NOT RECOMMENDED FOR PUBLICATION File Name: 22a0021n.06
Case Nos. 20-3472/3496
UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT
FILED ) Jan 10, 2022 CHRISTOPHER SMITH, ) DEBORAH S. HUNT, Clerk Petitioner-Appellee/Cross-Appellant, ) ) v. ) ON APPEAL FROM THE UNITED ) STATES DISTRICT COURT FOR WARDEN, TOLEDO CORRECTIONAL ) THE SOUTHERN DISTRICT OF INSTITUTION, ) OHIO Respondent-Appellant/Cross-Appellee. ) )
Before: SUTTON, Chief Judge; SILER and READLER, Circuit Judges.
SUTTON, Chief Judge. The State of Ohio convicted Christopher Smith of robbing a store.
He sought a new trial after learning that the State withheld DNA testing notes (but not the DNA
test results) from him, arguing that the lack of disclosure violated his due process rights. See Brady
v. Maryland, 373 U.S. 83 (1963). The state trial court denied his motion, and he did not appeal
that ruling. The Ohio Court of Appeals later determined that the decision not to appeal did not rise
to the level of ineffective assistance of counsel because the due process argument failed on the
merits anyway. Because this Ohio Court of Appeals decision did not contradict or unreasonably
apply Supreme Court precedent, we must reverse the district court’s grant of a writ of habeas
corpus. Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
I.
A.
In 2007, a man entered a Cincinnati Bell store wearing a facemask, a wig, and sunglasses.
He took out a gun, ordered the customers to the floor, and left with the cash drawer. Watching
from his car, Thomas Moore saw the whole thing, including the thief’s escape in the passenger
seat of a blue Ford Expedition. Moore, as it happens, had training as a military policeman. He
called 911, followed the Expedition, relayed a partial license plate number, and got a look at the
passenger from ten feet away. The police tracked the license plate to Jennifer Potts, who told the
police she bought the Expedition for Christopher Smith. Officers found the Expedition parked
near Smith’s residence, with a wig, sunglasses, and black t-shirt nearby.
In the hours after the robbery, Moore picked Smith out of a photo array as the thief. Police
checked Smith’s ankle monitor records, which he wore because he was on parole for armed
robbery. The records revealed the monitor was inactive between 2:35 and 4:20 pm on the day of
the robbery (October 17) and at the time of the robbery (3:40 pm). After a news segment on the
robbery aired later that evening, Smith cut off his ankle bracelet and disappeared on the run. He
texted Potts, telling her to “report the truck stolen,” and wrote someone else that he was “going to
get out of this.” R.12-3 at 106; R.12-4 at 82. The police found Charles Allen’s fingerprints in the
back seat of the Expedition and interviewed him. He told the police he had been in the car with
Smith several times.
A month later, police officers happened to stop a car in which Smith was a passenger. In
response, he fled. The police caught him and arrested him. During an interview after the arrest,
Smith revealed that he knew the street where the Expedition had been abandoned and claimed he
had loaned the car to “Chuck.” R.12-4 at 90–91. He also asserted that Charles Allen texted him
2 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
after the robbery to tell him what happened except that his phone records did not reveal any such
text. The officers showed him photos of Charles Allen, and Smith claimed that he did not know
him. Smith insisted that officers run DNA and polygraph tests and check his phone records to
prove his innocence.
A grand jury indicted Smith for aggravated robbery, robbery, and firearm possession, and
Smith opted for a bench trial. At trial, the defense argued that Allen—identified by Moore as the
getaway driver—borrowed the truck and committed the robbery without any involvement by
Smith at any point. The defense presented evidence that Smith’s ankle monitor regularly
malfunctioned. A state expert described the results of the DNA test—that it showed DNA on the
wig and t-shirt and that Smith’s DNA profile did not appear on the wig or t-shirt. The expert found
Charles Allen’s DNA profile on both items, prompting Smith to argue at trial that the DNA test
results showed that Allen was the sole culprit.
Allen testified at the trial after the prosecution promised not to indict him in return for his
truthful testimony. He testified that Smith had picked him up in the Expedition, committed the
robbery, and forced Allen to drive the getaway car. Allen testified that his girlfriend had several
wigs, that Smith had asked him whether she had one “laying around,” and that one went missing
around the time of the robbery. R.12-8 at 62–63. The independent testimony from the eyewitness,
Moore, supported Allen’s testimony.
After the bench trial, the trial judge convicted Smith on all counts.
B.
After the trial, the defense obtained the lab notes from the DNA testing. The notes revealed
Allen’s DNA at every place swabbed on the t-shirt and almost every place on the wig, providing
more evidence that Allen wore the wig and t-shirt found with the abandoned Expedition. Smith
3 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
moved for a new trial, alleging that the State’s failure to turn over the notes violated his rights
under Brady. Smith’s expert testified that the notes, together with the underlying DNA tests, were
consistent with Allen wearing the items for an extended period rather than simply touching them.
Throughout, she cautioned her conclusions were “not definitive” and did not preclude the
possibility that Smith wore the wig. R.12-11 at 33, 37–38.
The trial court denied the motion for a new trial. The judge concluded that the evidence
was not material and that no Brady violation occurred. He explained that Smith “might have an
argument if you’re trying it to a jury, but you’re trying the case to me, and I’m also deciding the
motion for a new trial.” Id. at 58. He had “no doubt that both these guys were involved in this.”
Id. at 58–59. And he pointed to all of the evidence that Smith participated in the robbery and
reasoned that Allen might have “wor[n] this outfit and had [the items] on numerous occasions on
other robberies, or even that day . . . and that would mask Chris Smith’s DNA.” Id. at 58–61.
C.
Smith appealed his conviction. In doing so, he did not raise a Brady claim. The Ohio
Court of Appeals affirmed Smith’s conviction.
Smith filed an application to reopen his appeal due to ineffective assistance of appellate
counsel, including the failure to raise the Brady claim. Ohio App. R. 26(B). The Ohio First District
Court of Appeals denied the motion. “Nondisclosure of the lab notes,” it concluded, “did not
violate” Brady “because the undisclosed evidence was not ‘material’ in that it could not
‘reasonably be taken to put the whole case in such a different light as to undermine confidence in
the verdict[].’” R.75 at 201 (quoting Kyles v. Whitley, 514 U.S. 419, 435 (1995)). The “assignment
of error,” it reasoned, “would not have presented a reasonable probability of success if it had been
4 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
advanced on appeal,” meaning that Smith “failed to demonstrate a genuine issue” as to whether he
received ineffective assistance. Id. at 202.
Smith sought federal habeas relief, raising a number of claims, Brady included. The district
court dismissed the petition. But it granted a certificate of appealability to this court on its
conclusion that several claims were procedurally defaulted, including the Brady claim, and its
merits determination on a freestanding ineffective assistance claim. A panel of this court ruled
that the district court was correct in ruling the claims procedurally defaulted. Smith v. Warden,
780 F. App’x 208, 231 (6th Cir. 2019). But it determined that ineffective assistance of appellate
counsel excused the default of the Brady claim, meaning the district court should adjudicate it on
the merits. Id. We remanded for the district court to consider the Brady claim and to assess the
freestanding ineffective assistance of appellate counsel claim. Id. at 230–31.
On remand, the district court granted an unconditional writ of habeas corpus on the Brady
claim, ordered Smith’s immediate release due to the COVID-19 pandemic, and rejected Smith’s
claim that he could not be retried. The warden appeals the district court’s Brady ruling, and Smith
appeals the district court’s denial of his motion to bar retrial.
II.
The Antiterrorism and Effective Death Penalty Act, commonly known as AEDPA, sets the
stage. It permits us to overturn a state court’s decision only if the state court’s decision “was
contrary to, or involved an unreasonable application of, clearly established Federal law, as
determined by the Supreme Court of the United States.” 28 U.S.C. § 2254(d)(1). In applying
these rules, we evaluate the “last state-court adjudication on the merits.” Greene v. Fisher, 565
U.S. 34, 40 (2011). To obtain relief under Brady, Smith must show that the lab notes were
(1) suppressed, (2) favorable, and (3) material. See Strickler v. Greene, 527 U.S. 263, 281–82
5 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
(1999). All of this means Smith may obtain relief only if the relevant state court determination
(1) was “contrary to” clearly established federal law or (2) was “an unreasonable application” of
clearly established federal law.
What was the “last state-court adjudication on the merits” of the Brady claim? The Ohio
Court of Appeals’ denial of Smith’s Rule 26(B) motion, in which he argued that his appellate
counsel failed to provide effective assistance when his attorney did not appeal the trial court’s
rejection of his Brady claim. The Court of Appeals’ “legal reasoning” expressly “address[ed]” the
underlying claim in finding no prejudice and made clear that its decision turned on its view of the
merits of the Brady claim. Fleming v. Metrish, 556 F.3d 520, 531–32 (6th Cir. 2009); see James
v. Brigano, 470 F.3d 636, 641–42 (6th Cir. 2006). The failure to disclose the notes, the state court
reasoned, did not violate Brady because “the undisclosed evidence was not ‘material’ in that it
could not ‘reasonably be taken to put the whole case in such a different light as to undermine
confidence in the verdict[].’” R.75 at 201 (quoting Kyles, 514 U.S. at 435). The court’s
consideration was not hypothetical—say that, “if the merits were reached, the result would be the
same.” Bell v. Miller, 500 F.3d 149, 155 (2d Cir. 2007). It was not an “offhand remark” expressing
“reservations” about the merits. Barton v. Warden, 786 F.3d 450, 461 (6th Cir. 2015) (per curiam).
The court ruled in straightforward terms that the Brady claim failed on the merits, precluding the
possibility that Smith’s counsel committed ineffective assistance by opting not to raise it. Once
the Ohio Court of Appeals addressed the Brady claim, “we would be acting contrary to Congress’s
intent . . . if we simply ignored” the court’s view and reconsidered it on our own fresh terms.
Fleming, 556 F.3d at 532.
6 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
Smith objects to this conclusion on the ground that the appellate court had in front of it an
ineffective-assistance claim, not a Brady claim. True enough. But the court did what appellate
courts frequently do. It resolved the underlying claim first. Put another way it resolved a necessary
premise of the ineffective-assistance claim. A lawyer does not provide ineffective assistance of
counsel by opting not to bring a losing claim. Under time-tested principles, a Sixth Amendment
claim of ineffective assistance of counsel requires the criminal defendant to show
(1) representation outside professional norms and (2) prejudice from failing to raise the claim.
Strickland v. Washington, 466 U.S. 668, 687 (1984). No prejudice results from failing to bring a
defective claim. See, e.g., United States v. Mahbub, 818 F.3d 213, 231 (6th Cir. 2016); United
States v. Holder, 657 F.3d 322, 332 (6th Cir. 2011); United States v. Jones, 489 F.3d 243, 255 (6th
Cir. 2007).
Nothing requires a state court to conduct its analysis under the heading of a specific federal
constitutional right to adjudicate it on the merits. See Bennett v. Brewer, 940 F.3d 279, 290–91
(6th Cir. 2019) (holding that a ruling that an ineffective assistance of trial counsel claim failed due
to lack of prejudice also served as a merits decision for an ineffective assistance of appellate
counsel claim on the same underlying error). When the substance of a right that the court addresses
is “generally coextensive” with another right under the Federal Constitution, that discussion is
“sufficient” to be a merits decision of the “related federal right.” Johnson v. Williams, 568 U.S.
289, 299 (2013). The test for materiality under Brady parallels the test for prejudice under
Strickland. Montgomery v. Bobby, 654 F.3d 668, 679 n.4 (6th Cir. 2011). Because the basis for
the state court’s decision “subsumes the [Brady] standard,” Johnson, 568 U.S. at 301, a ruling on
the integrated issue was “entirely dispositive” of the Brady claim and “necessarily resolved” it on
the merits, Jackson v. Smith, 745 F.3d 206, 210 (6th Cir. 2014) (emphasis omitted).
7 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
Other courts have rejected arguments like Smith’s. The Seventh Circuit held that when
“the due-process and ineffective-assistance-of-counsel standards effectively present the same
question,” a decision on the one means the “merits were effectively reached” on the other.
Sturgeon v. Chandler, 552 F.3d 604, 612 (7th Cir. 2009) (failure to conduct competency hearing);
see also Flint v. Carr, 10 F.4th 786, 796–97 (7th Cir. 2021) (double-jeopardy); Diez v. Sec’y, Fla.
Dep’t of Corr., 789 F. App’x 837, 838–39 (11th Cir. 2020) (per curiam) (Brady).
Our court has long done something similar in an analogous situation. Sometimes a state
court will address the merits of the underlying claim in conducting plain-error review. So long as
the state court “address[ed] whether an error had occurred,” we have held, that analysis is an
adjudication on the merits of the underlying claim for AEDPA deference purposes. Fleming, 556
F.3d at 532; Stewart v. Trierweiler, 867 F.3d 633, 638 (6th Cir. 2017); see Porter v. Eppinger, No.
19-3443, 2021 WL 3828113, at *4–5 (6th Cir. Aug. 27, 2021) (holding that an ineffective-
assistance claim failed because the underlying merits claim was “meritless”). We likewise have
held that Michigan state court decisions denying leave to appeal can be adjudications on the merits
when the state courts “specified” that they denied the application “for reasons involving the
substance.” Werth v. Bell, 692 F.3d 486, 494 (6th Cir. 2012).
Smith insists that our case law establishes that the ineffectiveness inquiry and the merits of
the underlying claim are distinct. But the cited cases determined only that the underlying claims
were procedurally defaulted because an independent and adequate state ground prevented their
consideration in state court and that adding the claim to a 26(B) application by itself did not suffice
to preserve the claim. Davie v. Mitchell, 547 F.3d 297, 311–13 (6th Cir. 2008); see also
Wogenstahl v. Mitchell, 668 F.3d 307, 338 (6th Cir. 2012). Whether a state court decision may
“resurrect an otherwise defaulted claim” is a distinct question from “whether AEDPA deference
8 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
applies.” Stewart, 867 F.3d at 638; see also Fleming, 556 F.3d at 531 (“[T]he question of whether
a claim should be addressed on collateral review under the judicially created doctrine of procedural
default is independent of the question of whether Congress requires deference pursuant to
AEDPA.”).
If, in the course of rejecting a claim on state procedural grounds, a state court “conducts
any reasoned elaboration of an issue under federal law” and “addresses” the merits, we owe
deference to its determination of the federal issue under AEDPA. Fleming, 556 F.3d at 531–32;
see Brooks v. Bagley, 513 F.3d 618, 624–25 (6th Cir. 2008) (explaining AEDPA requires deference
to a state court’s alternative holding on the merits even if the state court first holds that a claim is
procedurally defaulted). Likewise, when a state court addresses the substance of the underlying
claim in reviewing a 26(B) application, as the Ohio Court of Appeals did here, we have not
hesitated to consider it a merits determination for AEDPA purposes. See James, 470 F.3d at 641–
44; Patterson v. Haskins, 316 F.3d 596, 605, 608 (6th Cir. 2003); Winston v. Brunsman, 566 F.
App’x 422, 423–25 (6th Cir. 2014) (per curiam).
Our prior panel decision does not require a different conclusion. It did not apply AEDPA;
indeed it had no occasion to apply the federal statute. It thus did not assess the claim through the
lens of the last state court to reach it, the state court of appeals. That explains why it remanded the
case to the district court to resolve the claim in the first instance.
Smith’s reliance on the mechanics of a Rule 26(B) motion under Ohio law does not serve
him any better. Rule 26(B) requires an application to reopen the defendant’s direct appeal to state
“assignments of error that previously were not considered on the merits in the case . . . because of
appellate counsel’s deficient representation.” The application is granted only “if there is a genuine
issue as to whether the applicant was deprived of the effective assistance of counsel on appeal.”
9 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
Ohio App. R. 26(B)(5). This application is a “first stage” that requires the defendant to make a
“threshold showing” before he has “permission to file new appellate briefs.” Ohio App. R. 26 staff
note to 1993 amendment. If the applicant clears this bar, the “second stage” turns on the “merits
of the appeal” and whether it satisfies the ineffective assistance standard. Id. Smith argues that
this procedure prevents the merits from being considered “until the application is granted and the
appeal is officially reopened, at which point the case proceeds to full briefing and a merits review.”
Appellee’s Br. at 49 (emphasis omitted).
But this procedure does not prevent the Ohio appellate courts from denying a claim on the
merits at the first stage. A merits determination can come “after the court . . . heard and evaluated
the evidence and the parties’ substantive arguments.” Johnson, 568 U.S. at 302 (alteration in
original) (quotation omitted). Rule 26(B) allows “for a meaningful review of the record” even at
the first stage by requiring that applications include “portions of the record” and a statement of the
basis of the claim. State v. Davis, 894 N.E.2d 1221, 1224–25 (Ohio 2008). The appellate court
can “even conduct an evidentiary hearing.” Id. at 1225. “The clear intent of [Rule] 26(B) is for
the appellate court to function as the trier of fact in determining whether the defendant has
demonstrated a genuine issue as to the ineffectiveness of his appellate counsel.” Id.; see also
Hudson v. Bradley, 764 F. App’x 494, 496–97 (6th Cir. 2019) (concluding that the application
stage of the 26(B) process is not “so limited that it is ‘not worthy of deference under AEDPA’”).
Granting a 26(B) motion grounded in an underlying Brady claim and moving to the second stage
is not a conclusion that the Brady claim succeeds on the merits. But denying such a motion on
lack-of-prejudice grounds, where the standard mirrors a necessary element of a Brady claim, is
necessarily a conclusion that the Brady claim fails.
10 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
Smith adds that the Ohio Court of Appeals’ decision was not on the merits because its
analysis was conclusory. But there is no need for there even to be “an opinion from the state court
explaining the state court’s reasoning.” Harrington v. Richter, 562 U.S. 86, 98 (2011). The state
court made clear it was denying the 26(B) motion, made clear that it was doing so on the prejudice
prong of Strickland, and made clear that this followed from its conclusion that the Brady claim
failed. That was a merits determination by any measure.
To obtain relief under Brady, Smith must show that the lab notes were (1) suppressed,
(2) favorable, and (3) material. See Strickler, 527 U.S. at 281–82. To be material, the key question
at hand, the lab notes must create a “reasonable probability,” not a reasonable possibility, of a
different outcome. Kyles, 514 U.S. at 434 (quotation omitted).
Because the Ohio Court of Appeals addressed Smith’s Brady claim on the merits, we owe
its decision deference under AEDPA. 28 U.S.C. § 2254(d). Resolution of the case thus turns on
whether the Ohio Court of Appeals unreasonably applied the U.S. Supreme Court’s Brady case
law in determining that the undisclosed evidence was immaterial to the outcome of Smith’s trial.
See Williams v. Taylor, 529 U.S. 362, 409 (2000). To grant relief under § 2254, the state court’s
ruling must have been “so lacking in justification that there was an error well understood and
comprehended in existing law beyond any possibility for fairminded disagreement.” Harrington,
562 U.S. at 103. The more general the standard, the more “latitude” a state court has. Knowles v.
Mirzayance, 556 U.S. 111, 123 (2009). A “reasonable probability” standard operates at a high
level of generality. See Cyars v. Hofbauer, 383 F.3d 485, 493 (6th Cir. 2004).
The Ohio Court of Appeals did not unreasonably apply clearly established federal law as
determined by the U.S. Supreme Court. Considerable evidence linked Smith to the crime. See
11 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
Strickler, 527 U.S. at 293–94. A credible eyewitness (Moore) observed the robbery, which utilized
Smith’s car; he got a good look at the perpetrators, at one point just ten feet away; and he picked
out Smith from a photo array hours later. Police found the car near Smith’s residence. Smith’s
ankle monitor was inactive on the precise day and at the precise time of the robbery. Smith’s
actions after the robbery—cutting off his bracelet and running, sending inculpatory texts, and lying
to the police about receiving a text from Allen—all point toward guilt. Allen himself testified that
he had been there during the robbery and he watched Smith execute it. On top of all that, the trier
of fact had in front of him DNA test results consistent with the notes, namely that Smith’s DNA
did not appear to be on the wig and Allen’s DNA appeared to be on the wig. The State’s case was
not, in short, “a house of cards built on” a claim about the nature of the DNA evidence. Wearry v.
Cain, 577 U.S. 385, 392 (2016) (per curiam); see also McNeill v. Bagley, 10 F.4th 588, 603–04
(6th Cir. 2021).
The DNA notes, it is true, added some support to the defense’s theory: that Allen alone
wore the wig and alone committed the robbery. But keep in mind that this theory was already in
front of the fact finder based on the DNA results and expert testimony about those test results. The
DNA lab notes are hardly the kind of jarring new testimony that characterizes some Brady claims.
Instead of contradicting the evidence at trial, they simply elaborate upon information the State
already provided. Just like the DNA test results, the DNA lab notes could not pin down when
Allen might have worn the wig, and they could not pin down that Smith never wore it. Neither
the DNA test results nor the lab notes, moreover, “directly contradict” Moore’s eyewitness
testimony or the other evidence inculpatory of Smith. Smith v. Cain, 565 U.S. 73, 76 (2012);
McNeill, 10 F.4th at 603–04. In prior cases examining materiality, we have deemed Brady
12 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
evidence that impeaches one witness immaterial in the face of another credible eyewitness. France
v. Lucas, 836 F.3d 612, 630–31 (6th Cir. 2016); see also McNeill, 10 F.4th at 601–04.
Our earlier decision does not require a different approach. On fresh review, we held that
there “was a reasonable probability of a different outcome had the lab notes been disclosed.”
Smith, 780 F. App’x at 228. But today’s case presents a different lens through which to gauge the
claim—AEDPA’s lens that requires us to defer to reasonable state court rulings. The upshot of
AEDPA is that reasonable minds can differ about the answers to constitutional questions. In this
instance, our prior decision looks one reasonable way, and the state appellate court’s decision looks
another reasonable way. “[E]ven a strong case for relief does not mean the state court’s contrary
conclusion was unreasonable.” Harrington, 562 U.S. at 102. “Viewed alone,” the DNA evidence
supplemented by the lab notes “could cast doubt on the State’s theory,” but when considered
simultaneously with “the evidence supporting the State’s theory,” “the Ohio court was clearly not
unreasonable in its determination.” Montgomery, 654 F.3d at 680.
The district court did not consider the freestanding ineffective assistance claim after it
granted a writ of habeas corpus on the Brady claim. Both parties request a remand to the district
court to consider the ineffective assistance claim in the first instance. As a “court of review, not
first view,” we remand to the district court to consider the claim in the first instance. United States
v. Houston, 792 F.3d 663, 669 (6th Cir. 2015).
III.
After the State of Ohio failed to comply with the writ of habeas corpus, Smith moved under
Federal Civil Rule 60(b)(6) for an order amending the already issued writ of habeas corpus to
13 Case Nos. 20-3472/3496, Smith v. Warden, Toledo Corr. Inst.
prohibit the State from retrying him. Because we reverse the district court’s grant of the writ, the
appeal of the denial of the Rule 60(b) motion is dismissed as moot.
For these reasons, we reverse the grant of the writ of habeas corpus, remand the ineffective
assistance claim to the district court, and dismiss as moot the appeal from the motion to bar retrial.