Jimenez v. StanfordJimenez v. Stanford
Before: CABRANES, LYNCH, and LOHIER, Circuit Judges.
District Court for the Southern District of New York (Nathan, J.) denying his petition for a writ of habeas corpus. A New York jury found Jimenez guilty of murder after a trial in which the State‘s case-in-chief relied on the testimony of two eyewitnesses. But two decades later, one eyewitness recanted and claimed that a police detective improperly influenced his identification. Two alibi witnesses also came forward. A State court denied post-conviction relief, and Jimenez petitioned for a writ of habeas corpus, claiming actual innocence and Brady violations. The district court found that Jimenez had cast enough doubt on his guilt to excuse his untimely petition, but ultimately denied relief on the merits.
Jimenez argues on appeal that the district court erroneously deferred to the State court‘s conclusions, held his actual innocence claim to an impossibly high standard, and contravened factual findings made following an evidentiary hearing. His arguments rely on two unsettled legal premises: that a freestanding claim of actual innocence is constitutionally cognizable and may be established by clear and convincing evidence, and that the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA“) does not apply to claims of actual innocence. We assume without deciding that freestanding innocence claims are cognizable. We conclude, however, that AEDPA applies to such claims, and that the district court could not have granted habeas relief under the facts presented in this case because Jimenez‘s newly discovered evidence does not satisfy the substantially higher standard of proof required to prove actual innocence. We further conclude that there is no merit to Jimenez‘s Brady claim. We therefore AFFIRM the district court‘s judgment denying habeas relief.
Judge Cabranes concurs in the judgment in a separate opinion.
GLENN A. GARBER (Rebecca E. Freedman, on the brief), The Exoneration Initiative, New York, NY, for Petitioner-Appellant.
MATTHEW B. WHITE, Assistant District Attorney (David M. Cohn, Assistant District Attorney, on the brief), for Darcel D. Clark, District Attorney, Bronx County, Bronx, NY, for Respondent-Appellee.
Parvin D. Moyne and Andrew A. McWhorter, Akin Gump Strauss Hauer & Feld LLP, New York, NY; Juliana C. DeVries and Zara H. Shore, Akin Gump Strauss Hauer & Feld LLP, Washington, DC, for Amici Curiae Centurion Ministries, The National Association of Criminal Defense Lawyers, The New York State Association of Criminal Defense Lawyers, The Center for Appellate Litigation, and The Legal Aid Society, in support of Petitioner-Appellant.
GERARD E. LYNCH, Circuit Judge:
This is a troubling case about a weakly supported thirty-year-old murder conviction that may have condemned an innocent teenager to decades in prison. It is troubling because, despite our considerable doubt regarding the petitioner‘s guilt, we are bound to conclude that he is not entitled to a writ of habeas corpus based only on the contention that he is, in fact, innocent.
Petitioner-Appellant Rafael Jimenez appeals from a judgment entered on September 15, 2021, in the United States District Court for the Southern District of New York (Alison J. Nathan, J.), denying a writ of habeas corpus pursuant to
But when the district court reached the merits of Jimenez‘s claims, it denied his petition for a writ of habeas corpus. Jimenez v. Stanford (”Jimenez III“), 560 F. Supp. 3d 761, 764 (S.D.N.Y. 2021). It chiefly ruled that, although Jimenez‘s new exculpatory evidence was compelling enough to meet a high standard of proof for the purpose of excusing a procedural default, the same body of evidence did not meet the even higher standard required for freestanding claims of actual innocence and to overcome the deference to State court decisions mandated by the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA“),
Jimenez raises numerous objections on appeal. Most are addressed to the extraordinarily high hurdles set before freestanding innocence claims, including the central problem that to date no federal court has recognized the existence of such a claim as a basis for granting the writ. He also argues that the deference owed to State court decisions on the merits of habeas claims under AEDPA,
We disagree. Although we assume arguendo that freestanding claims of actual innocence are cognizable, and additionally assume that Jimenez can overcome certain other contested threshold barriers, we hold that freestanding actual innocence claims, even if cognizable, are evaluated under AEDPA‘s deferential standard, and that the standard of proof associated with such claims exceeds the proof of innocence required to excuse a belated petition. Accordingly, our analysis is constrained by AEDPA and the high standard applicable to freestanding innocence claims.
Under these legal constraints and on the facts of this case, the district court properly found that it could not grant Jimenez habeas relief. The evidence incriminating Jimenez is not compelling. Yet the newly surfaced exculpatory evidence, while troubling, is also subject to question. Even accounting for the expanded record developed in the district court, Jimenez – despite casting serious doubt on his guilt – has not carried the burden of proving his actual innocence under the standards set by AEDPA and the Supreme Court. Finally, like the district court, we find no merit in Jimenez‘s sole remaining claim for relief based on an alleged Brady violation at his trial.
We do not arrive lightly at these conclusions. We doubt that a reasonable jury would convict Jimenez in a new trial if presented with all of the evidence tending to prove his innocence. But doubt alone –
BACKGROUND
When assessing a claim of innocence for the purpose of considering whether to overcome a procedural bar to a habeas petition, courts are obliged to “consider all [record] evidence, old and new, incriminating and exculpatory, without regard to whether it would necessarily be admitted under rules of admissibility that would govern at trial.” House v. Bell, 547 U.S. 518, 538 (2006) (internal quotation marks omitted), quoting Schlup v. Delo, 513 U.S. 298, 327-28 (1995). In addressing the merits of a habeas claim, the Supreme Court has held that review under AEDPA,
differing evidentiary materials derived from Jimenez‘s original trial, the State court post-trial record, and the record of the “gateway” federal court hearing.2
I. The Murder of Michael Brana
Michael Brana was shot outside his apartment building on June 25, 1992, in front of his wife, Carmen Velazquez,3 and his friend and neighbor, Harry Ramos. Brana was home at the time only because Velazquez had told him by telephone that a group of men had harassed her earlier that day, including one man who spat at her, when she left a neighborhood bodega with her minor stepdaughter. Brana left work early, arrived home around 3:00 p.m., and then stood outside the building with Ramos, who urged him not to fight the harassers because they were dangerous. Velazquez joined them some time between 4:00 or 4:30 p.m.
Five men approached the trio at around 5:00 p.m. Accounts differ as to what occurred next. At trial, Velazquez testified that a short man stepped forward and taunted Brana in Spanish, and that she knew from his accent that he
was from the Dominican Republic. She said the short man asked “if the guy behind him was the guy that [Brana] was looking for,” and after Brana responded that he was “not looking for anybody,” the short man stepped aside and the man standing behind him fired three shots, killing Brana. App‘x 251-52. Velazquez was about eleven feet away and saw the shooter‘s face for about thirty seconds. She then fled indoors and called the police. She never heard the shooter speak.
Ramos testified at trial to a slightly different sequence of events: that almost
men then fled.
Importantly, Ramos testified at trial in Spanish through an interpreter, so the jury heard only the English translation of what the shooter said to Brana. When Ramos recanted almost two decades later, he attested that the shooter used a particular Spanish word translated as “cocksucker” in English: the exclusively Dominican slur “mamaguevo,” which Ramos insists would not be used by Spanish-speakers of any other national origin. The significance of the shooter‘s word choice, especially as a means of identifying him or determining his ethnicity, was never an issue at trial.
II. The Arrest of Rafael Jimenez
A. Eyewitness Descriptions
About a half hour after the shooting, New York police detective Floyd Coor interviewed Velazquez in her apartment, in the presence of another detective, Kenneth Thompson. The narrative portion of the police report, signed by Detective Coor, recorded that she described the man who had harassed her earlier that day as a five-foot, six- or seven-inch tall Dominican man, roughly 23 or 24 years old, with long black hair in a Jheri curls hairstyle,6 sporting a mustache, wearing a white tank top and denim shorts, and driving a red sports car. In the preset fields for entering a suspect‘s description at the top of the form, however, the report describes the shooter slightly differently: a Hispanic male, aged 25 or 26, about five feet, nine inches in height, with a light mustache and goatee, and wearing a white tank top and khaki shorts.7
The report does not clearly indicate whether the shooter and the harasser were different men, or whether these were inconsistent descriptions of the same person. As explained below, at trial, Velazquez insisted that she never saw the man who harassed her and that the only physical description she gave to Detective Coor was of the shooter. Detectives Coor and Thompson, however, testified that she described two different people – the harasser and the shooter.
B. Photographic Arrays
After the on-site interview, Velazquez went to a police precinct with Detective
Accounts of what happened next again differ. Detective Thompson‘s contemporaneous police report recorded that, after the interview, Ramos went through a photographic array containing several hundred pictures and “picked out a photo of a subject known to this dept. as RALF JIMENEZ.”8 Id. at 42. At the
gateway innocence hearing held over twenty years later, however, Ramos testified that Detective Thompson showed him only two or three photographs, and that he had refused to select one because Monaguillo was not among them, even though police officers urged him to select a particular photograph and told him that Brana‘s wife had already selected it. Detective Thompson, testifying at the same hearing, stated that he had no recollection of that day, but insisted that he would never show a witness a limited selection of photographs or suggest that the suspect was among them.
In any event, it is undisputed that on that same day, Detective Thompson obtained an updated photograph of Jimenez,9 included it in a photographic array with five other subjects, and showed the array to Velazquez, who positively identified Jimenez. At a pretrial hearing, Detective Thompson testified that Velazquez identified Jimenez “almost immediately.” Id. at 128.
C. Lineup
Detective Thompson could not immediately locate Jimenez at his known locations, so in September 1992, he prepared and circulated Jimenez‘s March 1992 photograph on a “Wanted” poster. The poster described Jimenez as a Hispanic male, aged seventeen, five-feet, eight-inches tall, with a medium complexion and short, black, curly hair.
Jimenez, however, was never very far from where Detective Thompson had been searching. He had been serving a probationary sentence for a prior drug offense and semi-regularly reported to his probation officer between March and November 1992 at a Bronx location close to both his own residence and the scene of the shooting. His probation officer testified at trial that Jimenez came to appointments on
It was not until October 14, 1992, that Detective Thompson arrested Jimenez several blocks from the crime scene. The next day, Jimenez was placed in the third position of a six-man lineup at a police precinct. Ramos arrived with his father and viewed the lineup at 7:12 p.m., but – according to Detective
Thompson‘s police report – he was “unable to identify” Jimenez. Id. at 52. Detective Thompson testified at trial that Ramos “refused to make an identification.” Id. at 381. At a pretrial hearing, Detective Thompson went further and expressed his belief that Ramos had refused to identify Jimenez only because his father, speaking in Spanish, appeared to discourage Ramos from getting involved. At the 2017 gateway innocence hearing, Ramos affirmed that he did not pick anyone out of the lineup and that his father told him not to get involved, but alleged that Detective Thompson pushed him to identify Jimenez by saying “it‘s number so and so.” Id. at 1037-38. Detective Thompson denied that allegation.
Velazquez next viewed the lineup at 7:16 p.m. and positively identified Jimenez. The reliability of that identification, however, was arguably diminished by an encounter Velazquez had on her way to the lineup. Brana‘s sister, who had copies of the Wanted poster with Jimenez‘s face on them in the backseat of her car, had driven Velazquez partway to the precinct. Velazquez saw the posters and discussed them with her sister-in-law. At trial, Velazquez admitted that she saw the Wanted poster just before viewing the lineup, but claimed that she picked Jimenez out of the lineup shortly thereafter based only on her observation of the shooter, not from the poster.
III. The Trial
Jimenez was principally charged with second degree murder in violation of
The trial lasted only four days.11 The State‘s case-in-chief relied entirely on eyewitness identification. Although the prosecutor‘s opening statement to the jury forecast only Velazquez as the State‘s principal witness, eventually, as described more fully below, both Velazquez and Ramos identified Jimenez from
the witness stand.12 The defense‘s opening also previewed that, because “this case is about mistaken identification,” it would primarily attack the reliability of Velazquez‘s identification. Id. at 210-11. Neither opening statement mentioned Ramos.
Velazquez and Ramos testified on the second day of trial. Velazquez took the stand first and, after delivering her account of the killing, identified Jimenez as the shooter in the courtroom. When asked how she described the killer to Detective Coor after the shooting, she testified that the killer was Dominican, about 25 years old, and had black hair about three to four inches in length in “like [Jh]eri curls.” Id. at 257. When asked to compare that description to Jimenez‘s appearance that day in the courtroom, she testified that he looked different than he had on the day of the shooting because “[h]is hair is shorter,” id. at 262, but said that she otherwise had no trouble recognizing him as the shooter.
On cross-examination, Velazquez denied that she gave two different descriptions – one of the harasser and one of the killer – to Detective Coor right
after the shooting. Rather, she testified that while she told Detective Coor about the harassment, she “didn‘t give him a description of the person who was harassing [her],” id. at 270, and insisted that she never even saw the harasser. She maintained instead that she only “described the killer,” id. at 276, including that the killer had “long, black [Jh]eri curls,” id. at 282, and that he “was Dominican,” id. at 293.13 She explained that she knew the killer was Dominican not from his accent (because she did not hear him speak) but that she could “tell this [was] a group of Dominicans” from the accent of the short man who first taunted Brana. Id. at 295. Finally, she testified that she had also described the killer to Detective Thompson, and that Detective Coor had not recorded on the police report every detail of the shooter‘s description that she had given to him because the top portion of the report recorded the suspect as Hispanic (rather than Dominican) and omitted the color of his skin and his “curly hair.” Id. at 328.
Velazquez‘s testimony conflicted with the testimonies of both detectives. Detective Thompson, called by the prosecution, testified that Velazquez had described two people, not one, in the post-shooting interview. And Detective Coor, called by the defense, more expansively testified that when he first arrived at the scene (before Detective Thompson), Velazquez gave him a description of the harasser, not the shooter, and told him that the harasser was a different man who sported long black Jheri curls, wore denim shorts (not khakis), and drove a red sports car. Detective Coor further testified that when Detective Thompson arrived later, it was he who took down Velazquez‘s separate description of the shooter – “a male Hispanic 25, 26,” wearing khakis, with no mention of curly hair, which is the description appearing in the preset fields “on the top” of the police report.14 Id. at 420. Detective Coor insisted that, based on what Velazquez told him, the shooter was not the same person
Velazquez‘s description of the shooter, moreover, did not match the description of Jimenez‘s appearance around the time of shooting as captured in
Jimenez‘s mugshots, and as recounted by Jimenez‘s probation officer at trial. Neither the March 1991 nor March 1992 mugshots of Jimenez that had been in the photographic arrays, nor the October 1992 mugshot taken at his booking, showed him wearing anything but short, mostly straight, hair – not Jheri curls, or even hair long enough to be styled into Jheri curls. Jimenez‘s probation officer, who had seen him regularly in face-to-face meetings during 1992, also testified that in June 1992, Jimenez had short hair similar to how he looked in the courtroom, and that he had never seen Jimenez with long Jheri curls (although he admitted that his recollection of Jimenez‘s appearance two years prior to the trial was vague).
B. Ramos
Ramos took the stand immediately after Velazquez and delivered brief, wavering testimony through a translator.15 He testified to his recollection of the shooting, including hearing the man he knew as Monaguillo call Brana the slur translated in English as “cocksucker.” But he also testified that he had picked
suspect number four in the lineup (not Jimenez), then appeared to change his mind, saying that he picked no one at all and, in fact, could not remember what the shooter looked like. He initially resisted the prosecutor‘s invitation to identify Monaguillo in the courtroom, responding that it had been a “long time” since he recognized him. Id. at 341. Later, after the prosecutor had reminded Ramos of his prior identification at the precinct, the prosecutor asked Ramos whether he saw “anyone in the courtroom that look[ed] like” the man who shot Brana, and Ramos pointed to Jimenez. Id. at 345.
On cross and redirect examination, Ramos described the basis for his identification: he was positive the shooter was Dominican,16 he recognized the killer from having seen Monaguillo before, and he knew the shooter‘s ethnicity because he had known Monaguillo “for about a year or two.” Id. at 352. Asked once more by the prosecutor if Monaguillo was in the courtroom - after the prosecutor assured him that “nobody is going to hurt you” and “[y]ou swore to tell the truth” - Ramos indicated Jimenez. Id. When defense counsel asked on recross if Ramos would describe the shooter as having Jheri curls, Ramos did not answer directly, instead responding that the shooter‘s hair was “[n]ot too long.” Id. at 354.
C. Closing and Verdict
The defense‘s closing argument followed its previewed strategy of arguing that Velazquez‘s identification was unreliable, primarily pointing to the short time she had to observe the shooter (thirty seconds), her distress and divided attention, her arguably tainted lineup identification, and the
The prosecutor, however, appeared to recognize that Velazquez‘s testimony was weaker than expected, because even while he argued that any inconsistencies in the details of Velazquez‘s testimony were attributable to her trauma and to a possible mistake by Detective Coor, he acknowledged she was “not the best witness in the world.” Id. at 598. He instead assured the jury that it could rely on Ramos as a disinterested observer of the shooting, remarking that, even though he had not expected Ramos to identify Jimenez because Ramos was clearly afraid of Jimenez, Ramos had “summ[on]ed up the courage . . . [and] finally did the right thing.” Id. at 598-99.
The jury deliberated for two days. It first asked for a readback of Ramos‘s description of the shooter, Velazquez‘s testimony about the harasser, Detective Thompson‘s testimony, and all physical evidence, and then it briefly deadlocked. After requesting another readback of Ramos‘s testimony, the jury returned a guilty verdict on October 4, 1994.
Jimenez was subsequently sentenced to an indeterminate term of 25 years to life in prison. His conviction and sentence were affirmed on direct appeal. People v. Jimenez, 647 N.Y.S.2d 947 (1st Dep‘t 1996) (rejecting Jimenez‘s unpreserved objection to the admission of Velazquez‘s lineup identification based on her “fortuitous viewing” of the Wanted poster), leave to appeal denied, 89 N.Y.2d 924 (1996). In 2001, he unsuccessfully moved pro se to vacate the judgment pursuant to
He did not try again for twelve years.
III. Newly Surfaced Evidence
A. Ramos‘s Recantation
Nineteen years after Jimenez‘s trial, Ramos recanted his testimony and declared that he was “completely certain that the person incarcerated is innocent.” App‘x 68. That certainty was based on Jimenez‘s ethnicity. In a handwritten affidavit dated April 11, 2013,17 Ramos averred that he knew that the killer, the man he recognized as Monaguillo, was Dominican because he heard him utter the exclusively Dominican slur “mamaguevo” just before shooting Brana. Following a chance encounter between Ramos‘s wife and Jimenez‘s sister in 2010, Ramos learned that Jimenez was actually Puerto Rican, not Dominican. That was when he realized that Jimenez was innocent, because “we Puerto Ricans use cabron” for the same slur, rather than “mamaguevo.” Id. at 67 (accent added).
That distinction in dialect was crucial, Ramos alleged, because he identified Jimenez in the courtroom at his trial only because Detective Thompson improperly influenced him. Ramos alleged that Detective Thompson, who had transported Ramos to the courthouse, misleadingly assured him that Jimenez was Dominican just before he took the stand. Specifically, Ramos averred:
[T]he black detective [Thompson] brought me to court to testified and the
placed me in a small room and showed me a picture allegedly of the person who killed [Brana] when they showed me the picture I asked if he was Dominican o Boricua [Puerto Rican] and he told me that he was Dominican I asked him because I wanted to make sure that he was not Puerto Rican because I knew that the person who killed him was Dominican. And when he showed me the picture I was still uncertain that its why I asked the detective if he was Dominican o Puerto Rican when he said to me that he was Dominican I thought that they had him and I thought that he was the one who killed [Brana]. When I asked him again if he was Dominican and he said he was I trusted the detective and the wife because the wife said he was that‘s what the detective said and I trusted him. When I was brought up to testify the detective told me not to look at anyone in the court when I was asked if he was there I then tried to look but I was still confused and had to point out the person that was there.
Id. at 67-68 (spelling and grammar retained).
At the gateway innocence hearing, Ramos further alleged that Detective Thompson not only showed him a photograph of Jimenez, but told him that Jimenez was the shooter, instructed him to point him out in the courtroom “without looking at him,” and promised him “all manner of castles and treasure.” Id. at 1044-45. Ramos maintained that he identified Jimenez at trial only “because of the pressure” applied by Detective Thompson. Id. at 1047.
Detective Thompson, at the same hearing, testified that he could not recall any such conversation with Ramos but denied that he would ever tell a witness that another witness already identified the defendant; he furthermore insisted that at the time, he did not understand any difference between Puerto Ricans and Dominicans, and classified all Spanish-speakers as Hispanics.
B. Alibi Witnesses
In 2013 and 2014, two former friends of Jimenez, Amancio Delgado and Danny Hernandez (the “Alibi Witnesses“), wrote affidavits swearing that on June 25, 1992, the day of the shooting, they had been with Jimenez all day on a street corner about a mile from Brana‘s apartment. That day was still memorable two decades later because, Hernandez averred, a group of his friends including Jimenez (nicknamed “Spaz“) were hanging out on their usual corner of Townsend Avenue and Mt. Eden in the Bronx to celebrate Hernandez‘s eighteenth birthday. They were there from the early afternoon until about 10:00 or 11:00 p.m. After Jimenez‘s arrest, Hernandez approached Jimenez‘s attorney and offered to testify, but he was never called.18 Hernandez never spoke with Jimenez again after his conviction.
Delgado similarly recounted that he regularly hung out with his friends on that street corner, and he had been with Jimenez (also nicknamed “Tito“) celebrating Hernandez‘s birthday between noon and
The Alibi Witnesses gave few additional details of relevance in their live testimony at the gateway innocence hearing. Hernandez testified that Jimenez had been with him most of the day, and in fact arrived at the corner before he did, but admitted that they had each separately left a few times to buy beer and marijuana and use the bathroom. Delgado, on the other hand, testified that Jimenez never left his line of sight that day.
IV. State Court Decision
In 2014, Jimenez once more sought to vacate his conviction in State court pursuant to
The State court denied Jimenez‘s motion without holding an evidentiary hearing. Id. at *1, *12. New York State law recognizes freestanding claims of actual innocence, but the State court, citing federal law, chiefly held that Jimenez failed to make a prima facie case of innocence because Ramos‘s recantation story was “implausible,” and in any event insufficient to establish Jimenez‘s innocence. Id. at *8. The State court found it notable that Ramos‘s affidavit was silent on the “foundation for defendant‘s initial identification,” which was that Ramos had claimed that he knew Monaguillo before the shooting and that Jimenez
The State court further found that Ramos‘s initial resistance to identifying Jimenez in the courtroom did not indicate coercion, but merely demonstrated that he had been a “hesitant, reluctant witness who was fearful to make an in-court identification and only did so when reminded that nobody was going to hurt him and that he swore to tell the truth.” Id. at *7. The State court accordingly rejected out of hand “the specious proposition that the two eyewitnesses, both of whom were Puerto Rican, would never erroneously identify a fellow Puerto Rican as Dominican.” Id. It reasoned in the alternative that, even if Ramos were credible, his recantation still could not demonstrate that Jimenez‘s innocence was “highly probable” in light of Velazquez‘s unequivocal in-court identification. Id. at *8. Instead, Ramos‘s recantation had merely “cast[] doubt” on Jimenez‘s guilt, because, “at most,” it was evidence in “direct conflict” with Velazquez‘s identification. Id.
The State court also found that the Alibi Witnesses’ affidavits lacked credibility because they claimed to remember their activities on a specific date over two decades ago. Id. at *9. But even overlooking credibility, the court reasoned that the alibi, taken as true, still placed Jimenez within a mile of the crime scene - close enough that the group could have “migrated the short distance to the scene of the homicide.” Id. The State court also pointed to Delgado‘s affidavit in support of its inference that, contrary to any suggestion of a sharp social separation between Dominicans and Puerto Ricans in the Bronx in the early 1990s, Jimenez‘s teenage group of friends included at least one Dominican from whom he could have easily picked up a Dominican profanity. Id.
Finally, the State court found there had been no Brady violation stemming from Detective Thompson‘s alleged statement to Ramos about Jimenez‘s ethnicity because that conversation was not material exculpatory evidence. Id. at *12-13. The State court‘s judgment became final when Jimenez‘s request for permission to appeal the judgment was denied in November 2015.
V. Gateway Innocence Hearing
Jimenez filed a petition for a writ of habeas corpus a few days after expiration of the one-year deadline to petition from a final judgment pursuant to
As relevant here, Magistrate Judge Francis found that, even according the factual conclusions of the State court a presumption of correctness as required by
The district court adopted that recommendation in full. Jimenez II, 2018 WL 2768644, at *1. It stressed, however, that its gateway innocence findings did not reflect its ultimate determination on the merits. See id. at *17. Shortly thereafter, in July 2018, Jimenez was released from State prison on parole.
VI. The District Court Opinion
On September 15, 2021, the district court denied habeas relief on the merits of Jimenez‘s claims. Jimenez III, 560 F. Supp. 3d at 764. The court principally held that, “considering the extraordinarily high showing that would be required for a freestanding claim of actual innocence and the deference due a state court‘s merits adjudication under AEDPA,” it had to deny Jimenez‘s actual innocence claim because he merely “cast considerable doubt on his guilt,” but did not affirmatively demonstrate “that he did not commit, or could not have committed” the crime. Id. at 771-72 (emphasis in original) (internal quotation marks omitted), first quoting House, 547 U.S. at 555 (2006); and then quoting Hyman v. Brown, 927 F.3d 639, 665 (2d Cir. 2019). That was so because there remained sufficient doubt of Jimenez‘s innocence, as the Alibi Witnesses did not rule out the possibility that Jimenez had traveled unnoticed to the scene of the shooting, and Velazquez‘s identification was not wholly unreliable. Id. at 772. The district court also found no merit to Jimenez‘s Brady/witness tampering claim based on Detective Thompson‘s allegedly misleading remarks to Ramos about Jimenez being Dominican. Id. at 773-74. This appeal followed.20
DISCUSSION
Jimenez maintains that he was wrongly convicted of a crime he did not commit. Under AEDPA, however, federal courts may entertain his petition for a writ of habeas corpus “only on the ground that he is in custody in violation of the Constitution or laws . . . of the United States.”
We decide only those issues necessary to our conclusion that the district court‘s decision to deny habeas relief from Jimenez‘s State conviction on the merits, which we review de novo, must be affirmed. Scrimo v. Lee, 935 F.3d 103, 111 (2d Cir. 2019). We also review the “district court‘s ultimate finding [relating to] actual innocence de novo.” Rivas v. Fischer, 687 F.3d 514, 543 (2d Cir. 2012) (internal quotation marks omitted), quoting Doe v. Menefee, 391 F.3d 147, 163 (2d Cir. 2004).
As a preliminary matter, however, we note that while we review the district court‘s decision without deference, our authority to grant Jimenez relief is tightly circumscribed because
Nonetheless, our review of the record confirms that Jimenez has not affirmatively demonstrated his innocence under any plausible set of assumptions or standard of review. Even after Ramos‘s recantation and Jimenez‘s newly introduced alibi, Velazquez‘s identification remains as sufficiently incriminating as it did almost thirty years ago. As a result, AEDPA deference or not, Jimenez has not proven his innocence. Nor has he shown that a Brady violation occurred at his trial. We must, therefore, deny habeas relief.
I. Freestanding Actual Innocence
The legal landscape of freestanding actual innocence is hostile territory for a petitioner seeking relief from a state conviction. At least four barriers stand between Jimenez and his claim for relief. The first is the open question of whether a freestanding innocence claim is even cognizable at all under federal law, especially in a noncapital case. The second barrier is the statute of limitations; the State argues that our analysis ought to stop there because, it contends, the district court erroneously reached the merits of Jimenez‘s claims by finding that he had sufficiently demonstrated his actual innocence to avoid that obstacle. The third barrier is the “extraordinarily high” standard of proof that the petitioner must meet to affirmatively establish innocence, assuming such a claim is cognizable. Herrera v. Collins, 506 U.S. 390, 416-17 (1993). Finally, after all these hurdles are surmounted, the fourth barrier is the deference owed under
Jimenez must surmount each of these barriers to obtain relief on his freestanding
A. Cognizability
As a first step in assessing the ultimate merits of Jimenez‘s petition, we address his threshold argument that a freestanding claim of actual innocence is cognizable under the Eighth and Fourteenth Amendments. To this day, “[w]hether . . . a federal right [based on a claim of actual innocence] exists is an open question.” Dist. Att‘y‘s Off. for Third Jud. Dist. v. Osborne, 557 U.S. 52, 71 (2009). The Supreme Court has provided scant guidance on the cognizability of a freestanding right to habeas corpus relief based on a claim of wrongful conviction after a fair trial free of error. Neither it, nor we, nor any of our sister circuits have recognized the existence of such a right or granted habeas relief based on such a claim.21 In Herrera v. Collins, the Court merely assumed without deciding that “in a capital case a truly persuasive demonstration of ‘actual innocence’ made after trial would render the execution of a defendant unconstitutional,” but concluded that the petitioner in that case had fallen “far short” of making that demonstration. 506 U.S. at 417-19 (emphasis added).
More importantly, Herrera announced that the burden of proof for a “threshold showing” of such a right “would necessarily be extraordinarily high.” Id. at 417. Because we “assum[e], arguendo, that [an innocence right] exists” for the purposes of this case, and further assume that its scope extends beyond a right against execution, we follow the path endorsed by the Supreme Court and other courts by bypassing “the difficult questions [recognizing] such a right would pose,” and turn instead to “the high standard any claimant would have to meet.” Osborne, 557 U.S. at 71.
Before proceeding, however, it is worth at least noting why those questions are so difficult. To the lay observer, it may seem intuitively obvious that the continued incarceration of an innocent person must violate the Constitution. Indeed, we have
B. Gateway Innocence
We begin by assuming, without deciding, that Jimenez‘s untimeliness has been excused and thereby bypass the State‘s alternative argument that his petition is time-barred.22 That said, the distinction between the more established standards of gateway innocence and the unsettled, hypothetical standards of freestanding innocence hinted at in Herrera is central to our analysis of the merits. Our discussion therefore starts with a necessary review of the strength of Jimenez‘s actual innocence claim under the threshold standards of gateway innocence.
Gateway innocence is an equitable exception allowing avoidance of AEDPA‘s one-year limitations period,
Untimeliness can be excused in a “narrow class of ‘truly extraordinary’ cases” of claims of actual innocence, Hyman, 927 F.3d at 656, quoting Schlup, 513 U.S. at 338 (Rehnquist, J., dissenting), where such claims are supported by new evidence of “factual innocence, not mere legal insufficiency,” Bousley v. United States, 523 U.S. 614, 623 (1998). Even if a petitioner overcomes this “demanding and rarely met” standard, however, gateway innocence “cannot itself afford [the petitioner] habeas relief from his state conviction. It can only open a gateway to federal review” of an otherwise barred claim “that, if itself successful, could afford him relief.” Hyman, 927 F.3d at 655, 662.
To reach the merits of such a claim, a habeas petitioner must make a “sufficiently credible and compelling” claim of innocence. Id. at 657; see also Rivas, 687 F.3d at 541 (“For the [innocence] claim to be ‘credible,’ it must be supported by ‘new reliable evidence . . . that was not presented at trial.‘“), quoting Schlup, 513 U.S. at 324. The second prong is our starting point for the standard of freestanding innocence.24 A compelling innocence claim must demonstrate that, “more likely than not, in light of the new evidence, no reasonable juror would find [the petitioner] guilty beyond a reasonable doubt.” House, 547 U.S. at 538. To make that determination, a habeas court must consider all record evidence, “old and new.”25 Id. If, after that exhaustive review, the petitioner‘s evidence of innocence is so compelling that the habeas court “cannot have confidence in the outcome of the trial unless the court is also satisfied that the trial was free of nonharmless constitutional error,” the court may adjudicate the merits of otherwise procedurally barred claims. Schlup, 513 U.S. at 316.
That is no nominal demand. Courts have seldom found compelling claims of gateway innocence since the Supreme Court in Schlup recognized the exception, and Schlup itself only suggested that a new alibi witness and eyewitness testimony incriminating a third-party suspect could be compelling evidence of innocence. Id. at 331. The Supreme Court in the end declined to decide whether that standard had been met on the facts of that case. See id. at 331-32.
The only other examples of sufficiently compelling gateway innocence claims were close cases. In House v. Bell, the Supreme Court found compelling evidence of the petitioner‘s innocence in new forensic DNA evidence, which “called into question” the “central forensic proof connecting [him] to the crime,” in combination with “substantial evidence” incriminating a third-party suspect who had essentially confessed. 547 U.S. at 540-41, 553-54. Yet, House cautioned that other
Similarly, in Rivas v. Fischer, we found compelling evidence of the petitioner‘s innocence in a new expert opinion that revised the victim‘s time of death to a period when he had an unchallenged alibi, as well as in unchallenged testimony discrediting the originally incriminating medical examiner‘s report. 687 F.3d at 546. But Rivas again stressed that it was a “close case” and cautioned that “we would not expect a lesser showing of actual innocence to satisfy the Schlup standard.” Id.
More recently, in Hyman v. Brown, we observed that Schlup, House, and Rivas shared a common feature: new evidence “directly supported petitioner‘s factual innocence by indicating either that he did not commit, or could not have committed, the crimes of conviction.” 927 F.3d at 665 (emphases in original). Hyman reasoned that in each case, compelling evidence of innocence came through a combination of eyewitness misidentification, testimony and forensic evidence ruling out the petitioner‘s ability to commit the crime, and evidence strongly incriminating a third party. Id. at 665-66. We concluded that the petitioner had not met his burden in that case because even though the only eyewitness to identify him as a participant in a large shootout (out of a total of four eyewitnesses) had recanted her testimony, that alone did not affirmatively demonstrate that the petitioner did not, or could not have, fired a gun when the witness‘s recantation was compared to the weight of the remaining incriminating evidence (the other three eyewitnesses’ testimonies, ballistics evidence, and the petitioner‘s concession that he had been at the scene). Id. at 666-71.
The State argues that Hyman forecloses the district court‘s finding that Jimenez made out a compelling claim of innocence because Velazquez‘s credible identification remains strongly incriminating. Velazquez‘s identification, however, is not nearly as strong as the remaining inculpatory evidence in Hyman. Unlike Hyman, here there is no physical evidence, no other eyewitnesses (given the recantation of Ramos), and no factual concessions by Jimenez linking him to the time or place of the crime. There is only the uncorroborated accusation of a single eyewitness, whose reliability had been vigorously contested through cross-examination at trial. Velazquez‘s description of the shooter‘s appearance - specifically the shooter‘s distinctive Jheri curls - is at odds with Jimenez‘s appearance in 1992 and further undermines confidence in her reliability; indeed, the prosecution itself acknowledged at trial that Velazquez was “not the best witness in the world.” App‘x 598. The weaker incriminating evidence in this record distinguishes Hyman. The closeness of this case is sufficient to make us prefer to simply assume arguendo that Jimenez demonstrated a sufficient likelihood of innocence to satisfy Schlup‘s “gateway” standard.
Yet that is not the end of our analysis. Although we assume that Jimenez has met his threshold burden, he has, at most, made out only a “close case” of gateway innocence. Rivas, 687 F.3d at 546. As in House and Rivas, the remaining inculpatory evidence is weak, and “it may be enough for the petitioner to introduce credible new evidence that thoroughly undermines the evidence supporting the
Jimenez‘s exculpatory evidence is, in fact, the same type of evidence as Velazquez‘s testimony - witnesses swearing to what they saw, heard, did, and thought. The resulting credibility contest, which may be compelling enough for the purposes of gateway innocence given the dearth of physical evidence cutting in either direction, does not surpass the heightened standard required on review of the merits of a freestanding innocence claim.
C. Standard of Proof
Having assumed that Jimenez has passed through the actual innocence gateway, what standard applies when, after opening the door, all that lies beyond is the same claim of innocence? The Supreme Court struggled with that question prior to AEDPA‘s passage in 1996. In Herrera and its progeny, it decided only that whatever the standard of proof, it is “extraordinarily high.” See Herrera, 506 U.S. at 417. Indeed, no court has ever found it met.
Jimenez‘s most potent argument on appeal is his effort to favorably calibrate Herrera‘s “extraordinarily high” standard of proof in freestanding innocence claims. Id. That effort, however, is ultimately unavailing. The trio of Supreme Court cases that form the backbone of what little federal law shapes the appropriate standard of proof demonstrates that, despite Jimenez‘s effort to articulate the exact quantum of proof required to show that he is truly innocent, he has not met his burden in this case.
Pre-AEDPA, Herrera vaguely described a cognizable actual innocence claim as requiring a “truly persuasive demonstration” in light of “the very disruptive effect that entertaining claims of actual innocence would have on the need for finality in capital cases, and the enormous burden that having to retry cases based on often stale evidence would place on the States.” Id. Schlup, which followed Herrera, then described the difference between the proof required to establish gateway innocence, as opposed to substantive innocence, by illustration:
In Herrera . . ., the evidence of innocence would have had to be strong enough to make his execution “constitutionally intolerable” even if his conviction was the product of a fair trial. For Schlup, the evidence must establish sufficient doubt about his guilt to justify the conclusion that his execution would be a miscarriage of justice unless his conviction was the product of a fair trial.
Schlup, 513 U.S. at 316 (emphases in original). Accordingly, Schlup concluded, if “there were no question about the fairness of the criminal trial, a Herrera-type claim would have to fail unless the federal habeas court is itself convinced that those new facts unquestionably establish [the petitioner‘s] innocence.” Id. at 317 (emphasis added). Over ten years later, and long after AEDPA‘s enactment, the Supreme Court in House finally reasoned that the “sequence of the Court‘s decisions in Herrera and Schlup - first leaving unresolved the status of freestanding claims and then establishing the gateway standard - implies at the least that Herrera requires more convincing proof of innocence than Schlup.” House, 547 U.S. at 555.
Like Herrera, Schlup,
We agree with the district court, though we express no view of its suggestion that a hypothetically lower standard of proof may be limited to actual innocence claims in capital cases. See id. at 769-70. We affirm only that, as a matter of law and as unequivocally set forth by the Supreme Court, the degree of proof required to make out a freestanding claim of innocence (assuming that such a claim is cognizable on habeas corpus even in a noncapital case) exceeds the proof required to establish gateway innocence. See House, 547 U.S. at 554-55. Jimenez‘s proof has not reached those heights. We need not identify the exact degree of proof required because, as we have already described, the same body of evidence that could, at most, only barely surmount the gateway innocence barrier in this case logically “falls short” of the greater certainty demanded in substantive innocence claims.26 Id. at 555.
Jimenez resists that conclusion. He raises three objections to the perceived contradiction in the district court‘s decision that, even though he prevailed at the gateway innocence stage, he cannot do so at the merits stage.
He and his supporting amici curiae primarily argue that the district court‘s reliance on the Hyman criteria - holding Jimenez to the burden of proving that he “did not commit, or could not have committed” the murder, 927 F.3d at 665 (emphases in original) - improperly imposed an onerous standard beyond what Herrera demands by requiring him to affirmatively prove that it was impossible that he had killed Brana.27 But, as Jimenez recognizes, Hyman
Jimenez next argues that he has met the standard required to affirmatively prove his innocence because he has shown that it is likely that no reasonable juror would convict him in light of Ramos‘s recantation and the Alibi Witnesses’ testimonies. That metric, however, reflects the less demanding gateway standard, which calls for an assessment of the probability that a rational jury would have acquitted him in a hypothetical trial in a manner akin to prejudice or harmless error standards in other constitutional claims.28 House, 547 U.S. at 538. Such a hypothetical second trial, moreover, is not easily assessed. One can, of course, predict that a trial without Ramos (or with a fatally impeached Ramos) and with the two Alibi Witnesses would present a difficult road for a prosecutor relying solely on Velazquez‘s testimony. But we return to the essential problem: how do we weigh Velazquez‘s testimony? Even assuming she is available to testify, a witness who was believable when testifying a few years after the event will necessarily find it harder to recall details thirty years later. An effort to cobble together a hypothetical trial that weighs the credibility of Velazquez‘s original testimony, which none of us saw in person, against the putative future testimony of the Alibi Witnesses (whom we have also not seen testify), is an inherently problematic exercise. Herrera, Schlup, and House, by contrast, are clear that demonstrating freestanding innocence calls for more than a sheer probability that a new trial, proceeding under impossible real-world conditions, would likely end in acquittal. House, 547 U.S. at 554-55; see also Herrera, 506 U.S. at 403 (rejecting a “probable innocence” standard that “would in effect require the State to retry [the] petitioner . . . simply because of a belief that in light of petitioner‘s new-found evidence a jury might find him not guilty at a second trial“).
Jimenez and amici finally argue that a “clear and convincing evidence” standard of proof satisfies Herrera‘s calling and should have been applied by the district court instead of its reliance on the Hyman criteria. “‘[C]lear and convincing evidence’ . . . means something more than ‘preponderance of the evidence,’ and something less than ‘beyond a reasonable doubt.‘” United States v. Chimurenga, 760 F.2d 400, 405 (2d Cir. 1985). The standard has also been more precisely defined as “plac[ing] in the ultimate factfinder an abiding conviction that the truth of [the] factual contentions are ‘highly probable.‘” Colorado v. New Mexico, 467 U.S. 310, 316 (1984); see also Addington v. Texas, 441 U.S. 418, 424 (1979) (observing that “clear and convincing evidence” is an “intermediate standard” of proof typically used in civil cases involving “quasi-criminal wrongdoing” and that it is used to “protect particularly important individual interests“).
Jimenez and amici contend that Herrera logically requires habeas petitioners to prove their innocence only to a degree of certainty one step greater than the “preponderance of the evidence” standard applicable in habeas review of other constitutional claims.29 We do not necessarily quarrel with that framing; it is clear to us that if a habeas court could ever grant relief on a freestanding innocence claim, such a claim would require at least clear and convincing evidence of actual innocence. Beyond that proposition, however, there is little authority for Jimenez‘s argument that clear and convincing proof of actual innocence is enough.30 Nothing in
Examples further afield of wrongful conviction claims also cut against any conclusion that clear and convincing evidence of innocence is sufficient to meet the threshold implied in Herrera. See, e.g., Sawyer v. Whitley, 505 U.S. 333, 336, 340 (1992) (contrasting innocence of the death penalty, which requires “clear and convincing” proof of innocence to excuse a death-row petitioner‘s defaulted claim, with the “prototypical example of ‘actual innocence’ . . . where the State has convicted the wrong person of the crime“). For instance, we have held that in habeas petitions by civil immigration detainees, the government must justify a civil immigration detention by clear and convincing evidence because “[t]he Supreme Court has consistently held the Government to a [higher] standard of proof . . . where liberty is at stake.” Velasco Lopez v. Decker, 978 F.3d 842, 856 (2d Cir. 2020). But that is because the “necessary scope of [habeas] review and resulting relief in part depends upon the [procedural] rigor of any earlier proceeding,” and “the most searching review” is called for when detention “occur[s] without the procedural protections required in courts of law.” Id. at 850 (emphasis added) (internal quotation marks omitted), quoting Boumediene v. Bush, 553 U.S. 723, 781 (2008). Criminal adjudications, by contrast, quintessentially require rigorous procedure, afford defendants maximum procedural protections, and hold the State to prove guilt beyond a reasonable doubt. See id. at 850-51; cf. Addington, 441 U.S. at 423-24 (explaining that “beyond a reasonable doubt” applies in criminal cases to protect defendants “by standards of proof designed to exclude as nearly as possible the likelihood of an erroneous judgment“).
As a result, we cannot be sure that a clear and convincing standard of proof is appropriate in light of the uniquely powerful competing interests repeatedly emphasized by the Supreme Court. “The function of a standard of proof . . . is to instruct the factfinder concerning the degree of confidence our society thinks he should have in the correctness of factual conclusions for a particular type of adjudication . . . [and] serves to allocate the risk of error between the litigants and to indicate the relative importance attached to the ultimate decision.” Addington, 441 U.S. at 423 (internal quotation marks and citation omitted). While the truth of a habeas petitioner‘s ultimate guilt or innocence is undoubtedly of the highest importance, the Supreme Court has repeatedly admonished that a state‘s countervailing interests - the “need for finality” and avoiding the onerous burden of retrying stale cases - are also important because “in state criminal proceedings the trial is the paramount event for determining the guilt or innocence of the defendant.” Herrera, 506 U.S. at 416-17 (emphasis added); see also Harrington v. Richter, 562 U.S. 86, 103 (2011) (declaring that habeas review involves an “intru[sion] on state sovereignty to a degree matched by few exercises of federal judicial authority“), quoting Harris v. Reed, 489 U.S. 255, 282 (1989) (Kennedy, J., dissenting). Under those parameters and our own struggle to envision a clear result in a hypothetical retrial of this ambiguous case, we cannot - and do not - definitively say that proof of innocence by clear and convincing evidence strikes a “fair balance” that “allocate[s] the risk of error,” Addington, 441 U.S. at 423, 431, between a convicted prisoner found guilty after a fair trial free of error, and the “special costs [imposed] on our federal system” by exercising our habeas authority to “override[] the States’ core power to enforce criminal law,” Shinn v. Ramirez, 596 U.S. 366, 376 (2022) (internal quotation marks omitted), quoting Engle v. Isaac, 456 U.S. 107, 128 (1982).
There may be a case that one day calls for such a definitive determination, but this is not that case. That is because, even accepting Jimenez‘s preferred burden of proof for the sake of argument, he has not demonstrated his innocence
by clear and convincing evidence. Accordingly, we have no need to decide whether petitioners’ evidentiary burden in freestanding innocence claims requires clear and convincing evidence of actual innocence, or some higher standard, such as proof beyond a reasonable doubt. See Herrera, 506 U.S. at 416-19.
Indeed, this case perfectly illustrates how evidence of innocence can be both “compelling” under Schlup, yet fall short of demonstrating substantive innocence under Herrera. Ramos‘s recantation is not a model of clarity. Nor is Jimenez‘s exculpatory evidence, taken as a whole, clear and convincing proof of his innocence in light of the lingering uncertainty from Velazquez‘s incriminating identification. Faced with a body of evidence that generates uncertainty as between guilt and innocence, we cannot conclude that Jimenez‘s innocence has been demonstrated to a “highly probable” degree. Colorado, 467 U.S. at 317 (emphasis added).
Ramos‘s recantation, taken as true, is more confusing than strongly probative of Jimenez‘s innocence. Ramos now insists that he knew the shooter was Dominican based solely on his word choice, but that does not square with the information Ramos gave to police the day after the shooting, or with his trial testimony that he knew that the shooter was Dominican because he recognized him as the man called “Monaguillo” with whom he was acquainted. Significantly, both Velazquez and Ramos testified at trial that they believed the shooter was Dominican for separate reasons: Velazquez because the shooter‘s compatriot spoke in a Dominican accent when taunting Brana, and Ramos because he had seen Monaguillo in the neighborhood “for about a year or two” prior to the shooting and knew him to be Dominican. App‘x 352. Ramos‘s present-day testimony that he actually knew Monaguillo to be Dominican only because he heard the shooter say “mamaguevo” is inconsistent with his statement to police the day after the shooting that he recognized Monaguillo using his sight and recollection (rather than his ears) and that Jimenez‘s picture, selected from a photographic array, matched that recollection. Even if Detective Thompson improperly influenced that selection as Ramos alleged at the gateway innocence hearing, Ramos‘s recantation does not undermine Velazquez‘s independent identification of Jimenez in another
Nothing in the record, moreover, convincingly supports the key premise of Ramos‘s recantation that, just because Jimenez is Puerto Rican - a trait that he and his counsel would have known distinguished him from the identifying features recounted by the eyewitnesses at trial - he would have never uttered a vulgarity used mainly by Dominicans. Both eyewitnesses agree that the shooter advanced on Brana in the company of at least one Dominican, and Delgado, one of Jimenez‘s alibi witnesses and close teenage friend, is also Dominican. Whatever Ramos‘s sincere beliefs about a lack of linguistic cross-pollination between Puerto Ricans and Dominicans in the Bronx, it is entirely possible that a Puerto Rican teenager fraternizing with Dominican neighbors might adopt a profanity from Dominican Spanish into his own manner of speaking.
The Alibi Witnesses leave us with the same nagging uncertainty. Technically, the only new affirmative evidence of innocence introduced in Jimenez‘s habeas petition is the alibi defense that his trial attorney never raised; Ramos‘s recantation, at most, reduces the total amount of inculpatory evidence. But even that alibi fails to inspire confidence that Jimenez is clearly innocent. Hernandez admitted that both he and Jimenez had left the street corner from time to time. And even granting that Delgado sincerely believes, as he testified at the gateway innocence hearing, that Jimenez remained at the corner celebrating Hernandez‘s birthday for ten straight hours on a day when a group of at least eight teenagers “came and left” from that public street corner, we cannot afford his assertion that Jimenez never “left [his] line of sight,” App‘x 1222, significant weight in light of Hernandez‘s admission.
At the very least, the alibi only contradicts Velazquez‘s identification. It does not undermine her trial testimony that she had no doubt that she saw Jimenez, a stranger to her, shoot her husband. We reiterate that, while Velazquez‘s reliability as an eyewitness was impeached at trial, nothing in Ramos‘s recantation or the Alibi Witness testimonies “thoroughly” destroys the basis for her identification of Jimenez. Rivas, 687 F.3d at 543.
Rather, the main reason to view Velazquez‘s identification with suspicion is not that her multiple post-shooting descriptions, as recorded in the police report by Detective Coor, contradicted her description of the killer at trial. Instead, it is the fact that her consistent description that the killer had Jheri curls bears little resemblance to Jimenez‘s recorded appearance in 1992. But even then, no evidence in the record, old or new, undermines Velazquez‘s uncoerced and close-in-time identification of Jimenez in the photographic array. She also consistently maintained at trial that, to the extent Detective Coor‘s testimony that she had given him two descriptions of two different men in the post-shooting interview conflicted with her own recollection, she simply had not described the harasser to the detectives (or even looked at him), allowing the fair inference that it was not Velazquez‘s initial description that was mistaken, but Detective
That remaining thread of incriminating evidence is enough to cause Jimenez‘s innocence case to fall short of the threshold implied in Herrera. The freestanding innocence standard does not require petitioners to eliminate all doubt of their guilt. But the facts of this case illustrate that it requires more than an affirmative demonstration of mere reasonable doubt, or even of “considerable doubt,” House, 547 U.S. at 555, that Jimenez is guilty of the crime. Put another way, the innocence claim, assuming it is cognizable, does not require Jimenez to prove to an absolute certainty that he is innocent, or that he could not have committed the crime. It does, however, require a much higher degree of certainty than this record provides.
Whatever the exact descriptor applied, the standard for freestanding innocence is set above the already demanding requirements of gateway innocence. That is enough to conclude that Jimenez would not meet his burden even if his freestanding innocence claim is cognizable and the standard of proof is set at “clear and convincing evidence” of actual innocence. And it is more than enough to conclude that, even if we thought differently, we would not be at liberty to grant Jimenez relief under AEDPA.
D. AEDPA
The foregoing discussion demonstrates that all that is clearly established in the law of freestanding innocence is the lack of clarity about the scope of any actual innocence right and the evidentiary burden necessary to demonstrate it. That brings us to AEDPA, the final barrier to both Jimenez‘s claim and to our own authority to decide differently. We cannot determine that the State court was wrong to deny Jimenez‘s claim, because even if we could characterize the doubt that Jimenez has cast on his guilt as sufficient to satisfy Herrera, the deferential requirements of AEDPA all but foreclose the possibility of granting relief on such a claim where there is little “clearly established [f]ederal law.”33
Deference is owed even if the State court does not supply any reasoning in support of its conclusion on the merits, or “cite or even [demonstrate] aware[ness] of [Supreme Court] cases under
We find no fault in that reasoning. As already discussed, Jimenez has not supplied new evidence so powerful that it “unquestionably establish[es]” his innocence. Schlup, 513 U.S. at 317. Therefore, because the State court‘s analysis is not “so flawed as to undermine confidence that the constitutional claim has been fairly adjudicated,” McCray, 45 F.4th at 640 (internal quotation marks omitted), quoting Cruz v. Miller, 255 F.3d 77, 86 (2d Cir. 2001), we are compelled to conclude that, on these facts, the district court simply could not have granted relief on Jimenez‘s actual innocence claim.
Jimenez does not contest the district court‘s conclusion that the State court did not misapply the law. Jimenez III, 560 F. Supp. 3d at 771. Instead, Jimenez argues that AEDPA does not apply to freestanding claims of actual innocence because that right, if it exists, is a fundamental constitutional right that requires petitioners to shoulder a heavier burden than required for other “arguably less fundamental” rights. Appellant‘s Br. 34. Therefore, innocence claims, which demand a “thorough inquiry for the truth above all,” must not fall within AEDPA‘s ambit. Id. at 32.
Jimenez more expansively contends that AEDPA unconstitutionally restricts federal review of the merits of his innocence claim. He argues that, because innocence “cuts to the heart of the ‘Great Writ,’ which has historically been a mechanism to protect the wrongfully convicted from erroneous incarceration,” Appellant‘s Br. 26, “logic suggests” that the deference mandated by AEDPA “improper[ly] interfere[s] with the truth/justice-seeking function” of habeas review. Id. at 32. To supply that logic, Jimenez points to In re Davis, 557 U.S. 952 (2009), a habeas petition in a capital case that the Court transferred to a Georgia district court with instructions to conduct factfinding on the petitioner‘s innocence. Id. at 952. Writing in concurrence, three Justices opined that even if AEDPA “applies in full, it is arguably unconstitutional to the extent it bars relief for a death row inmate who has established his innocence.” Id. at 953 (Stevens, J., concurring); see also id. at 953-54 (“Alternatively, the court may find in such a case that [AEDPA‘s] text is satisfied, because decisions of this Court clearly support the proposition that it ‘would be an atrocious violation of our Constitution and the principles upon which it is based’ to execute an innocent person.“), quoting In re Davis, 565 F.3d 810, 830 (11th Cir. 2009) (Barkett, J., dissenting).
That speculation about what is “arguably unconstitutional,” id. at 953, about AEDPA as applied to the availability of habeas relief in capital cases provides no guidance here, in a noncapital case.34 Moreover, AEDPA‘s
requirement
Ε. Reasonableness of the State Court‘s Decision
Jimenez finally argues that even applying AEDPA deference and an extraordinarily heavy burden of proof, the State court “unreasonabl[y] determin[ed] the facts” because its failure to hold an evidentiary hearing resulted in an incomplete development of the record supporting Jimenez‘s innocence.36
“Where ‘[r]easonable minds reviewing the record might disagree’ as to the relevant finding, that [disagreement] is not sufficient to supplant the state court‘s factual determination.” Id., quoting Rice v. Collins, 546 U.S. 333, 341-42 (2006). The State court even assumed that Ramos and the Alibi Witnesses were credible, but still concluded that their stories, taken together, did not show that it was “highly probable that [Jimenez] is innocent“; instead, they “merely cast[] doubt” on Jimenez‘s guilt and conflicted with Velazquez‘s identification. Jimenez I, 2015 WL 770457, at *8. A “state-court factual determination is not unreasonable merely because the federal habeas court would have reached a different conclusion in the first instance.” Wood v. Allen, 558 U.S. 290, 301 (2010). As we have made clear, we are far from reaching a different conclusion than the ones the State court reached in this case.
Accordingly, the district court did not erroneously determine that AEDPA deference all but compelled it to deny habeas relief on the merits. Even though it disagreed with the State court for the purposes of gateway innocence, it could not characterize that court‘s conclusions on the weight of the evidence as unreasonable. And, because the district court did not find differently, it could not grant Jimenez relief on his substantive claim of actual innocence. See Jimenez III, 560 F. Supp. 3d at 771-72. Neither can we.
II. Brady
Jimenez is not entitled to habeas relief based on his standalone actual innocence claim, but because his untimely petition passed through the innocence gateway, he might still be entitled to relief if he can demonstrate that Detective Thompson tampered with Ramos‘s testimony just before he took the stand at Jimenez‘s trial, and that that tampering was not disclosed to the defense, in violation of Brady. According to Ramos, Detective Thompson allegedly told him that Jimenez was Dominican, showed him a picture of Jimenez, and promised him “castles and treasure” if he identified Jimenez in the courtroom. App‘x 1044. The State court denied that claim without extended analysis. Jimenez I, 2015 WL 770457, at *12-13. Jimenez therefore argues on appeal that the State court unreasonably applied Brady law to Detective Thompson‘s undisclosed interference with Ramos‘s courtroom identification because, he urges, knowledge of that conversation was favorable to the defense and would have impeached Ramos‘s identification of Jimenez at trial.37
Brady held that “the government violates the Constitution‘s Due Process Clause ‘if it withholds evidence that is favorable to the defense and material to the defendant‘s guilt or punishment.‘” Turner v. United States, 582 U.S. 313, 315 (2017) (emphasis omitted), quoting Smith v. Cain, 565 U.S. 73, 75 (2012). The government has an affirmative duty to disclose favorable exculpatory evidence that would impeach the credibility of witnesses. Giglio v. United States, 405 U.S. 150, 154-55 (1972). To make out a Brady violation, a petitioner must show that material exculpatory or impeachment evidence was suppressed by the state, either willfully or inadvertently. United States v. Bagley, 473 U.S. 667, 674-75 (1985) (plurality opinion). “In other words, true Brady material must be (1) favorable, (2) suppressed, and (3) prejudicial.” United States v. Hunter, 32 F.4th 22, 31 (2d Cir. 2022).
The State court dismissed Jimenez‘s Brady claim on the basis that Detective Thompson‘s pre-testimony statement to Ramos was not exculpatory Brady material, but it did not identify whether it concluded as much because that conversation was not favorable, not suppressed, or not prejudicial. See Jimenez I, 2015 WL 770457, at *13. The district court, in turn, concluded that the State court‘s conclusion was supportable for three independently sufficient reasons. Jimenez III, 560 F. Supp. 3d at 773-74. One of those reasons was that “fairminded jurists could disagree over whether the Ramos affidavit identified any evidence favorable to the defense that prosecutors wrongfully withheld” because the conversation as alleged in that affidavit had nothing to do with the basis for Ramos‘s identification as understood by all parties at the time of the trial: his recognition of “Monaguillo.” Id. That conversation,
The allegedly improper encounter between Detective Thompson and Ramos was not favorable Brady material. Favorable impeachment evidence offered to discredit a witness - in this case, Ramos - includes evidence that could have significantly weakened his eyewitness account, undermined the “thoroughness and ... good faith” of the police investigation, Kyles v. Whitley, 514 U.S. 419, 441-45 (1995), or “could have helped the defense suggest an alternative perpetrator,” Boyette v. Lefevre, 246 F.3d 76, 91 (2d Cir. 2001). But in 1994, the fact that Ramos had based his identification of Jimenez on his mistaken belief that Jimenez was Dominican was, so far as his affidavit suggests, a private thought expressed to no one at the time. The content of his conversation with Detective Thompson demonstrates that none of the involved parties would have understood his queries as material to guilt or innocence that could have undermined the basis for his identification of Jimenez.
Ramos‘s affidavit describes only his internal narrative about why he considered information about Jimenez‘s ethnicity important. Ramos averred that he asked “because [Ramos] wanted to make sure that [Jimenez] was not Puerto Rican because [Ramos] knew that the person who killed [Brana] was Dominican,” and that he “was still uncertain” as to the shooter‘s identity until Detective Thompson mistakenly told him that Jimenez was indeed Dominican, at which point he “thought that they had him” and “thought that [Jimenez] was the one who killed [Brana].” App‘x 67-68. Nowhere in Ramos‘s attestation is there any suggestion that Detective Thompson knew or should have understood from his questioning that the reason Ramos wanted to know whether Jimenez was Dominican or Puerto Rican was to allay his uncertainty about identifying the killer, as opposed to idle curiosity.
Certainly, the record does not reflect that anyone understood that Ramos hesitated to identify Jimenez, either at the lineup or in the courtroom, because he was unsure whether Jimenez was actually Dominican. Such an inference on Detective Thompson‘s part would have been remarkable. Jimenez‘s ethnicity was not the basis for Ramos‘s identification; it bears repeating that he told Detective Thompson the day after the murder that he knew the shooter‘s identity because he recognized Monaguillo. Thus, even assuming that Detective Thompson‘s allegedly influential statement to Ramos was said at all, that it was material to the case, and that its withholding
In summary, the content of the conversation between Detective Thompson and Ramos would not have clearly signaled to either the defense or the State that Ramos was reluctant to identify Jimenez only due to his uncertainty about Jimenez‘s ethnicity. Because a fair jurist could reasonably reach the conclusion that Jimenez had not demonstrated a Brady violation, the State court‘s terse reasoning in denying that claim is not “so flawed as to undermine confidence that the constitutional claim has been fairly adjudicated.” McCray, 45 F.4th at 640 (internal quotation marks omitted), quoting Cruz, 255 F.3d at 86.
We are unpersuaded by Jimenez‘s counterarguments on appeal. He argues that the State court misapplied clearly established law by failing to recognize that Detective Thompson‘s undisclosed conversation with Ramos falls into certain broad categories of types of favorable Brady material: evidence that undermines the reliability of Ramos‘s trial identification of Jimenez, Smith, 565 U.S. at 76, and evidence that discredits the quality of the police investigation, Kyles, 514 U.S. at 441-45. But those are categories that have been found sufficiently favorable to defendants under case-specific sets of facts. See Yarborough, 541 U.S. at 664 (state courts have “more leeway” in “[a]pplying a general standard to a specific case“). On the unique facts of this case, it is certainly within the range of reasonable application of the law and facts for the State court to have concluded that Detective Thompson‘s putative incorrect statement to Ramos that Jimenez was Dominican was both incidental to the central inculpatory facts presented at trial, and not recognizable as the type of impeachment evidence prohibited by clearly established law.
Jimenez finally argues that the State court unreasonably found as a matter of fact that Ramos‘s affidavit was noncredible, because Ramos‘s multiple explanations for his identification of the killer were not incompatible with each other. The State court‘s credibility determinations on this front are not unreasonable. A fair jurist reviewing only Ramos‘s affidavit could rationally find it difficult to reconcile his statements to police in the immediate aftermath of the murder and his trial testimony about recognizing Monaguillo. As exhaustively explained, it is not obvious from the conversation recounted in the affidavit why Jimenez‘s ethnicity mattered to Ramos‘s identification, and it is doubly hard to see why it should have been obvious when all parties understood at the time that the reason Ramos identified Jimenez was because he recognized Monaguillo.40
Therefore, the State court did not unreasonably depart from the facts in the record before it by viewing the allegations in Ramos‘s affidavit as “far-fetched.” Jimenez I, 2015 WL 770457, at *8.
Under the forgiving analysis mandated by AEDPA, the State court did not unreasonably deny Jimenez‘s Brady claim as meritless.
We remain troubled by the possibility that Jimenez is innocent of the crime for which he was convicted. But given the unsettled law of the Supreme Court and the deferential review required by AEDPA, being troubled is simply not enough. Even assuming, as the Supreme Court has never held, that clear and convincing evidence of actual innocence permits a grant of habeas corpus, Jimenez has not satisfied that standard, let alone shown that the State court‘s conclusion that he failed to meet that standard was unreasonable. Accordingly, and for the reasons set forth above, we AFFIRM the district court‘s judgment denying Jimenez‘s petition for a writ of habeas corpus.
JOSÉ A. CABRANES, Circuit Judge, concurring in the judgment:
I concur only in the judgment of the Court insofar as it holds that Jimenez has not carried the burden of proving, under the standards set by the Antiterrorism and Effective Death Penalty Act of 1996,
Notes
The parties debate the reliability of her expert conclusions. Like the magistrate judge who presided over the gateway innocence hearing, we view the sociolinguist‘s contribution to the merits of this case as offering only minimal support to the fundamental issue of Jimenez‘s actual innocence claim, as her testimony bears only on the believability of Ramos‘s ability to distinguish between Spanish-speakers of different national origins by their manner of speech; it does not independently corroborate his (and Delgado‘s) specific discussion of the Dominican origin of the reported slur. Jimenez v. Lilley, No. 16cv8545(AJN)(JCF), 2017 WL 4535946, at *13 (S.D.N.Y. Oct. 10, 2017). Because we assume that Ramos‘s story is believable, we do not further address the sociolinguist‘s opinion.