Clinton Eldridge v. Catricia Howard

FOR PUBLICATION

UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT

CLINTON T. ELDRIDGE, No. 21-15616
Petitioner-Appellant,
D.C. No.
v. 4:21-cv-00081-
RCC-LAB
CATRICIA HOWARD; PATRICIA
K. CUSHWA, Acting Chairman of the
U.S. Parole Commission, OPINION
Respondents-Appellees.

Appeal from the United States District Court
for the District of Arizona
Raner C. Collins, District Judge, Presiding

Argued and Submitted December 5, 2022
Phoenix, Arizona

Filed June 2, 2023

Before: Kim McLane Wardlaw and Patrick J. Bumatay,
Circuit Judges, and Karen E. Schreier,* District Judge.

Opinion by Judge Schreier;
Dissent by Judge Bumatay

*
The Honorable Karen E. Schreier, United States District Judge for the
District of South Dakota, sitting by designation.
2 ELDRIDGE V. HOWARD

SUMMARY**

Habeas Corpus

The panel reversed the district court’s judgment
dismissing Clinton Eldridge’s amended habeas corpus
petition, which the district court construed as brought under
28 U.S.C. § 2241, and remanded to the district court to
decide the petition on the merits.
Eldridge filed the instant habeas petition in the District
Court for the District of Columbia in February 2020. In the
amended petition, Eldridge challenges, among other things,
the United States Parole Commission’s 2019 decision to
issue a three-year “set-off,” the time he must wait until his
next parole hearing. The District Court for the District of
Columbia transferred the case to the District of Arizona,
where Eldridge was incarcerated. That court dismissed
Eldridge’s petition as an impermissible second or successive
petition under the Antiterrorism and Effective Death Penalty
Act, and denied Eldridge’s motion to reconsider. Relying on
the abuse of the writ doctrine, the district court concluded
that Eldridge’s claims were substantially similar to the
claims he raised in at least two other § 2241 petitions.
The panel held that Eldridge need not obtain a certificate
of appealability (COA) to appeal the denial of the instant
petition because Congress did not define or include the
District of Columbia Superior Court as a “State court” in 28
U.S.C. § 2253(c), where it had expressly done so in that and
other statutes. The panel held that § 2253(c)(1)(A)’s

**
This summary constitutes no part of the opinion of the court. It has
been prepared by court staff for the convenience of the reader.
 ELDRIDGE V. HOWARD 3

language, “in which the detention complained of arises out
of process issued by a State court,” does not include the
District of Columbia Superior Court. Prisoners whose
detention arises out of process issued by a District of
Columbia court are not required to obtain a COA to appeal
the denial of habeas relief; thus, the COA jurisdictional
requirement does not pose a barrier to Eldridge’s appeal.
The panel further held that the district court erred in
dismissing the petition as an abuse of the writ when Eldridge
could not have possibly raised the same claims in prior
petitions. Looking to the substance of Eldridge’s claim—
that the Parole Commission acted arbitrarily and
capriciously in 2019 when it issued a three-year set-off—the
panel concluded that Eldridge did not have a fair opportunity
to raise this claim in 2016 or 2018 because the alleged
violation occurred only after the denial of his 2016 and 2018
habeas petitions. Additionally, the district court did not
address the merits of the set-off issues in its decision denying
Eldridge’s habeas petition in 2016, even though Eldridge
raised the issue. Thus, because no court has addressed
Eldridge’s three-year set-off claims regarding his 2016
parole denial, he did not abuse the writ by raising the 2019
denial issue in his instant habeas petition.
Judge Bumatay dissented. He wrote that this court
should have stuck with the consensus, embraced by the five
circuit courts to consider the question, that the D.C. Superior
Court is a “State court” under habeas law and prisoners
challenging detention arising from a D.C. Superior Court
conviction must obtain a COA before appealing. He wrote
that, all told, textual and contextual evidence supports that
overwhelming consensus. He wrote that the panel should
therefore have required Eldridge to obtain a COA before
exercising jurisdiction over this appeal, and that he would
4 ELDRIDGE V. HOWARD

conclude that Eldridge does not deserve one. He wrote that
in scheduling Eldridge’s 2010 and 2013 parole rehearings,
the Parole Commission’s erroneous use of the longer (three-
year) 2000 guidelines set-off period, rather than the shorter
(one-year) 1972 guidelines set-off period, did not result in
any increase in Eldridge’s incarceration, and that his ex post
facto constitutional challenge to those denials therefore
fails. He wrote that Eldridge’s claims related to his 2016 and
2019 rehearings, for which the Commission plainly used the
appropriate 1972 guidelines, fare no better. He concluded
that, even if the Commission had applied an ex post facto
law, which it did not, Eldridge still can’t succeed because
there was no basis to conclude that the longer set-offs under
the new law would extend his actual period of confinement.

COUNSEL

Keith J. Hilzendeger (argued), Assistant Federal Public
Defender; Jon M. Sands, Federal Public Defender, District
of Arizona; Federal Public Defender’s Office; Phoenix,
Arizona; for Petitioner-Appellant.
Timothy R. Cahill (argued) and Suzanne Grealy Curt,
Assistant United States Attorneys; Office of the United
States Attorney; Washington, D.C.; Melissa Marcus
Kroeger, Assistant United States Attorney; Christina M.
Cabanillas, Deputy Appellate Chief; Gary M. Restaino,
United States Attorney, District of Arizona; Office of the
United States Attorney; Tucson, Arizona; for Respondents-
Appellees.
 ELDRIDGE V. HOWARD 5

OPINION

SCHREIER, District Judge:

The issue here is whether the District of Columbia
Superior Court counts as a “State court,” for purposes of 28
U.S.C. § 2253(c)(1), when the statute is silent on the matter.
We conclude, for purposes of 28 U.S.C. § 2253(c)(1), that a
District of Columbia Superior Court is not a “State court,”
and thus petitioner Clinton Eldridge need not obtain a
certificate of appealability (COA) to appeal the district
court’s denial of his most recent habeas petition. We further
hold that because the district court erred in dismissing
Eldridge’s petition, the case is reversed and remanded to the
district court to decide Eldridge’s petition on its merits.
I.
Eldridge pleaded guilty in 1984 to nine counts, ranging
from burglary to rape. Eldridge v. United States, 618 A.2d
690, 694 (D.C. 1992). The District of Columbia Superior
Court sentenced him to prison for 40 to 120 years. Id.1
Eldridge first became eligible for parole in 2010. The United
States Parole Commission (Commission) denied Eldridge
parole in 2010, 2013, 2016, 2018, and 2019.2

1
The District of Columbia Court of Appeals vacated one of his
convictions, but the Superior Court imposed the same aggregate sentence
after remand. See Eldridge v. Davis, No. 10-1440, 2010 WL 3394708,
at *1 (D.D.C. Aug. 25, 2010).
2
The District of Columbia Board of Parole made parole decisions until
1997, when Congress “abolished the D.C. Board of Parole and directed
the U.S. Parole Commission to conduct parole hearings for D.C. Code
offenders.” See Daniel v. Fulwood et al., 766 F.3d 57, 59 (D.C. Cir.
2014).
6 ELDRIDGE V. HOWARD

In March 2016, Eldridge filed a habeas petition under 28
U.S.C. § 2241 challenging his 2010, 2013, and 2016 parole
denials. He alleged, among other things, that the
Commission improperly used guidelines it had issued in
2000, which were not in place at the time he committed his
offense, when deciding whether to release him and how
much time he must wait until his next parole hearing (also
known as a “set-off” date). The United States District Court
for the District of Colorado denied relief but did not
explicitly address Eldridge’s set-off arguments. See
Eldridge v. Oliver, 16-00690, 2017 WL 2812824, at *1–9
(D. Colo. June 29, 2017). The Tenth Circuit denied
Eldridge’s request for a COA and dismissed his appeal. See
Eldridge v. Oliver, 710 F. App’x 348, 349 (10th Cir. 2018).
In April 2018, Eldridge filed another 28 U.S.C. § 2241
petition challenging the denial of parole in 2010 and 2013.
He also alleged that he was improperly denied access to a
sex offender treatment program. Further, he argued that in
2016 the Commission improperly issued a three-year set-off
(the presumptive time under the Commission’s 2000
guidelines), rather than a one-year set-off (the presumptive
time under the District of Columbia Parole Board’s 1972
guidelines, which were in place at the time of his offense). 3
The District of Colorado ultimately dismissed Eldridge’s
2018 habeas petition as “malicious” and “repetitive” without
deciding the merits of his claims. See Eldridge v. U.S.
Parole Comm’n, 18-00797, 2018 WL 10426189, at *1–2 (D.
Colo. May 8, 2018). Once again, the Tenth Circuit denied
Eldridge’s request for a COA and dismissed his appeal. See

3
Eldridge’s April 2018 petition could not challenge his 2018 parole
denial because the Commission did not deny his parole request until
November 2018.
 ELDRIDGE V. HOWARD 7

Eldridge v. U.S. Parole Comm’n, 737 F. App’x 901, 902
(10th Cir. 2018).
In February 2020, Eldridge filed the instant habeas
petition in the District Court for the District of Columbia.
After a district court judge ordered him to do so, Eldridge
filed an amended petition. Although he styled his petition as
a writ of habeas corpus under D.C. Code § 16-1091(b), the
district court construed his petition under 28 U.S.C. § 2241.
See Eldridge v. Blanckensee, 20-1009, 2021 WL 325959, at
*3 (D.D.C. Feb. 1, 2021). In his amended petition, Eldridge
challenges, among other things, the Commission’s 2019
decision to issue a three-year set-off. In support, he points
to the Commission’s 2018 decision issuing a one-year set-
off. He correctly notes that the Commission justified both
its 2018 and 2019 set-off decisions for similar reasons: that
he is an untreated sex offender and needs to participate in a
sex offender treatment program.
The District Court for the District of Columbia
transferred the case to the District of Arizona, the district in
which Eldridge was incarcerated, and flagged that Eldridge’s
claims appeared “substantially similar” to claims he had
made in his 2016 and 2018 petitions. Id. at *2 n.2, 4–5. The
district court for the District of Arizona dismissed Eldridge’s
petition as an impermissible second or successive petition
under the Antiterrorism and Effective Death Penalty Act
(AEDPA), citing to 28 U.S.C. §§ 2244(a) and 2244(b).
Eldridge v. Blanckensee, 21-00081, 2021 U.S. Dist. LEXIS
36254, at *4 (D. Ariz. Feb. 24, 2021).
Eldridge moved the District of Arizona court to
reconsider its dismissal, arguing that the court improperly
dismissed the case “without any notice.” Eldridge contends
that his new petition challenges the Commission’s 2019
8 ELDRIDGE V. HOWARD

decision to issue a three-year set-off, which he “could not
[have] foresee[n]” when he filed his earlier habeas petitions.
The District of Arizona denied Eldridge’s motion to
reconsider. See Eldridge v. Blanckensee, No. CV 21-00081-
TUC-RCC (LAB), ECF No. 29 at *5 (D. Ariz. Mar. 18,
2021). Relying on the abuse of the writ doctrine, it
concluded that Eldridge’s claims were “substantially similar
to the claims [he] raised in at least two other § 2241
petitions.” See id.
This appeal followed. We have jurisdiction under 28
U.S.C. §§ 1291 and 2253(a), and we reverse and remand.
II.
We first must determine whether 28 U.S.C. § 2253(c)(1)
requires Eldridge to obtain a COA. If a COA is required in
this case, then it is a jurisdictional requirement that cannot
be waived, and we must decide whether to issue one. See
Gonzalez v. Thaler, 565 U.S. 134, 142 (2012) (noting that 28
U.S.C. § 2253(c)(1) is jurisdictional); Miller-El v. Cockrell,
537 U.S. 322, 336 (2003). Absent a COA requirement,
however, we have jurisdiction to decide whether the district
court erred in dismissing Eldridge’s instant petition. See,
e.g., Harbison v. Bell, 556 U.S. 180, 183–84 (2009)
(concluding that a COA was not required and thereafter
deciding the issue raised on appeal); Harrison v. Ollison,
519 F.3d 952, 958 (9th Cir. 2008) (“In the absence of a
statutory COA requirement for federal prisoners bringing
legitimate § 2241 petitions, we cannot require one as a
condition for our exercise of jurisdiction.”).
 ELDRIDGE V. HOWARD 9

Section 2253(c)(1) provides:

Unless a circuit justice or judge issues a
certificate of appealability, an appeal may not
be taken to the court of appeals from--
(A) the final order in a habeas corpus
proceeding in which the detention
complained of arises out of process
issued by a State court; or
(B) the final order in a proceeding under
section 2255.

(emphasis added).
We agree with both parties that § 2253(c)(1)(B) does not
apply because Eldridge properly brought his habeas petition
under 28 U.S.C. § 2241. See Herndon v. U.S. Parole
Comm’n, 961 F. Supp. 2d 138, 141 (D.D.C. 2013)
(describing § 2241 as the “exclusive federal avenue
available to a District of Columbia prisoner challenging the
manner of execution of a sentence, rather than the sentence
itself” (cleaned up)). Thus, the issue is whether
§ 2253(c)(1)(A)’s “State Court” requirement applies to
habeas corpus proceedings, such as Eldridge’s, which
challenge a detention that arose out of a process issued by
the District of Columbia Superior Court.4

4
Both parties assume that Eldridge’s detention “arises out of process
issued by” the District of Columbia Superior Court, and both cite
Hayward v. Marshall, 603 F.3d 546 (9th Cir. 2010) (en banc), overruled
on other grounds by Swarthout v. Cooke, 562 U.S. 216 (2011), in support
of their conclusion. Because the parties do not brief the issue, we
assume, without deciding, the same.
10 ELDRIDGE V. HOWARD

We begin with the text of § 2253(c)(1)(A). See Ross v.
Blake, 578 U.S. 632, 638 (2016). Here, the text requires a
COA for an appeal from a final order in a habeas proceeding
“in which the detention complained of arises out of process
issued by a State court[.]” 28 U.S.C. § 2253(c)(1)(A). None
of AEDPA’s habeas corpus amendments define “State
court.” See generally 28 U.S.C. §§ 2241–2255. Congress
also does not provide a broadly applicable definition of
“State” in the Dictionary Act. See 1 U.S.C. §§ 1–8.
Without a definition, we must look to the ordinary
meaning of the term “State court” and “state.” See, e.g.,
Wooden v. United States, 142 S. Ct. 1063, 1069 (2022)
(looking to the ordinary meaning of the word “occasion”);
Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1738 (2020)
(looking to the ordinary meaning of the language used in
Title VII at the time the statute was enacted). The ordinary
meaning of the word “state” does not—and did not at the
time § 2254 was enacted—include the District of Columbia.
See, e.g., Zachary B. Wolf, Why DC Should (and Should
Not) Be the 51st State, CNN (June 26, 2020, 8:09 PM)5;
Tessa Berenson, Here’s Why Washington D.C. Isn’t a State,
Time, (Apr. 15, 2016, 1:57 PM)6. The Supreme Court has
recognized that “[t]he District of Columbia is
constitutionally distinct from the States[.]” Palmore v.
United States, 411 U.S. 389, 395 (1973). And the Federal
Rules of Criminal Procedure expressly define the term
“State” to include non-state entities such as the District of
Columbia, Fed. R. Crim. P. 1(b)(9), suggesting that without

5
https://www.cnn.com/2020/06/26/politics/dc-statehood-101 (last visited
May 23, 2023).
6
https://time.com/4296175/washington-dc-statehood-history/ (last visited
May 23, 2023).
 ELDRIDGE V. HOWARD 11

this express definition, the term “state” would not include
the District of Columbia. See Palmore, 411 U.S. at 395.
Congress’s use of the word “state,” in Title 28, the same
Title to which § 2253 belongs, confirms this ordinary
meaning: when Congress wishes to treat the District of
Columbia as a state in Title 28, it expressly says so.7 See 28
U.S.C. § 1257 (expressly defining “highest court of a State”
to include the District of Columbia Court of Appeals, but not
defining “statute of any State” to include District of
Columbia statutes for purposes of Supreme Court
jurisdiction); 28 U.S.C. § 1332(e) (expressly defining
“States” to include, among other things, the District of
Columbia for purposes of diversity jurisdiction); 28 U.S.C
§ 1451(1) (expressly defining “State court” to include the
Superior Court of the District of Columbia for purposes of
removal jurisdiction).
Congress has defined “state” to include the District of
Columbia in other contexts too. See, e.g., 34 U.S.C.
§ 20142(f) (rules for closed circuit televising of court
proceedings for crime victims); 42 U.S.C. § 6802(6)
(electric utility rate design initiatives); 10 U.S.C.
§ 1045(e)(1) (voluntary withholding of state income tax
from retired or retainer pay for current and past members of

7
The dissent acknowledges that “absent strong textual reasons to do
otherwise, we ought to embrace the ordinary meaning of a term.”
However, the dissent concludes that here, we should not rely on the
ordinary understanding “that D.C. is not a ‘State’” once we take the term
“State court” in the context of the statute as a whole. The dissent ignores
the fact that before reaching our conclusion, we consider Congress’s use
of the word “state” in Title 28 and that we apply the relevant canons of
construction. Both of these contextual factors support our determination
that the ordinary use of the term “state” is appropriately used here.
12 ELDRIDGE V. HOWARD

military); 12 U.S.C. § 4001(21) (expedited funds availability
in banking); 5 U.S.C.
§ 5707a(f)(3) (federal employee travel expense
requirements and procedures). By specifying that the word
“state” includes the District of Columbia in certain
circumstances within these statutes, Congress is
acknowledging that the ordinary definition of the word
“state” does not include the District of Columbia.
The Supreme Court’s decision in Palmore confirms this
conclusion. In Palmore, the Court interpreted whether 28
U.S.C. § 1257 gave the United States Supreme Court
appellate jurisdiction. See 411 U.S. at 394. There, Palmore
was convicted in the District of Columbia Superior Court of
carrying an unregistered pistol after having been convicted
of a felony, which violated the District of Columbia Code.
See id. at 391. Palmore appealed his conviction to the
District of Columbia Court of Appeals, arguing that the
provision of the District of Columbia Code establishing the
Superior Court violated the Constitution because only an
Article III court could constitutionally try him for a felony
prosecution under the District of Columbia Code. See id. at
392–93. The District of Columbia Court of Appeals rejected
his argument. See id. at 393.
When Palmore sought review by the Supreme Court, he
argued that the Supreme Court had jurisdiction under
§ 1257(2) because the statute he challenged constituted a
“statute of any state.” See Palmore, 411 U.S. at 394–95. The
Supreme Court rejected his argument, noting that even
though Congress “plainly provided [in the text of § 1257]
that the District of Columbia Court of Appeals should be
treated as the ‘highest court of a state[,]’” Congress did not
provide that the phrase “‘statute of any state’” included the
 ELDRIDGE V. HOWARD 13

District of Columbia Code. See id. at 395. The Court
recognized that “[a] reference to ‘state statutes’ would
ordinarily not include provisions of the District of Columbia
Code,” because “[t]he District of Columbia is
constitutionally distinct from the States.” 8 See id. The
Supreme Court emphasized that Congress “legislated with
care, and that had Congress intended to equate the District
Code and state statutes for the purposes of [§] 1257, it would
have done so expressly, and not left the matter to mere
implication.” Id.
Here, we assume Congress “legislated with care” when
it did not expressly include District of Columbia courts as
state courts in 28 U.S.C. § 2253(c). Just as the Supreme
Court in Palmore determined that Congress would have
expressly provided that the District of Columbia Code
equated to a state statute if it intended such result, so too do
we when interpreting the phrase “process issued by a State
court.” 28 U.S.C. § 2253(c); see Palmore, 411 U.S. at 395.
If Congress meant for “State court” to include the District of
Columbia Superior Court, it could have easily said so. See,
e.g., 28 U.S.C § 1451(1) (defining “State court” to include
the Superior Court of the District of Columbia for purposes
of removal jurisdiction). Congress did not. A reference to a
state court ordinarily does not include a District of Columbia
court, and thus we must follow Congress’s express words
when interpreting 28 U.S.C. § 2253.
We also observe Palmore’s warning that
“[j]urisdictional statutes are to be construed ‘with precision

8
The dissent suggests that we look to the “legal (rather than a literal)
meaning” of “State court” here. But Palmore itself illustrates that the
Supreme Court has not distinguished between a “literal meaning” and
“legal meaning” of the term “state.”
14 ELDRIDGE V. HOWARD

and with fidelity to the terms by which Congress has
expressed its wishes[.]’” Palmore, 411 U.S. at 396 (quoting
Cheng Fan Kwok v. INS, 392 U.S. 206, 212 (1968)). The
Supreme Court has repeatedly instructed courts to exercise
caution when declaring statutory provisions jurisdictional,
and to do so only if a statute clearly states it is jurisdictional.
See, e.g., Fort Bend Cnty. v. Davis, 139 S. Ct. 1843, 1849
(2019); Sebelius v. Auburn Reg’l Med. Ctr., 568 U.S. 145,
153 (2013); Henderson ex rel. Henderson v. Shinseki, 562
U.S. 428, 435 (2011). Section 2253(c)(1) is a jurisdictional
requirement. See Gonzalez, 565 U.S. at 142; Miller-El, 537
U.S. at 336. Therefore, we must exercise caution and not
extend 28 U.S.C. § 2253(c)(1)’s text beyond its ordinary
meaning to include the District of Columbia Superior Court.
To do otherwise would limit Congress’s conferral of
jurisdiction without a clear textual basis, a move that the
Supreme Court has repeatedly cautioned against doing.
Appellees and our unpublished opinion in Johnson v.
Clay, 539 F. App’x 748, 748 (9th Cir. 2013), rely heavily on
our sister circuit’s decision in Madley v. U.S. Parole
Comm’n, 278 F.3d 1306 (D.C. Cir. 2002).9 Madley
concluded that the District of Columbia Superior Court is a
state court for purposes of 28 U.S.C. § 2253(c).10 278 F.3d

9
Although the dissent asserts that “our court treated it as settled that” the
District of Columbia qualifies as a “State court” for purposes of 28
U.S.C. § 2253(c)(1)(A), it also acknowledges that “we did so in a non-
precedential memorandum disposition and another panel of this court
may disagree with that conclusion.” The dissent thus concedes that this
question, which has never been decided in a precedential disposition in
our circuit, is far from “settled.”
10
We also recognize that several of our sister circuits have followed
Madley. See Sanchez-Rengifo v. Caraway, 798 F.3d 532, 535 (7th Cir.
2015); Eldridge v. Berkebile, 791 F.3d 1239, 1244 (10th Cir. 2015);
 ELDRIDGE V. HOWARD 15

at 1310. In Madley, the court relied on its prior precedent
that interpreted a previous version of § 2253(c). See id. at
1308–09. It noted that before AEDPA was enacted,
§ 2253(c) required a certificate of “probable cause” instead
of a COA. See id. at 1308. The relevant language of the
previous version was the same:

An appeal may not be taken to the court of
appeals from the final order in a habeas
corpus proceeding where the detention
complained of arises out of process issued by
a State court, unless the justice or judge who
rendered the order or a circuit justice or judge
issues a certificate of probable cause.

Id. (quoting Pub. L. No. 773, § 2253, 62 Stat. 869, 967
(1948) (emphasis added)). In 1986, before the passage of
AEDPA, the Court of Appeals for the District of Columbia
Circuit concluded that “District of Columbia prisoners are
‘state’ prisoners for purposes of [the probable cause]
requirement.” Garris v. Lindsay, 794 F.2d 722, 724 n.8
(D.C. Cir. 1986). Madley observed that when Congress
revised § 2253(c) it left unchanged the phrase “‘the
detention complained of arises out of process issued by a
State court,’” and “made no effort to disapprove Garris.”

Terry v. Deeboo, 473 F. App’x 282, 283 (4th Cir. 2012), Wilson v. U.S.
Parole Comm’n, 652 F.3d 348, 351–52 (3d. Cir. 2011). While we agree
with the dissent that “as a general rule, we decline to create a circuit split
unless there is a compelling reason to do so[,]” Padilla-Ramirez v. Bible,
882 F.3d 826, 837 (9th Cir. 2017) (internal quotation marks and citations
omitted), we find compelling reasons to do so here, including the
ordinary meaning of the word “state” and Congress’s omission of a clear
indication that the term “State court” includes the District of Columbia
in § 2253(c)(1)(A).
16 ELDRIDGE V. HOWARD

See 278 F.3d at 1309. Because Congress is “presumed to
adopt existing judicial interpretations of a statute when it re-
enacts without change[,]” Madley reasoned that Garris’s
decision and Congress’s later enactment of AEDPA means
that “a court of the District is a state court for the purposes
of [§ 2253(c)].” Id. (citing Lorillard v. Pons, 434 U.S. 575,
580 (1978)).
We disagree. At the time Congress passed AEDPA, it
was not well established that the phrase “the detention
complained of arises out of process issued by a State court”
included the District of Columbia Superior Court. In fact,
Garris appears to be the sole federal decision before the
passage of AEDPA that had interpreted the phrase to include
the District of Columbia Superior Court. See id.; Garris,
794 F.2d at 724 n.8. The only other court of which we are
aware to consider the District of Columbia Superior Court
issue before AEDPA was passed was a Tenth Circuit
decision, Blango v. Thornburgh, 942 F.2d 1487 (10th Cir.
1991), in which the court declared the issue was “still an
open question” and expressly declined to resolve it. See id.
at 1488 n.1 (“[W]e express no view on the matter” of
“[w]hether prisoners sentenced by the District of Columbia
Superior Court . . . should be treated as prisoners under
sentence of state court[.]”). Blango made no mention to
Garris in its opinion, which suggests that Garris’s
declaration on the issue was either unknown or
unremarkable.
Indeed, Garris declared its view in a single footnote, and
thus the issue was far from the central focus of the case. See
794 F.2d at 724 n.8. Rather, the main issue in Garris was
whether the court should issue a certificate of probable cause
to the prisoner, not whether the prisoner needed one. See id.
at 725 (“[W]e are called upon to determine whether we
 ELDRIDGE V. HOWARD 17

should [issue a certificate of probable cause.]”). Garris, the
sole case to state that the District of Columbia Superior
Court is a state court for purposes of a certificate of probable
cause under § 2253(c) prior to AEDPA, does not
demonstrate that such an interpretation was well-settled.
Nor does the passage of AEDPA suggest that Congress
approved of Garris’s holding. There is no evidence that
Congress approved, or was even aware of, Garris’s view
when it enacted AEDPA. Instead, “Congress enacted
AEDPA to reduce delays in the execution of state and federal
criminal sentences, particularly in capital cases[.]”
Woodford v. Garceau, 538 U.S. 202, 206 (2003); see also
Williams v. Taylor, 529 U.S. 362, 386 (2000). Congress
shortened the time for prisoners to file habeas petitions;
before AEDPA, “state prisoners had almost unfettered
discretion in deciding when to file a federal habeas petition.”
Calderon v. U.S. Dist. Ct. (Beeler), 128 F.3d 1283, 1286 (9th
Cir. 1997), partially overruled on other grounds by
Calderon v. U.S. Dist. Ct. (Kelly), 163 F.3d 530, 540 (9th
Cir. 1998), abrogated on other grounds by Garceau, 538
U.S. at 206. But “AEDPA dramatically changed this
landscape, shortening the time for filing a federal habeas
petition to one year.” Id.
Related to Congress’s goal of shortening the time period
for petitioners to file federal habeas petitions, it also sought
to promote principles of federalism and respect for state
court adjudications. See Shinn v. Ramirez, 142 S.Ct. 1718,
1730–32 (2022) (describing AEDPA’s design in preserving
states’ primary role in criminal law enforcement); Williams
v. Taylor, 529 U.S. 420, 436 (2000) (“There is no doubt
Congress intended AEDPA to advance [the principles of
comity, finality, and federalism.]”); Cook v. Kernan, 948
F.3d 952, 965 (9th Cir. 2020) (outlining the deferential
18 ELDRIDGE V. HOWARD

standard of review for habeas petitions filed on behalf of
petitioners in state custody and which were adjudicated in
state court).
Here, the issue we resolve—whether the District of
Columbia Superior Court is considered a “State court” for
purposes of deciding whether Congress requires a COA—
does not implicate the main purposes for which Congress
passed AEDPA. Thus, we decline to presume that Congress
decided, in re-codifying the relevant language after Garris’s
holding, that the District of Columbia Superior Court counts
as a “State court” for purposes of § 2253.
Even if the presumption on which Madley relies should
apply to this case, that presumption must yield to the
“cardinal canon” that “courts must presume that a legislature
says in a statute what it means and means in a statute what it
says there.” Conn. Nat’l. Bank v. Germain, 503 U.S. 249,
253–54 (1992); see also Babb v. Wilkie, 140 S. Ct. 1168,
1177 (2020) (“Where . . . the words of a statute are
unambiguous, the judicial inquiry is complete.”) (cleaned
up). Because the plain meaning of the word “state” does not
include the District of Columbia, and because Congress has
expressly defined “state” to include the District of Columbia
as a state in other contexts, we hold that § 2253(c)’s
language, “in which the detention complained of arises out
of process issued by a State court,” does not include the
District of Columbia Superior Court. Prisoners whose
detention arises out of process issued by a District of
Columbia court are not required to obtain a COA to appeal a
federal district court’s denial of habeas relief. Thus, the
COA jurisdictional requirement does not pose a barrier to
Eldridge’s appeal.
 ELDRIDGE V. HOWARD 19

III.
Next, we must decide whether the district court properly
dismissed Eldridge’s instant petition as an abuse of the writ.
We review dismissal of a habeas petition de novo. Alaimalo
v. United States, 645 F.3d 1042, 1047 (9th Cir. 2011).
The abuse of the writ doctrine “forbids the
reconsideration of claims that were or could have been raised
in a prior habeas petition.” Kelly, 163 F.3d at 538. Under
this doctrine, “a successive petition that raises identical
grounds for relief as a prior petition must be dismissed unless
the petitioner can show (1) cause for bringing a successive
petition and that prejudice would result or (2) that a
fundamental miscarriage of justice would result from failure
to entertain the claim.” Alaimalo, 645 F.3d at 1049
(emphasis added). The doctrine “refers to a complex and
evolving body of equitable principles informed and
controlled by historical usage, statutory developments, and
judicial decisions.” McCleskey v. Zant, 499 U.S. 467, 489
(1991).
Liberally construing Eldridge’s current petition, as we
must, Eldridge primarily argues that the Commission’s
decision to give him a three-year set-off in 2019 is arbitrary
and capricious, because the Commission gave him a one-
year set-off in 2018 and both parole requests were denied for
the same reason—namely, he was an untreated sex offender.
See Draper v. Rosario, 836 F.3d 1072, 1080 (9th Cir. 2016)
(noting court must construe pro se filings liberally)11.

11
Eldridge proceeded pro se when filing his original habeas petition.
The court later appointed counsel to represent Eldridge in this appeal.
Order at 1, Eldridge v. Blanckensee, No. 21-15616 (9th Cir. Oct. 20,
2021), ECF No. 6.
20 ELDRIDGE V. HOWARD

On appeal, Eldridge argues that this habeas petition was
not abusive “because it contained claims that he could not
have raised in his prior habeas filings—challenges to the
2018 and 2019 denials of parole, which had not occurred
when he filed those prior habeas petitions.” Thus, Eldridge
argues his 2019 denial is not second or successive. And even
if it is second or successive, Eldridge argues he has good
cause for not raising his challenge in 2016 because he could
not foresee the Commission’s future rationales in 2018 and
2019.
Our discussion in Brown v. Muniz, 889 F.3d 661, 674
(9th Cir. 2018) guides our analysis here. Drawing on pre-
AEDPA abuse of the writ principles, we stated that “to
determine whether [a claim] is ‘second or successive,’ [we]
must look to the substance of the claim . . . and decide
whether the petitioner had a full and fair opportunity to raise
the claim in the prior application.” Id. (quoting Magwood v.
Patterson, 561 U.S. 320, 346 (2010) (Kennedy, J.,
dissenting)). “[A] petitioner ‘had no fair opportunity to raise
the claim in the prior application’ if ‘[1] the claim was not
yet ripe at the time of the first petition, or [2] where the
alleged violation occurred only after the denial of the first
petition.’” Id. (quoting Magwood, 561 U.S. at 345–46
(Kennedy, J. dissenting)). Here, looking to the substance of
Eldridge’s claim—that the Commission acted arbitrarily and
capriciously in 2019 when it issued a three-year set-off—
Eldridge did not have a fair opportunity to raise this claim in
2016 or 2018 because “the alleged violation occurred only
after the denial of” his previous habeas petitions. Id.
Additionally, the district court did not address the merits
of the set-off issues in its decision denying Eldridge’s habeas
petition in 2016, even though Eldridge raised the issue. See
Eldridge v. Oliver, 2017 WL 2812824, at *1–8. Instead, it
 ELDRIDGE V. HOWARD 21

denied his petition on other grounds. See id. at *5–9. And
the district court dismissed Eldridge’s 2018 petition, which
also raised the three-year set-off issue, without deciding it on
its merits. See Eldridge v. U.S. Parole Comm’n, 2018 WL
10426189, at *2. Thus, because no court has addressed
Eldridge’s three-year set-off claims regarding his 2016
parole denial, he did not abuse the writ by raising the 2019
denial issue in his instant habeas petition. See Hill v. State
of Alaska, 297 F.3d 895, 899 (9th Cir. 2002) (“Because the
district court has never addressed [petitioner’s] claims
relating to mandatory parole on the merits, and those claims
could not have been included in earlier petitions challenging
his conviction and sentence, [petitioner] is not obliged to
secure this court’s permission prior to filing his habeas
petition in the district court.”).
The Government resists Eldridge’s argument and cites to
a Third Circuit decision in Benchoff v. Colleran, 404 F.3d
812 (3d. Cir. 2005) and an unpublished Eleventh Circuit
decision in Watson v. Coleman, 644 F. App’x 996 (11th Cir.
2016). Both cases are distinguishable from Eldridge’s
situation.
Benchoff was denied parole three times. See 404 F.3d at
817–18. Between the second and third denial, Benchoff
filed a habeas petition making claims related to his
underlying criminal trial. See id. at 813–14. Benchoff filed
another habeas petition between the second and third denial,
this time challenging his first two parole denials. See id. at
814. Benchoff’s third parole denial was initially for the
same reason as the first two, namely that the parole board
had determined that “the fair administration of justice cannot
be achieved through [his] release on parole.” Id. (internal
quotation marks omitted). The Third Circuit found that
“even one of the parole denials would have been sufficient
22 ELDRIDGE V. HOWARD

for Benchoff to formulate his complaint.” Id. at 818.
Because Benchoff did not have a legitimate excuse for
failing to raise the denial of the due process parole claim in
his first petition, which also challenged his underlying
criminal trial, the court found an abuse of the writ despite
Benchoff’s argument that he could not have challenged his
third parole denial in his first petition. Id. at 818–19.
Eldridge’s case differs from Benchoff. Eldridge’s
current petition raises a specific challenge to the
Commission’s decision to issue a three-year set-off in 2019,
after the Commission had issued him a one-year set-off for
similar reasons in 2018. Unlike in Benchoff, in which the
petitioner raised the same claims because the parole board’s
language in its notice of denial was identical, 404 F.3d at
818, Eldridge did not make an identical argument when he
filed his 2016 and 2018 habeas petitions because he could
not have known of the subsequent one-year set-off that was
issued later in 2018. Thus, he did not abuse the writ when
filing the instant petition.
It is true, as the Government highlights, that Eldridge
challenged the Commission’s 2016 decision to issue him
three-year set-offs in his 2016 petition, arguing it was
“punitive” because it would not take three years to complete
the sex offender treatment program. And in his 2018
petition, he argued his 2016 three-year set-off “went beyond
the ordinary 12-month rehearing guidelines[.]” But at the
time Eldridge lodged his 2016 and 2018 petitions
challenging his 2016 set-off, the Commission had never
issued him a one-year set-off, much less one for similar
reasons when it issued the three-year set-offs. Similarly, the
Commission had not yet issued its 2019 decision. Thus, the
Commission’s 2018 and 2019 decisions served as necessary
factual predicates for Eldridge’s current petition, and these
 ELDRIDGE V. HOWARD 23

facts were unavailable to Eldridge at the time of his 2016 and
2018 petitions.
The Government’s citation to the unpublished Eleventh
Circuit disposition in Watson similarly does not alter our
analysis. In Watson, a D.C. Code offender was convicted in
1978 and sentenced to serve a minimum of 30 years in
prison. See 644 F. App’x at 998. Watson had his first parole
hearing in 2004, where the Commission denied parole and
calculated his eligible parole date using the 2000 guidelines.
See id. In 2009, Watson filed a habeas petition challenging
the calculation of his eligible parole date. See id. The
district court rejected this claim, and the Eleventh Circuit
affirmed. See id. at 999. Watson then filed another habeas
petition, alleging among other things, that the Commission’s
use of the 2000 guidelines violated the Ex Post Facto Clause
when the Commission denied him parole in 2011. See
Watson v. Warden, FCC Coleman-USP I, 5:12-491-0C-27,
2015 WL 78775, at *2 (M.D. Fla. Jan. 6, 2015). The
Eleventh Circuit held that the abuse of the writ doctrine
barred his claim because his “central allegation—that the
2000 guidelines resulted in him receiving a harsher sentence
than what was available at the time of his crime—has been
available to him since the denial of parole in 2004.” Watson,
644 F. App’x at 1000. Thus, even though he could not have
challenged his 2011 eligible parole denial when he filed his
2009 petition, the Eleventh Circuit still found that the abuse
of the writ doctrine applied. See id.
Eldridge’s current petition differs from Watson’s: unlike
Watson, who could have alleged the Ex Post Facto violation
in 2009 because Watson knew all the relevant facts at that
time, Eldridge did not have all the necessary facts in 2016 to
argue that the Commission abused its discretion in 2019 in
light of the Commission’s 2018 decision.
24 ELDRIDGE V. HOWARD

Additionally, no district court has ruled on Eldridge’s
set-off claims on the merits, even though Eldridge raised the
three-year set-offs in his 2016 and 2018 petitions. In fact, a
similar occurrence happened in Watson. Prior to Watson’s
2009 petition, he had filed “numerous” habeas challenges
arguing that the Commission incorrectly calculated his
parole eligibility date. See Watson, 644 F. App’x at 998.
The district court originally dismissed Watson’s 2009
petition as an abuse of the writ, but the Eleventh Circuit held
that his petitions “had never been adjudicated on the merits
and that his case had been improperly dismissed by the
district court.” Id. at 999 (citing Watson v. United States,
392 F. App’x 737, 742 (11th Cir. 2010)). Thus, just as the
Eleventh Circuit held it would be inappropriate to dismiss
Watson’s 2009 petition as an abuse of the writ given the
district court’s failure to decide the merits of his previous
petitions, we hold that Eldridge’s instant petition is not an
abuse of the writ because his precise claim has not been
adjudicated on the merits. See id.
IV.
In conclusion, we hold that Eldridge need not obtain a
COA because Congress did not define or include the District
of Columbia Superior Court as a “State court” in 28 U.S.C.
§ 2253(c), where it had expressly done so in that and other
statutes. We further hold that the district court erred in
dismissing Eldridge’s instant petition as an abuse of the writ
when Eldridge could not have possibly raised the same
claims in prior petitions. We reverse and remand the case to
the district court to decide Eldridge’s instant petition on its
merits.
 ELDRIDGE V. HOWARD 25

BUMATAY, Circuit Judge, dissenting:

Congress has determined that we have no jurisdiction to
review habeas petitions filed by prisoners detained out of
process “issued by a State court” unless the state prisoner
first obtains a certificate of appealability. 28 U.S.C.
§ 2253(c)(1)(A). And we may only issue a certificate of
appealability “if the applicant has made a substantial
showing of the denial of a constitutional right.”
§ 2253(c)(2). In this case, we are asked to relitigate a
virtually settled question—whether the Superior Court of the
District of Columbia is a “State court” under
§ 2253(c)(1)(A).
Every circuit court to consider this question has reached
the same result—that the D.C. Superior Court is a “State
court” under habeas law and prisoners challenging detention
arising from a D.C. Superior Court conviction must obtain a
certificate of appealability before appealing. And this
consensus is significant—the D.C., Third, Fourth, Seventh,
and Tenth Circuits have all embraced this approach. See
Madley v. U.S. Parole Comm’n, 278 F.3d 1306, 1310 (D.C.
Cir. 2002); Wilson v. U.S. Parole Comm’n, 652 F.3d 348,
351–52 (3d Cir. 2011); Terry v. Deeboo, 473 F. App’x 282,
283 (4th Cir. 2012) (unpublished); Sanchez-Rengifo v.
Caraway, 798 F.3d 532, 535 (7th Cir. 2015); Eldridge v.
Berkebile, 791 F.3d 1239, 1243–44 (10th Cir. 2015).
And, until today, so had we. Ten years ago, our court
treated it as settled that “the Superior Court of the District of
Columbia qualifies as a state court under 28 U.S.C.
§ 2253(c)(1)(A).” Johnson v. Clay, 539 F. App’x 748 (9th
Cir. 2013) (unpublished). To be sure, we did so in a non-
precedential memorandum disposition and another panel of
this court may disagree with that conclusion. But that we
26 ELDRIDGE V. HOWARD

once thought this proposition of law so uneventful that it
didn’t warrant a published opinion just shows how widely
accepted it is.
Of course, we have “no warrant to ignore clear statutory
language on the ground that other courts have done so.”
Milner v. Dep’t of Navy, 562 U.S. 562, 576 (2011). But
“absent a strong reason to do so,” our general practice is to
refrain from creating “a direct conflict with other circuits.”
United States v. Cuevas-Lopez, 934 F.3d 1056, 1067 (9th
Cir. 2019) (simplified). And here, there’s no strong reason.
Indeed, given the text and its context, the best reading of
§ 2253(c)(1)(A) is that the D.C. Superior Court is a “State
court.” Rather than opening a circuit split, we should have
stuck with consensus.
So before considering this appeal, we should have first
determined whether Clinton Eldridge deserves a certificate
of appealability. On that score, I would conclude he does
not. I thus would have dismissed this appeal, and I
respectfully dissent.
I.
Section 2253(c)(1)(A) provides that, “[u]nless 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.”
28 U.S.C. § 2253(c)(1)(A). Thus, when it’s required, “[t]he
issuance of a certificate of appealability . . . is a
jurisdictional prerequisite to appeal in a habeas corpus
proceeding.” Lord v. Lambert, 347 F.3d 1091, 1094 (9th Cir.
2003). Because Eldridge’s detention arises from a D.C.
Superior Court conviction, our jurisdiction to hear this
 ELDRIDGE V. HOWARD 27

appeal turns on whether that court is a “State court” under
§ 2253(c)(1)(A).
In the majority’s view, no certificate of appealability is
necessary here because the D.C. Superior Court isn’t a “State
court.” The majority relies mainly on the fact that the
District of Columbia is not a “State” under the ordinary
meaning of the word. The majority then looks to other
federal statutes where Congress has expressly defined “State
court” to include the D.C. Superior Court and concludes that
Congress’s silence in § 2253(c)(1)(A) should be read as a
purposeful exclusion.
In fairness, the majority’s reading of § 2253(c)(1)(A) is
plausible. We all know that D.C. is not a “State.” And
generally, absent strong textual reasons to do otherwise, we
ought to embrace the ordinary meaning of a term. But
sometimes, adhering only to ordinary meaning—without
understanding legal context—may cross into literalism and
blind us to the best reading of a statute. Indeed, “[w]ords
are to be understood in their ordinary, everyday meanings—
unless the context indicates that they bear a technical sense.”
Antonin Scalia & Bryan A. Garner, Reading Law: The
Interpretation of Legal Texts 69 (2012) (“Reading Law”).
And in my view, our duty is always “to ‘seek the best
reading of the statute by interpreting the words of the statute,
taking account of the context of the whole statute, and
applying the agreed upon semantic canons.’” Rojas v. FAA,
989 F.3d 666, 694 n.5 (9th Cir. 2021) (en banc) (Bumatay,
J., dissenting in part) (quoting Brett M. Kavanaugh, Fixing
Statutory Interpretation, 129 Harv. L. Rev. 2118, 2121
(2016) (emphasis added)). As Learned Hand once wrote, “a
sterile literalism . . . loses sight of the forest for the trees.”
Reading Law, at 356 (quoting New York Tr. Co. v.
Commissioner, 68 F.2d 19, 20 (2d. Cir. 1933)). Instead, we
28 ELDRIDGE V. HOWARD

must recognize that “[t]he full body of a text contains
implications that can alter the literal meaning of individual
words.” Id.
While we should never breezily depart from ordinary
meaning, here “the context of the whole statute” and “agreed
upon semantic canons” all point to the best reading of
§ 2253(c)(1)(A) being that “State court” includes the D.C.
Superior Court.
A.
The “prior construction” canon derives from the
principle that when Congress enacts legislation borrowing
language with a settled meaning, the newly enacted
legislation “brings the old soil with it.” Sekhar v. United
States, 570 U.S. 729, 733 (2013) (quoting Felix Frankfurter,
Some Reflections on the Reading of Statutes, 47 Colum. L.
Rev. 527, 537 (1947)). “In other words, when Congress
adopts a phrase with a settled judicial interpretation, absent
some indication to the contrary, we presume that Congress
chose to give the phrase its established meaning.” United
States v. Randall, 34 F.4th 867, 875 (9th Cir. 2022)
(simplified). Under the prior-construction canon, “Congress
is presumed to be aware of an administrative or judicial
interpretation of a statute and to adopt that interpretation
when it re-enacts a statute without change.” Lorillard v.
Pons, 434 U.S. 575, 580 (1978); Chugach Mgmt. Servs. v.
Jetnil, 863 F.3d 1168, 1174 (9th Cir. 2017).
The current version of § 2253(c)(1)(A) was enacted as
part of the Anti-Terrorism and Effective Death Penalty Act
(“AEDPA”) in 1996. But Congress didn’t operate on a blank
slate. Prior to AEDPA, back in 1948, Congress adopted the
predecessor of § 2253(c)(1)(A), which required a “certificate
 ELDRIDGE V. HOWARD 29

of probable cause” rather than a “certificate of appealability”
to appeal a habeas petition. Before 1996, § 2253 said:

An appeal may not be taken to the court of
appeals from the final order in a habeas
corpus proceeding where the detention
complained of arises out of process issued by
a State court, unless the justice or judge who
rendered the order or a circuit justice or judge
issues a certificate of probable cause.

Pub. L. No. 80-773, § 2253, 62 Stat. 869, 967 (1948).
Ten years before AEDPA, the D.C. Circuit grappled with
the status of the modern D.C. court system under the 1948
language. Garris v. Lindsay 794 F.2d 722, 724 n.8 (D.C.
Cir. 1986) (per curiam). Given that Congress created the
D.C. courts to be analogous to State courts, the D.C. Circuit
concluded that “District of Columbia prisoners are ‘state’
prisoners for purposes of this requirement.” Id.; see also
Milhouse v. Levi, 548 F.2d 357, 360 n.6 (D.C. Cir. 1976)
(“[T]his Court has treated local courts as ‘state’ courts for
the purposes of exhaustion and federal habeas corpus
jurisdiction.”). That conclusion remained unchallenged up
through Congress’s enactment of AEDPA and
§ 2253(c)(1)(A).
When, as here, “Congress adopts a new law
incorporating sections of a prior law, Congress normally can
be presumed to have had knowledge of the interpretation
given to the incorporated law.” Lorillard, 434 U.S. at 581.
In other words, because § 2253(c)(1)(A) “perpetuat[es] the
wording” of the 1948 version, we can “presume[]” that
Congress meant to “carry forward” the uncontested
interpretation of “State court” from Garris. Tex. Dep’t of
30 ELDRIDGE V. HOWARD

Hous. & Cmty. Affs. v. Inclusive Cmtys. Project, Inc., 576
U.S. 519, 536 (2015) (quoting Reading Law, at 322). After
all, Congress presumably did not ignore the legal gloss that
the D.C. Circuit—the court with the greatest connection to
the District—placed on the meaning of the words of
§ 2253(c)(1)(A).
The D.C. Circuit followed this analysis in reaffirming its
position that § 2253(c)(1)(A)’s certificate of appealability
requirement applies to D.C. prisoners. See Madley, 278 F.3d
at 1309. As Judge Sentelle wrote, “Congress’s 1996
amendment to section 2253 left th[e] interpreted language
unchanged and made no effort to disapprove Garris.” Id.
Judges from the Third, Fourth, Seventh, Ninth, and Tenth
Circuits all found this analysis persuasive and relied on
Madley to require a certificate of appealability for D.C.
prisoners. See Wilson, 652 F.3d at 351–52; Terry, 473 F.
App’x. at 283; Sanchez-Rengifo, 798 F.3d at 535; Johnson,
539 F. App’x at 748; Eldridge, 791 F.3d at 1243–44.
To be fair, as the majority points out, the D.C. Circuit is
the only court that had to address whether the pre-AEDPA
§ 2253 applied to D.C. courts. Thus, in the majority’s view,
that court’s interpretation was not “well-settled” enough to
warrant applying the prior-construction canon. The majority
raises a valid question: What degree of judicial consensus is
needed for an interpretation to be “settled”? In its strongest
form, we can agree that meaning is definitively settled by a
ruling of “a court of last resort.” Reading Law, at 323. But
the prior construction canon also applies to “uniform
holdings of lower courts and even to well-established agency
interpretations.” Id. at 324; see also Bragdon v. Abbott, 524
U.S. 624, 645 (1998) (“When administrative and judicial
interpretations have settled the meaning of an existing
statutory provision, repetition of the same language in a new
 ELDRIDGE V. HOWARD 31

statute indicates, as a general matter, the intent to incorporate
its administrative and judicial interpretations as well.”). But
“how numerous must the lower-court opinions be, or how
prominent and long-standing the administrative interpretation,
to justify the level of lawyerly reliance that justifies the
canon?” Reading Law, at 325. While there is no universally
applicable answer, the “criterion ought to be whether the
uniform weight of authority is significant enough that [we]
can justifiably regard the point as settled law.” Id.; see also
Caleb Nelson, Statutory Interpretation 454 (2011)
(observing that the canon applies to “prominent decisions by
lower courts”).
Consider Fitzgerald v. Barnstable School Committee,
555 U.S. 246 (2009). In that case, the Supreme Court
examined whether Title IX of the Civil Rights Act allows for
parallel and concurrent 42 U.S.C. § 1983 claims. Id. at 258.
In deciding that question, the Court looked to how Title VI
of the Civil Rights Act (after which Title IX was modeled)
was understood at the time of Title IX’s passage. Id. at 258-
59. The Court found that Title VI was “routinely
interpreted” to allow for parallel § 1983 claims, citing as an
example only the decisions of the Fifth and Sixth Circuits.
Id. (citing Alvarado v. El Paso Indep. Sch. Dist., 445 F.2d
1011 (5th Cir. 1971); Nashville I–40 Steering Comm. v.
Ellington, 387 F.2d 179 (6th Cir. 1967); Bossier Par. Sch.
Bd. v. Lemon, 370 F.2d 847 (5th Cir. 1967)). The Court
“presume[d] Congress was aware of this when it passed Title
IX” and “[i]n the absence of any contrary evidence,” the
Court held that “it follows that Congress intended Title IX
to be interpreted similarly.” Id. at 259.
Here, unlike some of our sister courts, I would not rely
solely on Garris to give us a “well-settled” meaning. But
other clues reinforce carrying forward the D.C. Circuit’s
32 ELDRIDGE V. HOWARD

interpretation of § 2253 after AEDPA. Include, for example,
the significant judicial consensus that the District of
Columbia’s local courts were to be viewed as state courts.
In 1970, Congress created the “new local court system” of
trial and appellate courts of general jurisdiction in the
District of Columbia and “transferred in its entirety . . .
responsibility for processing local litigation” to the D.C.
Superior Court. Swain v. Pressley, 430 U.S. 372, 375 (1977)
(citing the District of Columbia Court Reform and Criminal
Procedure Act of 1970, 84 Stat. 475, Pub. L. No. 91-358,
Title I, § 111 (1970); D.C. Code § 11–101); see also
Theodore Voorhees, The District of Columbia Courts: A
Judicial Anomaly, 29 Cath. U. L. Rev. 917 (1980) (providing
a history of courts in the District of Columbia). Congress
then took pains to establish the D.C. Superior Court as a
“State court” for jurisdictional purposes. See 28 U.S.C.
§§ 1257, 1451(2), 2113.
The Supreme Court thus recognized that “[o]ne of the
primary purposes” of the local courts’ creation “was to
restructure the District’s court system so that the District will
have a court system comparable to those of the states.”
Pernell v. Southall Realty, 416 U.S. 363, 367 (1974)
(simplified) (emphasis added). The result was “a system of
courts analogous to those found in the States.” Swain, 430
U.S. at 375 n.4. And based on Congress’s actions, the
Supreme Court viewed judgments of the District of
Columbia Court of Appeals, the highest court in the
District’s local court system, as it would the judgments of
the “highest courts of the several States.” Pernell, 416 U.S.
at 368 (simplified).
So, along with Garris, we can presume that Congress
was aware of the judicial consensus that the D.C. Superior
 ELDRIDGE V. HOWARD 33

Court was viewed as a “State court” when it enacted
§ 2253(c)(1)(A). Even so, I think we should require more
evidence before departing from ordinary meaning. And
here, contextual evidence tilts the analysis in favor of
reading “State court” to include the D.C. courts.
B.
Statutory language “cannot be interpreted apart from
context.” Smith v. United States, 508 U.S. 223, 229 (1993).
Thus, one of the most important interpretive tools we have
is the “whole-text canon, which calls on the judicial
interpreter to consider the entire text, in view of its structure
and of the physical and logical relation of its many parts.”
Rojas, 989 F.3d at 696 n.6 (Bumatay, J., dissenting in part)
(quoting Reading Law, at 167). So when interpreting a
statutory phrase, we consider “the language itself, the
specific context in which that language is used, and the
broader context of the statute as a whole.” Robinson v. Shell
Oil Co., 519 U.S. 337, 341 (1997).
And sometimes, this broader context means recognizing
that laws are “part of an entire corpus juris,” Reading Law,
at 252, and may require analyzing the use of the same terms
across related provisions or laws. At its most useful, this
consistent-use canon “provides that identical words and
phrases within the same statute should normally be given the
same meaning.” Powerex Corp. v. Reliant Energy Servs.,
Inc., 551 U.S. 224, 232 (2007). But, as a matter of logic, the
interpretive principle may extend to identical terms across
different statutes closely related in time and subject matter—
this is sometimes called the “in pari materia” canon. See
Erlenbaugh v. United States, 409 U.S. 239, 243–44 (1972).
After all, “a legislative body generally uses a particular word
with a consistent meaning in a given context” and it’s a fair
34 ELDRIDGE V. HOWARD

assumption that “whenever Congress passes a new statute, it
acts aware of all previous statutes on the same subject.” Id.;
see, e.g., United States v. Stewart, 311 U.S. 60, 63–65 (1940)
(interpreting the phrase “income derived” similarly across
two statutes because they dealt with the same subject
matter—tax exemptions for farm loan bonds).
First, look at the use of “State” elsewhere in AEDPA.
Congress defined “State” two times in AEDPA. And both
times, Congress expressly said that a “State” includes the
“District of Columbia”:

• AEDPA § 235, Closed circuit televised
court proceedings for victims of crime:
“As used in this section, the term “State”
means any State, the District of
Columbia, or any possession or territory
of the United States.” 34 U.S.C.
§ 20142(f).
• AEDPA § 811, Federal Bureau of
Investigation: “For purposes of this
paragraph, the term ‘State’ means any
State of the United States, the District of
Columbia, the Commonwealth of Puerto
Rico, the Virgin Islands, American
Samoa, Guam, and the Commonwealth of
the Northern Mariana Islands[.]” 28
U.S.C. § 531 Statutory Notes.

Thus, when Congress felt the need to define “State” in
AEDPA, it had the District of Columbia in mind. While not
dispositive in our case, it provides further evidence of what
Congress meant by “State court” in § 2253(c)(1)(A). See,
e.g., United States v. Garcia-Paz, 282 F.3d 1212, 1214 (9th
 ELDRIDGE V. HOWARD 35

Cir. 2002) (importing the definition of “merchandise” from
Title 19 of the U.S. Code into Title 18 because the two
statutes were “part of the same act,” even though the Title 19
definition was expressly limited to “the purposes of th[at
particular] chapter”).
Second, at the time of AEDPA’s enactment, every time
Congress defined “State court” in a Title 28 jurisdictional
statute, the definition invariably included the District of
Columbia’s local court system:

• 28 U.S.C. § 1257(b): defining “highest
court of a State” to include “the District
of Columbia Court of Appeals” for the
writ of certiorari
• 28 U.S.C. § 1451(1)–(2): defining “State
court” to include “Superior Court of the
District of Columbia” and “State” to
include “the District of Columbia” for
removal jurisdiction
• 28 U.S.C. § 2113: defining “State court,”
“State courts,” and “highest court of a
State” to include “the District of
Columbia Court of Appeals” for filing the
writ of certiorari in criminal appeals

And whenever Congress sought to define a “State” for
jurisdictional questions, it included D.C.:

• 28 U.S.C. § 1332(e): defining “States” to
include “the District of Columbia” for
diversity jurisdiction
36 ELDRIDGE V. HOWARD

• 28 U.S.C. § 1367(e): defining “State” to
include “the District of Columbia” for
supplemental jurisdiction

To my knowledge, no jurisdictional provision of Title 28
excludes the D.C. Superior Court from the definition of
“State court.”
With such uniformity in jurisdictional statutes, we can
presume that Congress was “aware of all previous statutes
on the same subject,” Erlenbaugh, 409 U.S. at 244, and used
“State court” in § 2253(c)(1)(A) like it did in other
jurisdictional provisions—to include the D.C. Superior
Court. In other words, absent a reason to believe otherwise,
“State court” in § 2253(c)(1)(A) takes on a legal (rather than
a literal) meaning, and we should interpret the phrase
consistently with other jurisdictional provisions. So I
disagree with the majority’s view that silence in
§ 2253(c)(1)(A) reflects Congress’s intent to exclude D.C.
prisoners from the certificate of appealability requirement.
Instead, the better view is that Congress was legislating
within a well-known legal backdrop that included the D.C.
Superior Court within the legal meaning of “State court.”
Also, contrary to the majority’s reading, Palmore v.
United States, 411 U.S. 389 (1973), supports this reading of
“State court.” The majority argues that Palmore shows that
we can’t treat the District as a “State” for all purposes. Of
course, that’s true. Without strong textual evidence
otherwise, we should construe that term to apply only to the
50 States. So in considering whether a provision of the
District of Columbia Code counts as a “statute of any State”
under 28 U.S.C. § 1257, the Court said it does not. Palmore,
411 U.S. at 395. That’s because “nowhere in § 1257, or
elsewhere,” had Congress ever identified the D.C. Code as a
 ELDRIDGE V. HOWARD 37

“state statute.” Id. (emphasis added). But that’s not the case
when it comes to D.C. courts. Congress has identified them
as a “State court” “elsewhere”—in fact, it has done so at least
three times within Title 28. See 28 U.S.C. §§ 1257, 1451(2),
2113. So Palmore confirms that we ought to look to related
provisions of law to determine the meaning of a
jurisdictional term like “State court.”
* * *
All told, textual and contextual evidence supports the
overwhelming consensus among courts that the D.C.
Superior Court is a “State court” for purposes of the
certificate of appealability requirement. See also O’Neal
Smalls, Habeas Corpus in the District of Columbia, 24 Cath.
U. L. Rev. 75, 85 (1974) (“The history of the [District of
Columbia Court Reform and Criminal Procedure] Act, its
language, and experience with the local courts indicate that
the District of Columbia Superior Court and Court of
Appeals should be treated, at least for habeas corpus
purposes, as state courts[.]”). While the question is
concededly a close one, we should have declined to open a
circuit split.
II.
Because the D.C. Superior Court is a “State court” within
the meaning of § 2253(c)(1)(A), we should have required
Eldridge to obtain a certificate of appealability before
exercising jurisdiction over this appeal. When the district
court denies a certificate of appealability on procedural
grounds, we only issue such a certificate when a prisoner
shows both that (1) “jurists of reason would find it debatable
whether the petition states a valid claim of the denial of a
constitutional right” and (2) “jurists of reason would find it
debatable whether the district court was correct in its
38 ELDRIDGE V. HOWARD

procedural ruling.” Slack v. McDaniel, 529 U.S. 473, 484
(2000). Eldridge’s constitutional claim fails to meet the first
prong of this standard and so we should have declined a
certificate of appealability and dismissed this appeal.
Eldridge alleges that the U.S. Parole Commission
violated the Ex Post Facto Clause in its application of parole
guidelines. In 1984, Eldridge was convicted of several
criminal offenses, including rape, armed robbery, and
burglary. The D.C. Superior Court sentenced him to a term
of 40 to 120 years of imprisonment. After he became
eligible for parole and was initially denied, the Commission
again denied him parole on rehearings in 2010, 2013, 2016,
2018, and 2019. At the time of Eldridge’s conviction, 1972
parole guidelines governed. Under those guidelines,
prisoners denied parole after serving sentences of five years
or longer were eligible for a parole rehearing “ordinarily”
after a one-year set-off period, but the parole board had
discretion to “establish a rehearing date at any time it fe[lt]
such would be proper.” 9 D.C.R.R. § 103 (1972); see also
Daniel v. Smoot, 316 F. Supp. 3d 79, 82 (D.D.C. 2018). In
2000, the guidelines changed. The 2000 guidelines
lengthened the “presumptive” set-off period to three years
while permitting the Commission to shorten that period at its
discretion. Smoot, 316 F. Supp. 3d at 83 (quoting 28 C.F.R.
§ 2.75(a)(1)(iv), (e)).
In his petition, Eldridge argues that the Commission
improperly applied the longer 2000 guidelines “set-off”
period to his 2010, 2013, 2016, and 2019 parole rehearings,
rather than the shorter 1972 guidelines period. The Ex Post
Facto Clause forbids laws “that change[] the punishment,
and inflict[] a greater punishment, than the law annexed to
the crime, when committed.” Peugh v. United States, 569
U.S. 530, 532–33 (2013) (simplified); see U.S. Const. art. I,
 ELDRIDGE V. HOWARD 39

§ 9, cl. 3. A denial of parole can violate the Ex Post Facto
Clause when the retroactive use of revised guidelines creates
“a sufficient risk of increasing” the duration of a prisoner’s
incarceration. Garner v. Jones, 529 U.S. 244, 250 (2000)
(simplified).
Eldridge is right that the Commission used the 2000
guidelines in scheduling his 2010 and 2013 rehearings. But
even so, Eldridge can’t show an ex post facto violation.
Because the Commission continued to deny Eldridge’s
parole after his 2013 rehearing, he would not have been
released any earlier under the proper one-year set-off period.
Thus, the erroneous use of the 2000 guidelines in 2010 and
2013 did not result in any increase in Eldridge’s
incarceration and his constitutional challenge to those parole
denials fails. See Garner, 529 U.S. at 250.
Elridge’s claims related to his 2016 and 2019 rehearings
fare no better. The Commission plainly used the appropriate
1972 guidelines for those rehearings. While the
Commission imposed three-year set-off periods for those
rehearings, it did so under the 1972 guidelines’ discretionary
authority to extend the “ordinary” one-year period. See 9
D.C.R.R. § 103 (1972). Take the 2016 rehearing. The
Commission expressly said that the “ordinary” one-year set-
off period was improper because Eldridge needed to
complete drug and sex offender treatment programs. The
same occurred in 2019. Then, the Commission
acknowledged that it was “exceeding the normal rehearing
schedule” because Eldridge was “an untreated sexual
predator.” Thus, it’s not debatable that the Commission
applied an ex post facto law; it did not.
Even if the Commission did, Eldridge still can’t succeed
because he only challenges the extended set-off periods—
40 ELDRIDGE V. HOWARD

not that the Commission should have released him under the
1972 guidelines. In California Department of Corrections
v. Morales, 514 U.S. 499 (1995), the Court considered the
retroactive application of a California law that allowed the
state parole board to hold hearings every three years rather
than the annual rehearing required by law at the time of the
offender’s crime. See id. at 503. Because the amended law
did not change the substantive standards for parole eligibility
and only lengthened the set-off period, the Court held the
amendment “simply alters the method to be followed in
fixing a parole release date under identical substantive
standards.” Id. at 508 (simplified). The new law only
relieved the parole board from conducting time-consuming
annual parole hearings for offenders “who have no
reasonable chance of being released.” Id. at 507. In this
situation, the retroactive application of the law would only
result in “the most speculative and attenuated possibility of
producing the prohibited effect of increasing the measure of
punishment” for the defendant. Id. at 509. Because there
was no basis to conclude that the longer set-offs under the
new law “would extend any prisoner’s actual period of
confinement,” they didn’t alter the “quantum of
punishment,” and so did not create an ex post facto violation.
Id. at 513 (simplified). Thus, Eldridge’s constitutional
claims must fail.
For these reasons, Eldridge is not entitled to a certificate
of appealability.
III.
While not an easy decision, the weight of interpretive
guidance counsels against creating a circuit split here. I
would instead follow our sister circuits and read
§ 2253(c)(1) harmoniously with prior court decisions, the
 ELDRIDGE V. HOWARD 41

rest of AEDPA, and related jurisdictional provisions. We
thus should have interpreted “State court” to include the
D.C. Superior Court. Under this reading, Eldridge needed a
certificate of appealability, which is unwarranted here. We
should have dismissed this case.
I respectfully dissent from the judgment of the court.