.1987) ("This rule, however, is not jurisdictional and may be waived by this court in certain exceptional circumstances.... Our decision to consider an argument for the first time [on appeal] is left to our discretion, based on the facts of each case.").

As before, none of the relevant habeas statutes suggests that Congress has eliminated our traditional authority to address waived issues. Consequently, the FPC's second argument fails.

Having demonstrated that a circuit court of appeals has the power — even in the habeas corpus context — to consider sua sponte issues that a petitioner fails to preserve either in the district court or on appeal, we are still left with the issue of whether it was appropriate for this court to have exercised its discretionary authority in this particular case. We are clearly obligated to raise questions concerning our subject-matter jurisdiction sua sponte in all cases. See, e.g., Arthur v. Haley, 248 F.3d 1302, 1303 n. 1 (11th Cir.2001) ("While neither party has raised the issue of whether we have subject matter jurisdiction over this case, we are obliged to address the issue sua sponte."). We may raise nonjurisdictional issues sua sponte in habeas proceedings only if they implicate "important federal interest[s]." Esslinger, 44 F.3d at 1524; see, e.g., Moon v. Head, 285 F.3d 1301, 1315 n. 17 (11th Cir.2002), cert. denied, 537 U.S. 1124, 123 S. Ct. 863, 154 L. Ed. 2d 807 (2003) ("We cannot discern — nor did the district court find — any important federal interest in this case to justify raising the [procedural] bar [issue] sua sponte. We therefore consider the merits."). For this reason, we will generally not consider exhaustion issues sua sponte because this doctrine is only "intended to give the state the initial opportunity to decide alleged violations of federal constitutional rights." Hopkins v. Jarvis, 648 F.2d 981, 984 n. 2 (5th Cir. Unit B June 1981). Determining whether § 2241 or § 2254 was the proper federal statute under which a state prisoner may challenge the calculation of his parole date is a sufficiently important federal issue to consider sua sponte.

A court of appeals may also consider an issue sua sponte when it can fairly be characterized as a "threshold" matter to another question properly before it. In Jorss v. Gomez, the Ninth Circuit stated, "The district court granted a Certificate of Appealability (`COA') as to `whether equitable tolling applied' to Jorss's petition. Because we find that a determination of timeliness under the statute is a necessary predicate to the question of whether equitable tolling should be applied, we hold ... that Jorss's petition was timely filed." 311 F.3d 1189, 1191 (9th Cir.2002) (emphasis added). The court went on to explain, "[A] court must first determine whether a petition was untimely under the statute itself before it considers whether equitable tolling should be applied. As a matter of logic, where a petition is timely filed within the one-year statute of limitations imposed by AEDPA, then equitable tolling need not be applied." Id. at 1192 (internal citation omitted); see also United States Nat'l Bank v. Indep. Ins. Agents of Am., Inc., 508 U.S. 439, 447, 113 S. Ct. 2173, 2178, 124 L. Ed. 2d 402 (1993) ("[A] court may consider an issue `antecedent to ... and ultimately dispositive of' the dispute before it, even an issue the parties fail to identify and brief." (quoting Arcadia v. Ohio Power Co., 498 U.S. 73, 77, 111 S. Ct. 415, 112 L. Ed. 2d 374 (1990))).

Here, the original COA concerned whether the district court correctly applied the procedural bar provisions of 28 U.S.C. § 2254. Given our doubts as to whether Thomas's petition (which had been filed under 28 U.S.C. § 2241) should have been considered under § 2254 at all, it would have been absurd to interpret this statute without first ascertaining whether it was the applicable law. Due to the possibility that § 2254 did not govern Thomas's habeas petition, an opinion reviewing the district court's application of § 2254 without resolving the threshold issue of the statute's applicability would have run the risk of being a constitutionally prohibited advisory opinion of an inapplicable statute. See NLRB v. Gen. Cinema Corp., 526 F.2d 427, 429 (5th Cir.1976) ("Article III of the Constitution prohibits the rendition of advisory opinions.").

Where a circuit panel decides to raise an issue sua sponte, the preferred method of doing so is by requesting supplemental briefing from the parties and permitting oral argument, as was done here. See Adam A. Milani & Michael R. Smith, Playing God: A Critical Look at Sua Sponte Decisions by Appellate Courts, 69 Tenn. L. Rev. 245, 252-53 (2002) ("[W]hen appellate courts identify an issue not raised by the litigants, they should order supplemental briefing as a matter of course, so that the parties have an opportunity to be heard on the issue."); Barry A. Miller, Sua Sponte Appellate Rulings: When Courts Deprive Litigants of an Opportunity to Be Heard, 39 San Diego L. Rev. 1253, 1310 (2002) ("Notice and an opportunity to be heard before deciding a case sua sponte ... makes the unequal application of waiver more fair to the litigants (and increases the accuracy of the decisionmaking process in some cases).").

There is no statutory or doctrinal prohibition against an appellate court issuing a COA sua sponte on issues not specified by a habeas petitioner. This is a well-established practice in the majority of both district and circuit courts to have addressed the issue, including this circuit. Because of the importance of the §§ 2241/2254 issue, and because it would be somewhat silly to debate how § 2254 applies to Thomas without first satisfying ourselves that it is, in fact, the governing statute, we are correct in exercising our discretion to consider this question sua sponte.

This brings us to the actual issue embraced by the second COA — whether the district court erred in assuming that Thomas's petition should be examined under § 2254 rather than § 2241. There is very little precedent other than Medberry v. Crosby, 351 F.3d 1049 (11th Cir.2003), directly on point; even the Supreme Court has cited the statutes together without distinguishing between them. See, e.g., Francis v. Henderson, 425 U.S. 536, 538, 96 S. Ct. 1708, 1710, 48 L. Ed. 2d 149 (1976) ("There can be no question of a federal district court's power to entertain an application for a writ of habeas corpus in a case such as this. 28 U.S.C. §§ 2241, 2254."). Prior to the enactment of AEDPA, several of our own rulings also conflated these provisions, allowing a single habeas petition to be filed under the authority of both. See, e.g. Grace v. Hopper, 566 F.2d 507, 508 (5th Cir.1978) ("This is a habeas corpus case brought pursuant to 28 U.S.C. §§ 2241, 2254...."); Glass v. Heyd, 457 F.2d 562, 563 (5th Cir.1972) (per curiam) ("In this habeas corpus proceeding filed under 28 U.S.C. §§ 2241, 2242 and 2254 petitioner ... challenges two convictions for criminal contempt....").

I begin, as usual, with a plain-meaning analysis of the statutory text of §§ 2241 and 2254 in Subpart A, demonstrating that each section creates a separate route through which a petitioner may seek federal habeas corpus relief. Subpart B discusses how the Supreme Court's ruling in Ex parte Yerger, 75 U.S. (8 Wall.) 85, 19 L. Ed. 332 (1868), requires that we treat these sections as creating independent vehicles for relief. Subpart C offers a structural analysis of federal habeas corpus law, demonstrating that when the various sections are read together, they strongly suggest that relief under § 2241 is distinguishable from relief under § 2254. Subpart D demonstrates how the canon of statutory construction concerning implied repeals leads to the same result. Subpart E contrasts my interpretation of §§ 2241 and 2254 with that of other circuits, showing the flaws in their reasoning. Finally, Subpart F briefly concludes.

To determine the proper statute under which Thomas's claims should have been considered, we begin with the statutory text. 28 U.S.C. §§ 2241(a), (b)(3) states, in relevant part, "Writs of habeas corpus may be granted by ... the district courts ... within their respective jurisdictions.... The writ of habeas corpus shall not extend to a prisoner unless ... [h]e is in custody in violation of the Constitution or laws or treaties of the United States." The language of this section appears to make the writ available to any federal and state prisoners who have federal constitutional claims, whether or not they have been convicted. See Stacey v. Warden, Apalachee Corr. Inst., 854 F.2d 401, 403 n. 1 (11th Cir.1988) ("[Section] 2241 ... applies to persons in custody regardless of whether final judgment has been rendered.").

Section 2254, in contrast, applies only to convicted state prisoners. It provides,

[A] district court shall entertain an application for a writ of habeas corpus in behalf of a person in custody pursuant to the judgment of a State court only on the ground that he is in custody in violation of the Constitution or laws or treaties of the United States.

28 U.S.C. § 2254(a) (emphasis added). This language is essentially identical to that employed in § 2241, except that it is limited to state prisoners. Where the statutory text of two separate sections of the United States Code authorize a convicted state prisoner to bring exactly the same type of claim — that "he is in custody in violation of the Constitution or laws or treaties of the United States" — we are forced to give these sections their plain and natural meaning and conclude that the claim may be filed under either. See Sorenson v. Sec'y of the Treasury, 475 U.S. 851, 860, 106 S. Ct. 1600, 1606, 89 L. Ed. 2d 855 (1986) ("[I]dentical words used in different parts of the same act are intended to have the same meaning.") (internal marks and citations omitted). Thus, while "[p]re-trial habeas petitions ... are properly brought under 28 U.S.C. § 2241," Stacey, 854 F.2d at 403 n. 1, it seems that a convicted state prisoner like Thomas may challenge any aspect of his conviction or sentencing, or the execution of his sentence (including determinations concerning parole) under either § 2241 or § 2254.

Shortly after the Civil War, there was a statutory habeas scheme much like the one that exists today, where there were two statutes through which a federal prisoner could seek relief. The Supreme Court interpreted them as providing separate routes for relief, however, and held that changes to one did not affect the other. Based on this precedent, I believe we are bound to treat §§ 2241 and 2254 as independent as well.

Our nation's first federal habeas statute was enacted in the Judiciary Act of 1789:

[A]ll the before-mentioned courts of the United States, shall have power to issue writs of ... habeas corpus.... And ... either of the justices of the supreme court, as well as judges of the district courts, shall have power to grant writs of habeas corpus for the purpose of an inquiry into the cause of commitment. — Provided, That writs of habeas corpus shall in no case extend to prisoners in gaol, unless where they are in custody, under or by colour of the authority of the United States....

Judiciary Act of Sept. 24, 1789, ch. 20, § 14, 1 Stat. 73, 81-82. The Supreme Court interpreted this Act as granting federal district and circuit courts original jurisdiction, and the Supreme Court appellate jurisdiction, over the habeas petitions of federal prisoners. See Ex parte Yerger, 75 U.S. at 101, 19 L. Ed. at 337 ("The jurisdiction [in habeas matters] thus given in law [the Act of 1789] to the Circuit and District Courts is original; that given by the Constitution and the law [the Act of 1789] to this court is appellate.").

In 1867, Congress enacted another statute, under which the writ of habeas corpus was made available to both federal and state prisoners. The law stated,

[T]he several courts of the United States, and the several justices and judges of such courts, within their respective jurisdictions, in addition to the authority already conferred by law, shall have power to grant writs of habeas corpus in all cases where any person may be restrained of his or her liberty in violation of the constitution, or of any treaty or law of the United States.... From the final decision [of any circuit court in a habeas case,] an appeal may be taken... to the Supreme Court of the United States....

Act of Feb. 5, 1867, ch. 28, § 1, 14 Stat. 385, 385. After the enactment of this statute, there were two different laws with language allowing a federal prisoner to seek a writ of habeas corpus or appeal a lower court's denial of a habeas writ to the Supreme Court — the Act of 1789 and the Act of 1867. As we shall see below, the Supreme Court interpreted these statutes as being two separate and independent vehicles for seeking habeas relief, rather than holding that one narrowed or eliminated the other.

In 1868, Congress repealed part of the Act of 1867.13 See Act of Mar. 27, 1868, ch. 34, § 2, 15 Stat. 44, 44 ("And be it further enacted, [t]hat so much of the act approved February five, eighteen hundred and sixty-seven... as authorizes an appeal from the judgment of the circuit court to the Supreme Court of the United States ... is[] hereby repealed."). If the Act of 1867 had limited or eliminated the Act of 1789 as a means through which federal prisoners could appeal to the Supreme Court a circuit court's refusal to issue the writ, then the repeal of the Act of 1867 would have left federal prisoners without a way of obtaining Supreme Court review of adverse federal habeas rulings.14

In Yerger, the Supreme Court concluded that the Acts of 1789 and 1867 were entirely separate and independent from each other, and that federal prisoners could seek federal habeas relief or appeal adverse rulings under either. The main issue in Yerger arose because of Congress's 1868 repeal of the 1867 Act. The Court had to determine whether this repeal eliminated all of its appellate jurisdiction over federal habeas cases or whether the appellate jurisdiction conferred by the Act of 1789 remained. See Ex parte Yerger, 75 U.S. at 103 (addressing "whether the 2d section of the act of March 27th, 1868, takes away or affects the appellate jurisdiction of this court under the Constitution and the acts of Congress prior to 1867."). The Supreme Court held that the 1868 repeal "affect[ed] only appeals and appellate jurisdiction authorized by that [1867] act. They do not purport to touch the appellate jurisdiction conferred by the Constitution, or to except from it any cases not excepted by the act of 1789. They reach no act except the act of 1867." Id. at 105. The Supreme Court further stressed, "It is true that the exercise of appellate jurisdiction, under the act of 1789, was less convenient than under the act of 1867, but the provision of a new and more convenient mode of its exercise does not necessarily take away the old." Id. Consequently, the Act of 1867 did not prevent federal prisoners from seeking a writ of habeas corpus or appealing a denial of habeas corpus to the Supreme Court under the Act of 1789; the Act of 1867 merely provided a separate, alternate route through which they could seek relief. See id. at 105 (denying that "the act of 1789, so far as it provided for the issuing of writs of habeas corpus by this court, was already repealed by the act of 1867"). Neither the enactment nor repeal of the 1867 statute affected interpretation of the Act of 1789, or the rights it gave federal prisoners.

Thus, the Yerger opinion is crucial to this case in two respects. First, when two habeas statutes each extended the writ of habeas corpus to federal prisoners, the Supreme Court treated each as an independent vehicle through which relief could be sought. The Act of 1789 made the writ of habeas corpus available to federal prisoners. The Act of 1867 made the writ of habeas corpus available to both federal and state prisoners. Confronted with these two Acts, the Supreme Court declared,

The appeal given by the act of 1867 extended, indeed, to cases, within the [Act of 1789];.... But this effect does not take from the act [of 1867] its character of an additional grant of jurisdiction, and make it operate as a repeal of jurisdiction theretofore allowed [under the Act of 1789].

Id. at 106 (emphasis added). Applying Yerger to the instant case (in which two separate statutes address federal habeas relief for state prisoners), we should treat §§ 2241 and 2254 as separate and independent vehicles for obtaining the writ, and not interpret § 2241 so as to "make it operate as a repeal of jurisdiction theretofore allowed [under § 2254]." Id.

Second, Yerger shows us that Congress has previously enacted redundant habeas statutes for certain classes of prisoners. It is indubitable that, between the enactment of the Act of 1867 and its 1868 repeal, a federal prisoner could appeal a denial of habeas relief to the Supreme Court under either the Act of 1789 or the Act of 1867. Thus, we should not be overly concerned about concluding that a state prisoner may seek relief under either §§ 2241 or 2254; this would not be the first time Congress made habeas relief available to people under two separate statutes.

In Section A, I considered the text of §§ 2241 and 2254 as discrete sections. Even if we adopt a structuralist view, however, and consider them in the context of the entire statutory habeas scheme, it becomes clear that each is actually a substantive, independent route through which a state prisoner may seek habeas relief. Subsection 1 arrives at this result by contrasting these sections with the language of § 2255, which establishes the "motion to vacate" — the equivalent to habeas corpus relief for federal prisoners. Subsection 2 looks to variations in the language of § 2244, which establishes conditions on the availability of habeas relief, noting that while some of them apply to habeas petitions under either §§ 2241 or 2254, others are tailored exclusively to § 2254, thereby suggesting that petitions under §§ 2241 and 2254 are not interchangeble or equivalent.

28 U.S.C. § 2255 allows a convicted federal prisoner to move the district court to vacate his sentence. This statute states, "An application for a writ of habeas corpus [under §§ 2241] in behalf of a prisoner who is authorized to apply for relief pursuant to this section [§ 2255] shall not be entertained if it appears that the applicant has failed to apply for relief, by motion [under § 2255]." I shall refer to this passage as § 2255's "exclusivity provision." Due to this exclusivity provision, a § 2255 motion to vacate is an alternative to § 2241 habeas corpus relief.

The existence of § 2255's exclusivity provision demonstrates that Congress did not believe that the mere creation of a motion to vacate for convicted federal prisoners would intrinsically preclude such prisoners from seeking habeas relief under § 2241. Congress felt the need to enact the above-quoted language to preclude federal prisoners from filing habeas petitions under § 2241. Since § 2255 needs an exclusivity provision to prevent convicted federal prisoners from seeking relief under § 2241, it stands to reason that § 2254 would need an exclusivity provision to prevent convicted state prisoners from seeking relief under § 2241. "Where Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion." United States v. Granderson, 511 U.S. 39, 63, 114 S. Ct. 1259, 1272, 127 L. Ed. 2d 611 (1994) (Kennedy, J., concurring) (quoting Gozlon-Peretz v. United States, 498 U.S. 395, 404, 111 S. Ct. 840, 846-47, 112 L. Ed. 2d 919 (1991)); see also Chan v. Korean Air Lines, Ltd., 490 U.S. 122, 132-33, 109 S. Ct. 1676, 1683, 104 L. Ed. 2d 113 (1989) ("[G]iven the parallel structures of these provisions [of the Warsaw Convention] it would be a flouting of the text to imply in [one provision] a sanction not only withheld there but explicitly granted elsewhere."). Consequently, the absence of such an exclusivity provision in § 2254 indicates that § 2241 relief is available to prisoners who also qualify for § 2254 relief.

Indeed, were we to interpret the mere enactment of § 2254 as precluding state prisoners from seeking § 2241 relief, this same interpretation would necessarily apply to § 2255, thereby rendering § 2255's exclusivity provision superfluous and meaningless. However, "[a]n interpretation of statutory language that causes other language within the statute to be meaningless contravenes the `elementary canon of [statutory] construction that a statute should be interpreted so as not to render one part inoperative.'" In re City of Mobile, 75 F.3d 605, 611 (11th Cir.1996).

Section 2254 was enacted at the same time and in the same law as § 2255. It authorizes habeas relief for all state prisoners. Nevertheless, unlike its sister provision, § 2254 does not contain an express provision preventing state prisoners from seeking relief under § 2241. We must assume that this variation in language was deliberate and purposeful. See Johnson v. United States, 225 U.S. 405, 415, 32 S. Ct. 748, 751, 56 L. Ed. 1142 (1912) ("A change of [statutory] language is some evidence of a change of purpose...."). Congress knew how to restrict access to § 2241 when it wanted to, and it chose not to do so for state prisoners. We must respect this choice.

Section 2255 is not the only extrinsic provision in federal habeas corpus law that suggests that §§ 2241 and 2254 are independent. Section 2244 contains various restrictions on the availability of federal habeas corpus relief. Some of these provisions are made applicable to "habeas corpus application[s] under section 2254." See, e.g., 28 U.S.C. § 2244(b)(1) (setting forth a general rule to govern "second and successive" habeas petitions under § 2254); id. § 2244(b)(2)(B) (addressing the proper way to deal with claims based on newly discovered evidence of innocence). Other provisions within this section, however, use broader language; they are applicable in any "habeas corpus proceeding brought in behalf of a person in custody pursuant to the judgment of a State court." See, e.g., id. § 2244(c) (setting forth a "law of the case" doctrine for habeas cases involving state prisoners); id. § 2244(d) (establishing a one-year statute of limitations for federal habeas claims by state prisoners).

Section 2254 makes habeas relief available to any "person in custody pursuant to the judgment of a State court." Id. § 2254(a). If the majority is correct, then all convicted state prisoners must bring habeas petitions under § 2254. Consequently, there would have been no reason for Congress to use different language in different provisions of § 2244. That is, there would be no way to explain why some provisions within § 2244 refer to "habeas corpus application[s] under section 2254" and others refer to "habeas corpus proceeding[s] brought in behalf of a person in custody pursuant to the judgment of a State court."

My theory, however, explains why Congress varied this language. Some of the restrictions within § 2244 were intended to apply to all habeas petitions brought by state prisoners, regardless of the statutory provision (§ 2241 or § 2254) under which they were brought. Other restrictions, however, were expressly intended to apply only to petitions brought under § 2254. The majority's interpretation would nullify this deliberate and purposeful variation in language. As discussed above, we are bound to give effect to this variation, and assume that the two phrases — "habeas applications brought under § 2254" and "habeas proceedings brought by state prisoners" — are not interchangeable. See Smith v. United States, 508 U.S. 223, 235, 113 S. Ct. 2050, 2057, 124 L. Ed. 2d 138 (1993) (holding that related statutory provisions should be given different meanings where Congress "carefully varied the statutory language" used in each); American Nat'l Red Cross v. S.G., 505 U.S. 247, 263, 112 S. Ct. 2465, 2475, 120 L. Ed. 2d 201 (1992) (applying "the canon of statutory construction requiring a change in language to be read, if possible, to have some effect"). Put another way, this variation in terminology strongly suggests that it is possible for a state prisoner to bring a habeas petition under a statute other than § 2254; it suggests state prisoners may seek relief under § 2241, as well.

Although both plain text and structural analyses of §§ 2241 and 2254 demonstrates that each section establishes a separate route through which convicted state prisoners may seek federal habeas relief, we can reach this same conclusion by applying the canons of statutory construction. According to the doctrine against implied repeals, when interpreting two statutes, we must endeavor to apply both fully and prevent one from limiting the legal effect of the other. See Rodriguez v. United States, 480 U.S. 522, 524, 107 S. Ct. 1391, 1392, 94 L. Ed. 2d 533 (1987) ("It is well settled, however, that repeals by implication are not favored, and will not be found unless an intent to repeal is clear and manifest." (internal marks and citations omitted)); Knight v. Georgia, 992 F.2d 1541, 1546 (11th Cir.1993) ("The canon of statutory construction strongly disfavors findings of implied repeal, and we decline to draw such an inference in this case." (internal citations omitted)).

The language of both §§ 2241 and 2254 is broad enough to allow state prisoners to seek writs of habeas corpus. Section 2254, however, does not explicitly purport to amend, repeal, limit, or revise § 2241. Consequently, we should not interpret the elaborate restrictions Congress established for § 2254 — particularly those set forth in the 1996 AEDPA amendments — as curtailing or eliminating a convicted state prisoner's right to seek relief under § 2241. If Congress wishes to change the broad language of § 2241, it is free to do so, but neither the original enactment of, nor subsequent amendments to, § 2254 is enough to accomplish that task.

Having discussed my own approach to the §§ 2241/2254 issue, I believe it would be helpful to contrast it with interpretations offered by other circuits. Section 1 discusses the circuits that, like the majority, conclude that convicted state prisoners may never bring habeas petitions under § 2241. Section 2 discusses the Tenth Circuit's ruling that state prisoners may bring certain kinds of challenges under § 2241, and must bring other challenges under § 2254.

Many circuits have held that convicted state prisoners must bring all habeas claims under § 2254. For example, the Seventh Circuit ruled that, with regard to convicted state prisoners, § 2254 "in effect implements the general grant of habeas corpus authority found in § 2241." Walker v. O'Brien, 216 F.3d 626, 633 (7th Cir.2000). It went on to declare without explanation that "[e]ven though § 2254 does not contain exclusivity language ... as a practical matter the requirements of § 2254 must be met by all state prisoners filing petitions for writs of habeas corpus after conviction." Id. Consequently, "§ 2254 [is] the exclusive vehicle for prisoners in custody pursuant to a state court judgment who wish to challenge anything affecting that custody...." Id. The Seventh Circuit "permits resort to Section 2241 only when Section 2254 is not available." McLean v. Smith, 193 F. Supp. 2d 867, 871 (M.D.N.C.2002).

As discussed earlier, see supra Section III.C.1, I believe the presence of exclusivity language in § 2255, but not § 2254, is significant, and disagree with the Seventh Circuit's decision to overlook this crucial difference. Moreover, the Seventh Circuit's opinion begins by assuming that § 2254 simply "implements" § 2241; this essentially begs the question of whether they exist independently of each other. Indeed, this assumption is undermined by Congress's decision to make certain restrictions applicable to all habeas petitions filed by convicted state prisoners and others applicable only to habeas petitions brought under § 2254. See supra Section III.C.2.

Other circuits, turning the Seventh Circuit's reasoning on its head, have nevertheless reached the same conclusion. In James v. Walsh, the Second Circuit held that § 2241 was unavailable to convicted state prisoners because

[s]ection 2254(b)(1) requires state prisoners to exhaust all available state court remedies before filing a Section 2254 petition, whereas Section 2241 contains no such exhaustion requirement. Had Congress intended to make Section 2241 available to state prisoners, it would likely have required, in the interests of comity, that state prisoners challenging the execution of their state-imposed sentences first exhaust their remedies in the state courts.

308 F.3d 162, 167 (2d Cir.2002). The Third Circuit echoed the Second Circuit's view, holding:

Allowing [a state prisoner] to file the instant petition in federal court pursuant to Section 2241 without reliance on Section 2254 would circumvent [§ 2254's] restriction[s]... and would thereby thwart Congressional intent. Thus, applying the "specific governs the general" canon of statutory construction to this action, we hold that [state prisoners] must rely on Section 2254 in challenging the execution of [their] sentence[s].

Coady v. Vaughn, 251 F.3d 480, 484-85 (3d Cir.2001). Like the Second Circuit, the Third Circuit construes Congress's erection of additional hurdles for habeas writs under § 2254 as an implicit bar to applications under § 2241.

These courts point to the differences between § 2241 and § 2254 to conclude that Congress could not possibly have intended to allow convicted state prisoners to take advantage of § 2241. I believe the fact that different restrictions apply to each statute suggests that the statutes create different routes through which habeas relief may be obtained. This is hardly unusual; there are many occasions in federal law where Congress provides multiple ways of doing the same thing with differing levels of procedural or substantive difficulty. I am simply unwilling to make the tremendous leap these courts made. Had Congress intended § 2254's restrictions to apply to all habeas applications by convicted state prisoners, it would not have limited the scope of those restrictions to only applications filed under § 2254.

The Eighth Circuit also believes that convicted state prisoners are limited to filing petitions under § 2254. See Singleton v. Norris, 319 F.3d 1018, 1023 (8th Cir.2003) (en banc) ("§ 2254 is the only means by which `a person in custody pursuant to the judgment of a State court' may raise challenges to the validity of his conviction or sentence or to the execution of his sentence."), cert. denied, 540 U.S. 832, 124 S. Ct. 74, 157 L. Ed. 2d 59 (2003); Crouch v. Norris, 251 F.3d 720, 723 (8th Cir.2001) ("Not only is § 2254 an appropriate vehicle for Crouch's proposed claims, it is, as a practical matter, the only vehicle."). To support its conclusion, the Eighth Circuit simply cited the Supreme Court's ruling in Felker v. Turpin that "[a]uthority to grant habeas relief to state prisoners is limited by § 2254, which specifies the conditions under which such relief may be granted to `a person in custody pursuant to the judgment of a State court.'" Crouch, 251 F.3d at 723, quoting 518 U.S. 651, 662, 116 S. Ct. 2333, 2339, 135 L. Ed. 2d 827 (1996).

The Eighth Circuit took this quote out of context. Felker involved a habeas petition to the Supreme Court under § 2254; the Court was not speaking to whether habeas relief was available through other avenues such as § 2241. Instead, it was merely pointing out the indisputable fact that § 2254 has many restrictions by which courts considering § 2254 habeas petitions must abide.

Unlike the courts discussed above, the Tenth Circuit has permitted convicted state prisoners to bring at least some types of collateral attacks under § 2241 rather than § 2254. The Tenth Circuit has held, "Petitions under § 2241 are used to attack the execution of a sentence, in contrast to § 2254 habeas and § 2255 proceedings, which are used to collaterally attack the validity of a conviction and sentence." McIntosh v. United States Parole Comm'n, 115 F.3d 809, 811 (10th Cir.1997) (internal citations omitted). Under this approach, claims that a court acted unconstitutionally during a prisoner's trial or sentencing hearing must be brought under § 2254, while most other claims (including claims concerning parole board actions) must be brought under § 2241. A district court within our jurisdiction, without explanation, came to the same conclusion. See Vargas v. Sikes, No. 1:98-CV-0651-TWT, 1998 U.S. Dist. LEXIS 22232, at *9 (N.D.Ga. Feb. 9, 1998) ("Ostensibly an action under 28 U.S.C. § 2254, [petitioner's habeas suit] was an action under 28 U.S.C. § 2241 because it challenged the execution of the petitioner's sentence and not the conviction or the imposition of the sentence.").

The Rules Governing Section 2254 Cases in the United States District Courts lend support to the Tenth Circuit's view that § 2254 is limited to attacks on the judgments or sentences of courts. For instance, Rule 2(d) states, "A petition shall be limited to the assertion of a claim for relief against the judgment or judgments of a single state court (sitting in a county or other appropriate political subdivision)." This provision does not seem to contemplate challenges to adverse parole determinations. Rule 5 provides that the state's answer in a § 2254 proceeding "shall indicate what transcripts(of pretrial, trial, sentencing, and post-conviction proceedings) are available...." While the term "post-conviction proceedings" is arguably broad enough to embrace parole hearings, when read in context this provision appears to suggest that the "main thrust" of § 2254 is to review what happened in court.

The Form included in the Rules' appendix also clearly contemplates challenges to trials or sentences, and not to administrative proceedings such as parole hearings. For example, the Form asks whether the petitioner has appealed his conviction, and not whether he has appealed any adverse administrative actions. The Form specifies the ten "most frequently raised grounds for relief in habeas corpus proceedings" — all concerning trial-related rights. Moreover, other than "[d]enial of effective assistance of counsel" and "[d]enial of right of appeal," all the grounds specified in the Form expressly attack the underlying conviction (e.g., "Conviction obtained by use of coerced confession.").

While intriguing, these facts are less than compelling. Moreover, the structure of § 2255 strongly suggests that neither § 2241 nor § 2254 are limited in the types of constitutional claims that may be brought under either. Section 2255 authorizes a federal prisoner to file a motion to vacate his sentence (the equivalent of a § 2254 habeas petition) only if "the sentence was imposed in violation of the Constitution or laws of the United States, or that the court was without jurisdiction to impose such sentence, or that the sentence was in excess of the maximum authorized by law, or is otherwise subject to collateral attack ...." (emphasis added). If § 2241 were limited to attacks on the execution of a sentence, or § 2254 were limited to attacks on a trial or sentencing proceeding, we would have expected those sections to contain similar limiting language. "Where Congress includes particular language in one section of a statute but omits it in another... it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion." Keene Corp. v. United States, 508 U.S. 200, 208, 113 S. Ct. 2035, 2040, 124 L. Ed. 2d 118 (1993). The Tenth Circuit's conclusion also seems inconsistent with Supreme Court cases in which the Court entertained habeas petitions under § 2254 that involved constitutional challenges to adverse parole decisions, an area that falls outside the Tenth Circuit's conception of the scope of this statute. See, e.g., Spencer v. Kemna, 523 U.S. 1, 118 S. Ct. 978, 140 L. Ed. 2d 43 (1998) (rejecting on mootness grounds prisoner's challenge to the revocation of his parole, brought under § 2254); Cal. Dep't of Corr. v. Morales, 514 U.S. 499, 504, 115 S. Ct. 1597, 1600-01, 131 L. Ed. 2d 588 (1995) (rejecting, on the merits, a petition for a writ of habeas corpus under § 2254 alleging that a change in state parole laws violated the Ex Post Facto Clause).

Finally, I am unwilling to pile inference on top of inference and hold that, because Congress must have intended to limit § 2254 to certain types of habeas petitions, § 2241 is necessarily limited to petitions falling outside of those categories. Courts undoubtedly have an obligation to harmonize the various provisions within a statutory scheme. See United Sav. Ass'n v. Timbers of Inwood Forest Assoc., Ltd., 484 U.S. 365, 371, 108 S. Ct. 626, 630, 98 L. Ed. 2d 740 (1988) ("Statutory construction, however, is a holistic endeavor. A provision that may seem ambiguous in isolation is often clarified by the remainder of the statutory scheme...."); Mountain States Tel. & Tel. Co. v. Pueblo of Santa Ana, 472 U.S. 237, 252, 105 S. Ct. 2587, 2596, 86 L. Ed. 2d 168 (1985) (adopting the "reading of the statute that better harmonizes the two clauses [at issue] with the structure of the entire Act [at issue]"). Nevertheless, the leaps and bounds required to arrive at the Tenth Circuit's admittedly elegant and tempting conclusion are simply too great.

Thus, in light of the expansive language of § 2241, I reject the proposition that § 2254 is the exclusive route through which convicted state prisoners may seek federal habeas relief. My conclusion is not entirely without precedential support. While this circuit has never directly addressed the issue before Medberry, we have previously ruled on the merits of at least one challenge to FPC rulings brought under § 2241, without suggesting that it should have been brought under § 2254. See Van Zant v. Fla. Parole Comm'n, 104 F.3d 325 (11th Cir.1997) (per curiam).

Needless to say, this interpretation is not without serious problems. Most notably, it forces us to reject the otherwise sensible suggestion that there is only one habeas statute under which a prisoner may bring a particular claim. Moreover, it allows state prisoners to evade at least some of the statutory requirements for obtaining habeas petitions under § 2254 by simply filing under § 2241. The Medberry panel correctly noted,

If § 2254 were not a restriction on § 2241's authority to grant the writ of habeas corpus, and were instead a freestanding, alternative post-conviction remedy, then.... Section 2254 would be a great irrelevancy because a state prisoner could simply opt out of its operation by choosing a different label for his petition.

351 F.3d at 1060-61.

This necessarily happens, however, whenever Congress creates two routes to achieving the same goal and one is easier or otherwise more attractive than the other. While this conclusion may seem undesirable, the fault lies with Congress, who decided to enact two virtually identical statutory provisions (at least with regard to convicted state prisoners), yet made certain requirements applicable only to one of them. It is not up to this court to rewrite federal habeas statutes. Moreover, as discussed in the next Part, many of the statutory restrictions that apply to § 2254 are based on common law constraints on habeas corpus that apply to petitions under § 2241.

Section 2254 contains a wide range of requirements and restrictions that do not directly apply to petitions brought under § 2241. See, e.g., 28 U.S.C. §§ 2254(b)(1) (requiring exhaustion of state remedies); 2254(d)(1) (limiting reversals to cases involving "unreasonable application[s] of, clearly established Federal law"); 2254(e)(1) (establishing presumption of correctness for factual determinations by state courts); 2254(e)(2) (establishing stringent requirements for evidentiary hearings); 2254(i) (barring claims for ineffective assistance of counsel the petitioner received in state or federal post-conviction proceedings). Section 2241 proceedings, in contrast, are governed by the common law requirements for habeas petitions. Some of § 2254's restrictions on habeas relief, however, are simply codified versions of these common law principles. See Williams v. Taylor, 529 U.S. 362, 380 n. 11, 120 S. Ct. 1495, 1506, 146 L. Ed. 2d 389 (2000) ("It is not unusual for Congress to codify earlier precedent in the habeas context.").

Among the most fundamental common law requirements of § 2241 is that petitioners must first exhaust their state court remedies. "The exhaustion doctrine is a judicially crafted instrument which reflects a careful balance between important interests of federalism and the need to preserve the writ of habeas corpus as a swift and imperative remedy in all cases of illegal restraint or confinement." Braden v. 30th Judicial Circuit Court, 410 U.S. 484, 490, 93 S. Ct. 1123, 1127, 35 L. Ed. 2d 443 (1973) (internal marks and citation omitted). The Third Circuit, interpreting Braden, held, "[A]n exhaustion requirement has developed through decisional law, applying principles of federalism.... [A]lthough there is a distinction in the statutory language of §§ 2254 and 2241, there is no distinction insofar as the exhaustion requirement is concerned." Moore v. DeYoung, 515 F.2d 437, 442 (3d Cir.1975); see also Schandelmeier v. Cunningham, 819 F.2d 52, 53 (3d Cir.1986) ("The state court exhaustion requirement is mandated by statute under 28 U.S.C. § 2254(b) and has developed through decisional law in applying principles of comity and federalism as to claims brought under 28 U.S.C. § 2241.").

I agree with the Third Circuit. In Ex parte Hawk, the Supreme Court held,

Ordinarily an application for habeas corpus by one detained under a state court judgment of conviction for crime will be entertained by a federal court only after all state remedies available, including all appellate remedies in the state courts and in this Court by appeal or writ of certiorari, have been exhausted.

321 U.S. 114, 116-17, 64 S. Ct. 448, 450, 88 L. Ed. 572 (1944).15 The Court later explained that "[t]his exhaustion requirement is also grounded in principles of comity; in a federal system, the States should have the first opportunity to address and correct alleged violations of state prisoner's federal rights." Coleman v. Thompson, 501 U.S. 722, 731, 111 S. Ct. 2546, 2555, 115 L. Ed. 2d 640 (1991). Thus, convicted state prisoners bringing suit under § 2241 must exhaust their state court remedies not because of § 2254(b)(1)(A),16 but because this has been long recognized as an integral part of § 2241 proceedings. See Haig v. Weber County, 468 F. Supp. 887, 889 (D.Ut.1978) ("When Congress revised the habeas statute in 1948, its intent was to embody in the statute the judicially developed principle of exhaustion as expressed in Hawk.").

The four issues Thomas raised in his federal habeas petition had been previously raised in both his state mandamus and state habeas petitions. Because Thomas failed to appeal the denial of his state mandamus petition, the state habeas court held that Thomas had not exhausted his state level remedies, and so was procedurally barred from reasserting many of his claims. Hawk emphasizes that a prisoner must pursue all "available" state level remedies, which include appellate review. Of course, Thomas is, by now, undoubtedly time-barred under Florida law from seeking appellate review of the mandamus ruling, and has already been procedurally barred from pursuing habeas relief in state court, so in one sense he has exhausted all of his "available" state remedies.

The Hawk rule, however, cannot be satisfied through such technicalities; the common law exhaustion requirement clearly contemplates active pursuit of constitutional claims on the state level before resort to federal habeas proceedings. While a prisoner is not obligated to seek every conceivable extraordinary writ available in state court, he must at the very least afford the state "a full and fair opportunity to resolve federal constitutional claims before those claims are presented to the federal courts ... [by] invoking one complete round of the State's established appellate review process." O'Sullivan v. Boerckel, 526 U.S. 838, 845, 119 S. Ct. 1728, 1732, 144 L. Ed. 2d 1 (1999). Consequently, Thomas is procedurally barred under § 2241' s common law exhaustion requirement from bringing his federal habeas petition.17

It should be noted that Boerckel and Coleman are not directly controlling on these issues because they interpret the statutory exhaustion requirement for habeas petitions under § 2254, which do not apply to § 2241 petitions. Nevertheless, their reasoning and conclusions regarding exhaustion are so persuasive and consistent with the contours of the common law's exhaustion requirement that I would incorporate their doctrines into the evolving common law governing § 2241 habeas petitions.

Thus, while Thomas had the right to have his petition considered under the standards of § 2241 rather than § 2254, the district court still acted properly in rejecting it.