Doev.Menefee

United States Court of Appeals, Second CircuitNov 19, 2004
391 F.3d 147 (2d Cir. 2004)

Cases citing this case

How cited

  • State v. Beach

    …To determine if the evidence is “reliable,” the reviewing court must analyze “whether the new evidence is…

  • State v. Beach

    …To determine if the evidence is "reliable," the reviewing court must analyze "whether the new evidence is…

lock 455 Citing caseskeyboard_arrow_right

Summaries written by judges

Summaries

  • concluding after evidentiary hearing that recantation of victim and testimony of petitioner were insufficiently reliable to meet Schlup standard

    Summary of this case from Arnold v. Dittmann

  • holding that a state collateral proceeding commenced after the statute of limitations has run does not reset the limitations period

    Summary of this case from Cabeza v. Griffin

  • holding that, even where extraordinary circumstances involve attorney incompetence, petitioner must demonstrate that he himself made reasonably diligent attempts to ensure that his petition was filed on time

    Summary of this case from Jones v. Stephens

Analyses written by attorneys

Analyses

1 Analysiskeyboard_arrow_right

No. 03-2432.

Argued: December 17, 2003.

Decided: November 19, 2004.

Appeal from the United States District Court for the Southern District of New York, Brieant, J.

Michael S. Pollok, Hoffmann Pollok, LLP, New York, NY, for plaintiff-appellant.

John J. Sergi, Assistant District Attorney (Jeanine Pirro, District Attorney of Westchester County, on the brief; Joseph M. Latino, of counsel), White Plains, NY, for defendant-appellee the Attorney General of the State of New York.

Before: MESKILL, POOLER, and SOTOMAYOR, Circuit Judges.

Judge POOLER dissents in a separate opinion.



SOTOMAYOR, Circuit Judge.

Petitioner-appellant John Doe appeals from the decision of the United States District Court for the Southern District of New York (Brieant, J.), denying his petition for a writ of habeas corpus. Although Doe seeks to challenge his 1995 New York State conviction for second-degree sodomy, he failed to file his petition within the limitations period provided by the Anti-Terrorism and Effective Death Penalty Act ("AEDPA"), 28 U.S.C. § 2244(d)(1). Doe argues that the limitations period should be tolled because he is actually innocent of the offense to which he pled guilty, because the attorney whom he hired to file his state post-conviction motion incompetently failed to file the motion in time to trigger the tolling provision provided in 28 U.S.C. § 2244(d)(2), and because the unavailability of New York State case law in the library of the federal prison in which he was incarcerated rendered him unable to file his state post-conviction motion pro se. Although the district court found that Doe was actually innocent, it declined to hold that actual innocence provides a basis for tolling AEDPA's limitations period, and it also rejected Doe's other asserted bases for tolling.

After oral argument, we requested supplemental briefing on whether "John Doe" should have been permitted to litigate under a pseudonym. The judge below granted use of the pseudonym ex parte, without giving appellees an opportunity to be heard. However, upon learning of the order, appellees failed to make a motion for reconsideration, and have not raised the issue on appeal. We decline to address the complex question of the applicable standards for litigating under a pseudonym under these circumstances.

Doe's petition challenges his state custody, resulting from his New York State conviction for second-degree sodomy, pursuant to 28 U.S.C. § 2254. The official caption reflects the fact that Doe named as one of the respondents Frederick Menefee, warden of the federal correctional institution in which Doe was incarcerated at the time that he filed this petition. The parties have agreed that Attorney General of the State of New York is the only proper respondent, however, and have litigated this matter accordingly. The district court recognized this in an order dated January 4, 2002.

On appeal, Doe argues that because he has presented a credible claim of actual innocence, the Suspension Clause, the Eighth Amendment, and due process mandate that AEDPA's limitations period be equitably tolled. With respect to the attorney competence issue, he asserts that our decision in Baldayaque v. United States, 338 F.3d 145 (2d Cir. 2003), issued after the district court's ruling on the issue, necessitates reversing the court's finding that the incompetence of Doe's attorney did not warrant tolling. Finally, Doe argues that the district court erred in holding that his claim of prison library inadequacy could not serve as a basis for equitable tolling. Because any one of these grounds is sufficient to toll the limitations period, Doe maintains, he should be given the opportunity to have his constitutional claims adjudicated on their merits.

We hold that the district court relied on clearly erroneous findings of fact in determining that Doe had presented new reliable evidence of innocence and was actually innocent. With respect to Doe's attorney incompetence argument, we affirm the court's conclusion that Doe failed to exercise reasonable diligence, and thus do not reach Doe's contention that his attorney's incompetence is sufficiently egregious to warrant tolling under Baldayaque. We also affirm the court's holding that Doe is not entitled to tolling based on purported inadequacies of the prison library because Doe's post-conviction motion relied largely on federal law.

BACKGROUND

I. Doe's Second-Degree Sodomy Conviction

The following facts are taken from the affidavits and transcripts in the record, as well as from the testimony given at the evidentiary hearing conducted before the district court in March 2003, and are essentially undisputed. Any factual disputes that have arisen during the habeas proceedings will be treated in greater detail in the discussion of the hearing testimony, below.

John Doe's second-degree sodomy conviction was the result of a long-running investigation, conducted jointly by federal and state law enforcement agencies, into a ring of pedophiles based in New York, New Jersey and Maryland. The ring was composed of men who called themselves "Boy Lovers," a phrase that Doe defined as men who liked to have sex with young boys, and its activities consisted of exchanging and creating child pornography, "seducing" and sharing victims, and soliciting child prostitution. Although the exact dates of Doe's affiliation with the group are unclear, it appears that he was friendly with its principals, and participated in its activities, beginning in the early 1980s and continuing at least through 1993. Doe's activities as a member of the group apparently never gave rise to a criminal investigation prior to 1993, although Doe did have one sex offense conviction in Maine, stemming from his 1975 attempt to pay two young boys to pose for pornographic pictures.

Doe provided a detailed description of the Boy Lovers' victims and "seduction" methods in his testimony as a federal cooperator against two fellow Boy Lovers. The men would choose victims, generally pre-teen and teenage boys from disadvantaged backgrounds, and ingratiate themselves by taking the boys to arcades and fast food restaurants. Doe stated that sometimes a boy's parents would be given financial assistance, especially if the Boy Lover thought that the parents were aware that their son was being groomed for sex. After having obtained the boy's trust, the Boy Lover would invite him to a nude beach, which would serve the dual purposes of acclimating the boy to a more erotic relationship, and allowing the pedophile to evaluate the boy's willingness to engage in eventual sexual acts. According to Doe, eventually the process would culminate in a sexual relationship, during which the Boy Lover would continue to give money to, and do favors for, the boy and his friends and family. The boy would also be coached on how to deceive the authorities, and might be used to create child pornography, which would then be shared amongst the Boy Lovers. Some boys were shared between Boy Lovers, and certain boys would be used to recruit potential new victims.

During the summer of 1993, New York and New Jersey state and federal investigators were investigating the activities of two members of the group, Richard Bagarozy and Edward Federowicz, who were suspected of creating and warehousing child pornography and transporting boys from New York to New Jersey for sex. Investigators for the Westchester County District Attorney's Office in New York ("the DA's Office") conducted several wiretaps of Bagarozy and Federowicz, and became alerted to Doe's existence when, on August 17, 1993, they overheard Bagarozy and Federowicz discussing how Doe had recently had sex with a boy named Edwin and had paid him afterwards. In later wiretaps, Doe was overheard discussing Edwin with Federowicz in a sexually explicit manner. Based on this information, the DA's Office attempted to locate Edwin in order to determine whether a crime had been committed, but did not succeed in doing so until February 1994.

The surnames of all victims who were minors at the time of the events in question have been redacted throughout this litigation in order to protect the victims' privacy.

Although these wiretaps were not introduced into evidence below, two witnesses testified as to their substance. In addition to the conversation between Bagarozy and Federovich, investigators overheard a conversation between Bagarozy and Doe, in which they discussed Edwin, and Bagarozy informed Doe that Edwin preferred Doe to fellow Boy Lover Bill O'Rourke. In subsequent conversations, Doe mentioned Edwin in the course of a sexually explicit conversation about boys' genitalia. As discussed below, the Supreme Court's actual innocence case law would allow us to consider these tapes or transcripts as substantive evidence for purposes of the actual innocence determination, even though some of the statements contained on tapes might not be admissible at a trial. See Schlup v. Delo, 513 U.S. 298, 327, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995). Because respondents refused to proffer the tapes as evidence, even after the court inquired as to their evidentiary value, we will not consider them as substantive evidence here.

Meanwhile, the FBI also had begun to investigate Doe as a member of the Boy Lovers. On September 9, 1993, the FBI and investigators from the DA's Office and other state authorities searched Doe's Westchester County apartment as part of a coordinated series of searches and arrests intended to break up the ring. The investigators recovered items of child pornography from Doe's apartment, and the United States Attorney's office for the Southern District of New York initially considered prosecuting him on child pornography charges, although it is not clear how far these deliberations progressed. Six days after the search, on September 15, Doe and his attorney, Murray Richman, met with federal investigators for the purpose of determining whether Doe had sufficient knowledge of the Boy Lovers' ring to warrant offering him a plea bargain in return for his cooperation. Doe spoke with the federal investigators pursuant to a proffer agreement that guaranteed that the United States would not use any of his statements against him in any judicial proceeding, except for purposes of cross-examination or rebuttal. During the proffer session, Doe provided a detailed description of the Boy Lovers' "seduction" methods, the names of their victims, as well as an account of his personal involvement with Bagarozy and Federowicz and his own sexual acts with children. He volunteered that "in the last couple months," he had had sex with a boy named Edwin, who was also sexually associated with Bagarozy. This would place the encounter in July or August of 1993. According to Doe, "the sex included oral copulation and masturbation."

The session was run and recorded by the federal investigators, including representatives of the FBI and the United States Attorney's Office of the Southern District of New York. Although at least one representative of the New Jersey state authorities was present, both the evidentiary hearing testimony and the documents recording the session indicate that nobody from the Westchester DA's Office participated in the session. Richman testified at an evidentiary hearing, however, that somebody from the DA's Office was present.

It is unclear whether the agreement binds any state law enforcement agency, as the agreement refers only to "this Office," a reference to the United States Attorney's Office, and the only representative of a state law enforcement agency who signed the agreement appears to have signed as a witness rather than as a party. The agreement itself indicates that the DA's Office was not a party to the agreement, however, as no representative of that office has signed the agreement in any capacity. See United States v. Pelletier, 898 F.2d 297, 302 (2d Cir. 1990) (holding that scope of use immunity provided in proffer agreement is governed by contract law principles).

Doe participated in a second proffer session in October 1993, where he provided more information about Bagarozy's and Federowicz's activities, but apparently did not discuss his own activities with Edwin. Although Doe himself was still under investigation for federal crimes at that time, at some point the federal government decided not to charge him. Nonetheless, Doe continued to work with Special Agent James Kyle of the FBI, providing information regarding the ongoing activities of the Boy Lovers and assisting the FBI in locating Bagarozy's stash of child pornography. In February 1995, Doe testified for the government in the federal trial of Bagarozy and Federowicz, where he provided extensive evidence of the Boy Lovers' activities and victims, as well as the defendants' individual proclivities and creation of child pornography.

The parties, including a representative of the Westchester DA's Office, signed a different immunity agreement during this proffer session. Because all of Doe's admissions with respect to Edwin occurred during the September 15 proffer session, however, Doe's statements during the October session are not included in the record, and the October immunity agreement is irrelevant for our purposes.

Meanwhile, investigator Pat Storino of the DA's Office was investigating the connection between Doe and Edwin on the basis of the August wiretaps. Storino located Edwin on February 3, 1994, but did not attempt to question him about his contact with Doe. Also in early February, Doe told Special Agent Kyle that Edwin had attempted to contact Doe because O'Rourke had recently gotten out of prison and had coached Edwin and another underage victim, Edward R., on how to deceive the authorities. According to Doe, Edwin wanted Doe to know that he had promised not to say anything incriminating to any law enforcement authorities. Special Agent Kyle transmitted this information to the DA's Office on February 10.

In March 1994, the DA's Office interviewed Edwin on two separate occasions. Assistant District Attorney ("ADA") Elizabeth Cronin of the Special Prosecutions Division, which handles domestic violence and child abuse cases, conducted the interviews. In accordance with the DA's Office's customary practice when dealing with underage victims, the interviews were not recorded and Edwin was not asked to sign a written statement, so Cronin's handwritten notes provide the only record of the substance of Edwin's statements. During the first interview, Cronin elicited general information about Edwin's relationship with Doe and Doe's practice of taking Edwin and his friends out for hotdogs at Nathan's, a fast food restaurant, to arcades, and back to Doe's apartment to play video games. Cronin's notes taken during the second interview, on March 18, 1994, indicate that Edwin told her that Doe had on earlier occasions attempted to "grab [Edwin's] crotch" while driving him to Nathan's, and had at least once tried to force Edwin onto the bed in Doe's apartment. On August 13, 1993, Edwin went to Doe's apartment to play computer games, and Doe tried to pull down Edwin's pants while Edwin was sitting at the computer. Doe then grabbed Edwin, took him to the bedroom, threw him down on the bed, pulled his pants down, and put his mouth around Edwin's penis. Edwin was able to shake Doe off by kicking him, and Doe later drove Edwin home and gave him $50. Edwin also provided a detailed description of Doe's apartment, which Cronin later verified, and indicated that he remembered when the alleged forcible contact had occurred because it was the Friday before he moved on August 15, 1993, which places the incident on August 13.

Ms. Cronin is now the Director of the Office of Legal Affairs in this Court.

In October 1994, Cronin notified Doe's attorney, Murray Richman, that the DA's Office intended to prosecute Doe for attempting to have sex with Edwin. Richman protested, asserting that the use immunity agreements that Doe had signed during his proffer sessions prohibited the DA's Office from prosecuting Doe for any offense that he had described during the proffer session. Cronin maintained that the September 15 use immunity agreement did not purport to bind the DA's Office, and offered to allow Doe to plead to second-degree sodomy, with a sentence of probation, rather than to a forcible sodomy charge that would mandate a prison sentence. Angry that Richman had apparently misconstrued the immunity agreement, Doe fired him and retained another lawyer, who negotiated the plea bargain.

In February 1995, Doe pled guilty in Westchester County Court to one count of second-degree sodomy. During the plea allocution, Doe was asked, "Do you admit that you, in the Village of Tarrytown . . . on or about and between July 1st of 1993 and August 13th of 1993, being 18 years of age or more, engaged in deviate sexual intercourse with a person less than 14 years of age, consisting of contact between mouth and penis?" Doe, under oath, admitted the charge. Approximately ten months after entering the guilty plea, Doe rehired Richman and attempted to withdraw his plea on the basis of an alleged jurisdictional defect, but did not assert his actual innocence. The motion was denied, and Doe was sentenced to five years' probation in January 1996. The Appellate Division affirmed the conviction in relevant part in September 1997, and denied leave to appeal on August 19, 1998. The AEDPA limitations period therefore began running on November 19, 1998, after the 90-day period to seek certiorari with the United States Supreme Court expired. See Pratt v. Greiner, 306 F.3d 1190, 1195 n. 1 (2d Cir. 2002). Doe did not immediately file a state collateral challenge to his conviction pursuant to New York Criminal Procedure Law § 440.10.

In November 1998, Doe was arrested on federal charges of conspiring to receive and possess child pornography over the internet. After it became apparent that the New York second-degree sodomy conviction would adversely affect Doe's sentencing under the United States Sentencing Guidelines, Doe retained attorney Patrick Wall in February 1999 to collaterally attack the conviction in New York Supreme Court by filing a § 440 motion. Because Doe was incarcerated in a federal facility at the time, he gave his brother, Peter Gould, a power of attorney to enable him to retain, pay and supervise Wall. Despite Doe's and Gould's supervision, Wall decided to delay filing the § 440 motion until the federal pre-sentence report was completed, at least in part because the motion might not be necessary if the pre-sentence report did not use the sodomy conviction to enhance Doe's sentencing calculation. When the AEDPA limitations period expired on November 19, 1999, Wall had not yet filed the § 440 motion, thereby losing the opportunity to invoke 28 U.S.C. § 2244(d)(2)'s provision for tolling the limitations period during the pendency of a state post-conviction motion. By February 2000, it was apparent that Wall's various health problems were preventing him from working on the motion, and Doe fired him. Doe finally filed the § 440 motion in May 2000, alleging that the DA's Office withheld exculpatory evidence, committed prosecutorial misconduct by breaching the federal use immunity agreements, and coerced his guilty plea. The motion was denied in June 2001, and the Appellate Division denied leave to appeal in November 2001.

The federal complaint alleged that Doe, by his own admission, had downloaded some 46,000 images of child pornography, digitized a video of children having sex that was filmed by his co-defendant, and possessed the equipment necessary to create, duplicate, and enhance the viewing quality of videos and CDs containing child pornography.

Doe pled guilty to the federal child pornography charges in the Eastern District of New York in April 2000. Because of his prior conviction, the district court applied a five-level enhancement, resulting in a statutory mandatory minimum sentence of sixty months.

In considering the § 440 motion, the County Court denied Doe's challenges on their merits and did not address whether the affidavit in which Edwin asserted that he never accused Doe of sexual assault raised any issue of actual innocence.

II. Doe's Federal Habeas Petition

A. Initial Proceedings

Doe filed the instant petition on November 30, 2001, just over two years after AEDPA's limitations period had run. He conceded that his petition was untimely, but raised the three tolling arguments at issue here, although in a less developed form. With respect to the actual innocence issue, Doe proffered an affidavit in which Edwin stated that Doe never molested him and that he never told the DA's Office that Doe had attempted to have any sexual contact with him. The district court found, however, in an order dated February 2002, that a claim of actual innocence did not provide a basis for tolling AEDPA's limitations period, noting "the clear intent of AEDPA to bring repose to state convictions in the federal courts, whatever the social cost." The court also denied Doe's tolling claim based on attorney incompetence, finding that our decision in Smaldone v. Senkowski, 273 F.3d 133 (2d Cir. 2001) (per curiam), mandated the conclusion that Wall's lack of diligence was insufficient to constitute "extraordinary circumstances" meriting equitable tolling. Finally, the court implicitly held that Doe could have filed his petition pro se despite his incarceration.

On appeal, we declined to review the merits of the district court's rulings. We stated that this Court has "not yet determined whether the Constitution requires an actual innocence exception to [AEDPA's] statute of limitations," but that we would decide this issue in the context of a case in which the petitioner's credible showing of actual innocence "compels us to do so." Doe v. Menefee, 49 Fed.Appx. 340, 341-42 (2d Cir. 2002) (unpublished opinion) ("Doe I") (internal quotation marks and citations omitted). Noting that "Doe has presented evidence, the affidavit of his alleged victim, which might well lead a reasonable juror to find him not guilty," but that "the respondent also has presented extensive evidence undermining the victim's affidavit," we remanded the case to the district court to determine whether "Doe has presented a credible claim of actual innocence." Id. at 342. We also directed the court to consider whether a petitioner must demonstrate that he pursued his claim of actual innocence with reasonable diligence, as is required in asserting other grounds for equitable tolling. Id. Finally, we declined to consider Doe's arguments with respect to attorney incompetence and the inadequacy of the prison library. Id. B. The Evidentiary Hearing on Remand

On remand, the district court conducted a hearing on the actual innocence issue, after which it determined that Doe had "established his innocence by a preponderance of the credible evidence." The court heard testimony from Doe, Edwin, Richman, Cronin, Storino, and ADA Greg Bernhard on the issue of actual innocence, and from Peter Gould on the issue of Doe's reasonable diligence in pursuing his claim. The testimony that is relevant to the disposition of this appeal is summarized below.

1. Doe

Doe testified that he met Edwin through Edward R., one of the boys most closely connected with Bagarozy, early in the summer of 1993. Doe assertedly did not realize at the time, however, that Edward R. was sleeping with Bagarozy, and was used by Bagarozy and Federowicz to identify and attract new victims. Although Doe was not very close to Bagarozy and Federowicz and was not aware of the bulk of their activities, he associated with many of the same boys as did the two convicted pedophiles. Doe attributed this overlap in association to the fact that he routinely made himself available to underprivileged boys like Edward R. and Edwin, taking them to arcades, to Nathan's for hotdogs, and occasionally back to his apartment to watch movies and play video games.

On August 13, 1993, the date of the alleged assault, Doe picked up Edwin from the Bronx and drove him, alone, to Doe's apartment in Tarrytown. Doe had just finished work, and wanted to change into casual clothes before he took Edwin to Sportstime, an arcade. At Doe's apartment, Edwin waited while Doe changed, and the two were alone in the apartment for perhaps ten minutes. Doe denied having sex with Edwin or having attempted to force him into sexual contact.

Doe maintained that even after August 13, Edwin continued to call him and ask to be taken out. They appear to have met just once, though, in July 1994, when Doe, O'Rourke, and Edwin went to Playland Park in Westchester County. Doe stated that Edwin told him that the DA's Office had approached him with questions about Doe, but that Edwin had denied any improprieties.

When questioned as to why he pled guilty to second-degree sodomy when he was innocent of the offense, Doe asserted that he felt coerced into pleading because he was frightened that he would be charged with a more serious offense — and face prison time — if he did not plead. Moreover, Doe was anxious to help the federal government convict Bagarozy and Federowicz, because he had been completely unaware of their more unsavory activities — including having sex with Edward R. and others, and using the boys to make child pornography — and felt betrayed because he himself would not have countenanced such exploitation. Doe believed that if he did not plead guilty to the sodomy offense, the federal prosecutors running the Bagarozy/Federowicz trial would be angry, and that the possibility that he would have to invoke the Fifth Amendment while on the witness stand could jeopardize his ability to function as a cooperator.

Doe also attempted to explain his admission at the September 1993 proffer session that he had had sex with Edwin by stating that he had "panicked" when asked about Edwin. He stated that the proffer was held "in a small, airless, windowless room" "at the end of the summer . . . [and] it was extremely hot." Doe was worried that he "was not going to be able to explain to these agents . . . what [his] role in [the Boy Lovers] was, which was not a criminal role," and he therefore "lied" and stated that he had had sex with Edwin.

2. Edwin

Edwin's testimony roughly corroborated Doe's, although the two accounts vary in some respects. Edwin stated that he was never alone with Doe in Doe's apartment, and had never had sexual contact, consensual or forcible, with Doe.

Edwin was approached by the DA's Office in early 1994, and went to one meeting with ADA Cronin. He recalled that Cronin was the only person involved in the interview until a man came in towards the end, but he admitted that his recollection of the particulars of the interviews was unclear. Cronin asked Edwin if Doe ever "gave [him] money for sexual favors or anything like that," and Edwin replied that no sexual contact had ever occurred.

Edwin's affidavit states that Edwin attended two meetings.

With respect to his larger involvement with the circle of Boy Lovers, Edwin testified that he played on a basketball team organized by Bagarozy, and knew of O'Rourke through Bagarozy but had never met him. He further testified that Bagarozy invited him to a nude beach once, but he refused. None of the men ever gave Edwin money or other financial assistance.

Edwin learned that Doe had been convicted of an offense involving the August 1993 incident when Edwin was incarcerated at Rikers Island in 1998. There, Edwin was approached by Edward R.'s brother, Angel R., who had been closely associated with O'Rourke and was known as one of the Boy Lovers' main boys. Angel R. told Edwin that Doe was "in jail because of" Edwin, leading Edwin to attempt to contact Doe. Doe refused to speak with Edwin because of an order of protection arising from the 1995 conviction, but soon after Edwin attempted to initiate contact, Doe sent a private investigator to speak to Edwin. Doe's attorney then generated an affidavit that Edwin signed, asserting that he had never told the DA's Office that Doe assaulted him. Doe also retained an attorney for Edwin, ostensibly to protect Edwin against "pressure" from the DA's Office. Finally, Edwin maintained that he was not promised anything in exchange for his testimony at the hearing; he was willing to testify because "the truth needs to be told."

3. Elizabeth Cronin

Cronin recounted her interviews with Edwin after refreshing her recollection with the notes that she took during the interviews. Cronin testified that of seven pages of notes, the first four, dated March 18, 1994, were taken during her second interview with Edwin. The next two pages, undated, were taken during the first interview, and the last page of notes may have been taken later, when Cronin was preparing for a proceeding. The notes were then offered into evidence, and are included in the record.

According to Cronin, the DA's Office conducted two interviews with Edwin because it was standard practice to conduct at least two interviews when dealing with a child victim. ADA Greg Bernhard, who had handled the investigation until the DA's Office decided to transfer the prosecution to the Special Prosecutions Division, was also present for both interviews. During the first interview, Cronin elicited general information about Edwin's relationship with Doe and how the two had met. The second interview covered the alleged sexual contact and its aftermath.

Cronin stated that Edwin had met Doe through Edward R. in early summer, and that Doe would drive Edwin to various places. Sometimes Doe would attempt to touch Edwin's groin in the car. At some point before the alleged contact that formed the basis of the sodomy charge, Edwin, unaccompanied by any other boys, was playing on the computer in a back room in Doe's apartment when Doe came up behind him and attempted to put his hands down Edwin's pants. Doe told Edwin to stand up, and when Edwin struggled, Doe carried Edwin into the bedroom. When Edwin continued to struggle, Doe became angry and threatened not to drive Edwin home.

On August 13, 1993, Edwin again accompanied Doe back to his apartment. Doe's advances began as they had on the earlier occasion but went further this time; once Doe had maneuvered Edwin onto the bed, he pulled Edwin's pants down and put his mouth around Edwin's penis. Edwin pushed Doe's hands away and kicked him, and Doe retreated. Doe later accused Edwin of having provoked Doe into coming on to him, and gave Edwin $50 or $60. Cronin asked Edwin why he had accompanied Doe to his apartment after the first incident, which itself had involved an element of compulsion, and Edwin answered that he had thought that because he had made it clear to Doe that he did not like the advances, Doe would not come on to him again.

In accordance with standard practice, Cronin also asked Edwin to describe Doe's apartment in detail, in order to verify that Edwin had actually been there. Cronin was later able to verify Edwin's description with Storino's and Bernhard's knowledge of the apartment, based on their search of the property.

Based on the interviews with Edwin, the DA's Office decided to charge Doe with a sex offense, and Cronin contacted Doe's attorney, Murray Richman, to discuss the charges. Cronin told Richman that the DA's Office believed it had sufficient evidence to charge Doe with forcible sodomy, a Class B felony that would carry a sentence of prison time, but that because Doe was cooperating with the federal government in the Bagarozy and Federowicz trial, the DA's Office was willing to consider charging Doe with second-degree sodomy, a Class D felony that would probably result in a sentence of probation.

4. Greg Bernhard

ADA Bernhard, who supervised the wiretaps and the search for Edwin and was present at Cronin's interviews, corroborated Cronin's testimony in its entirety. Bernhard testified that he had not reviewed Cronin's interview notes in preparation for his testimony, because he remembered the interviews "vividly." Bernhard had not expected Edwin to describe the sexual contact as forcible, because the wiretapped conversations had created the impression that because Doe paid him, Edwin was a willing participant in the encounter.

When asked to describe what allegedly occurred during Edwin's second visit to Doe's apartment, Bernhard stated that Edwin told him and Cronin that he was playing computer games and that, while he was on the computer, Doe began grabbing at his legs and trying to turn him from the computer. Eventually Doe pulled Edwin off of the chair, threw him onto the bed, pulled his pants down, and put his mouth on Edwin's penis. Edwin kicked Doe and told him to stop.

Bernhard was asked only about Edwin's second visit to the apartment, from which the sodomy charge arose. The first visit, presumably, was the one described in Cronin's notes, in which Doe attempted to molest Edwin, but did not succeed in having any sexual contact with him.

5. Investigator Pat Storino

Finally, Storino testified as to his investigation into Edwin's whereabouts and the process by which the DA's Office learned of Edwin and the potential crime. Although the wiretaps were joint federal-state endeavors, no representative of the DA's Office was present at the September 15 proffer session in which Doe admitted to having sex with Edwin. The DA's Office's investigation into the incident was therefore conducted, at least at first, without knowledge of Doe's admission that he had sex with Edwin. Storino also testified that Special Agent Kyle contacted him in February and relayed Doe's statements that Edwin had been coached on what to say to the authorities, and had promised not to divulge any incriminating information.

C. The District Court's Findings and Conclusions

Following the hearing, the district court issued an order detailing its findings of fact and conclusions of law. The court analyzed the testimony in light of its conception of the standard for demonstrating actual innocence, stating that "the standard of proof is at the lowest possible level. Innocence need only be demonstrated to be more likely than unlikely."

The district court found Edwin's testimony credible in its entirety, as Edwin was "forthright . . . and responsive" in answering questions. The court concluded that although Edwin had "socialized" with Doe and other Boy Lovers, he "had not personally engaged in any sexual activity with Doe or any other man." Because crediting Edwin's accounts of what he had said during his interviews with Cronin required that the court discount or discredit Cronin's and Bernhard's directly conflicting testimony, the court concluded that Cronin's recollection was based primarily on her notes, "at least some of which are not contemporaneous records." Thus, although the court explicitly refused to make an adverse credibility determination with respect to Cronin, it found that Edwin's testimony was entitled to more weight. The court did not analyze or acknowledge the implications of Bernhard's testimony, however.

Turning to Doe's testimony, the court credited Doe's account of his relationship with Edwin and his explanations of his previous inconsistent statements. The court placed weight on the fact that although Edwin stated that he was never alone with Doe in his apartment, Doe admitted to having been alone with Edwin once, an inconsistency that the court believed suggested Doe's forthrightness. The court also concluded that Doe had pled guilty because he was genuinely afraid of receiving jail time if he did not plead, and that his sense of betrayal arising out of Bagarozy's and Federowicz's activities made him particularly reluctant to endanger his ability to testify against them by refusing to plead guilty. The court also found Doe's discharge of Richman just before he pled guilty consistent with Doe's innocence, because Doe may have been loath to accept Richman's advice that he plead. This reasoning in turn suggested that Doe's admission of guilt during his plea allocution was false. Finally, the court found Doe's statement that he "panicked" at the proffer session and felt compelled to lie about having sex with Edwin "somewhat strained," but apparently did not feel that Doe's 1993 admission threw the veracity of his hearing testimony into doubt.

The court therefore concluded that Doe had established his innocence by a preponderance of the evidence. It next determined that Doe had exercised reasonable diligence in pursuing the claim of actual innocence, but reiterated its initial conclusion that AEDPA does not allow tolling for actual innocence, and the Suspension Clause does not require such an exception to AEDPA's limitations period.

DISCUSSION

Doe challenges the district court's ruling that AEDPA's limitations period cannot be tolled based on a credible claim of actual innocence as well as its February 2002 rulings that Doe's claims of attorney incompetence and prison library inadequacy were insufficient to toll the limitations period. Doe argues that each of these grounds entitles him to equitable tolling, and that he therefore should have the opportunity to have his constitutional claims adjudicated on the merits.

Even if Doe were to prevail on his argument that he is entitled to tolling based on actual innocence, however, it is not clear that such tolling would in itself be sufficient to render his petition timely. Doe never clearly established, and the district court made no finding with respect to, when Doe first learned of the new evidence that purportedly demonstrates his actual innocence. There is some indication in the record that Doe first learned of and investigated his alleged victim's willingness to testify that no sexual assault occurred as early as 1998 or 1999, well before the limitations period expired. If this were the case, any tolling that might arise from Doe's belated discovery of the new evidence might not in itself render his petition timely. We note, however, that the question of how tolling for actual innocence will be calculated need not be decided until we determine whether the limitations period may be tolled for actual innocence.

Doe also argues that there is sufficient evidence in the record to enable us to reach the merits and grant his habeas petition. In light of our disposition of the tolling issues, we need not reach this argument.

Although AEDPA does not provide that its limitations period may be tolled for any reason other than the pendency of a state post-conviction motion, see 28 U.S.C. § 2244(d)(2), in "rare and exceptional circumstances" a petitioner may invoke the courts' power to equitably toll the limitations period. Smith v. McGinnis, 208 F.3d 13, 17 (2d Cir. 2000) (per curiam) (internal quotation marks and citations omitted). To qualify for such treatment, the petitioner must establish that "extraordinary circumstances prevented him from filing his petition on time," and that he "acted with reasonable diligence throughout the period he seeks to toll." Id. We have established only a limited number of circumstances that may merit equitable tolling, such as where an attorney's conduct is so outrageous and incompetent that it is truly extraordinary, see Baldayaque, 338 F.3d at 152, and where prison officials intentionally obstruct a petitioner's ability to file his petition by confiscating his legal papers, see Valverde v. Stinson, 224 F.3d 129, 133-34 (2d Cir. 2000). We have specifically reserved the question of whether a claim of actual innocence based on newly discovered evidence constitutes an extraordinary circumstance that merits equitable tolling, however, as well as the question of whether the Constitution would require equitable tolling for actual innocence. See Lucidore v. New York State Div. of Parole, 209 F.3d 107, 114 (2d Cir. 2000).

With respect to Doe's claim of actual innocence, we hold that the district court misapprehended the standard for determining actual innocence as it is delineated in governing Supreme Court case law, and that its findings with respect to the credibility of Doe's and Edwin's hearing testimony were clearly erroneous. In adherence to our repeated statements that we will decide whether equitable tolling is available on the basis of actual innocence only in a case in which the petitioner has made a credible claim of actual innocence, therefore, we do not reach Doe's arguments that various constitutional provisions require that AEDPA's limitations period be tolled for actual innocence. With respect to the attorney incompetence issue, we agree with the district court's conclusion that Doe did not exercise reasonable diligence during the period that he seeks to have tolled. Finally, we affirm the district court's ruling that the purported inadequacies of the library of the prison in which Doe was incarcerated do not provide a basis for tolling the limitations period, as Doe has not established that the library lacked adequate resources for him to prepare his state post-conviction motion.

I. Tolling for Actual Innocence

The doctrine of actual innocence was developed to mitigate the potential harshness of the judicial limitations placed on a petitioner's ability to file successive or otherwise procedurally defaulted habeas petitions in the federal courts. See Schlup v. Delo, 513 U.S. 298, 318-21, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995) (describing development of limitations on the writ of habeas corpus and establishing contours of actual innocence exception). Prior to AEDPA's passage, procedural limitations on the right to seek habeas relief were judicially created to prevent the erosion of the values of comity and finality, which were threatened by repeated federal examination of state convictions. These procedural limitations were crafted and applied as equitable limitations that could yield in the face of countervailing considerations. See id. at 320-21, 115 S.Ct. 851. Thus, where a petitioner appears to be the "victim of a fundamental miscarriage of justice," the principles of comity and finality must yield to the imperative of federal judicial review of the conviction. See Murray v. Carrier, 477 U.S. 478, 495-96, 106 S.Ct. 2639, 91 L.Ed.2d 397 (1986). Accordingly, in a series of decisions, the Supreme Court established that prisoners may obtain federal habeas review, despite their state procedural default or previous federal habeas petitions, if they can demonstrate cause for the default and resulting prejudice. See, e.g., McCleskey v. Zant, 499 U.S. 467, 494, 111 S.Ct. 1454, 113 L.Ed.2d 517 (1991) (establishing cause and prejudice standard in cases where petitioner is accused of abusing the writ); Francis v. Henderson, 425 U.S. 536, 542, 96 S.Ct. 1708, 48 L.Ed.2d 149 (1976) (establishing cause and prejudice standard for federal review following state procedural default).

An independent category of cases in which petitioners may suffer miscarriages of justice if they are procedurally barred from filing habeas petitions is composed of those cases in which the petitioners claim that they are actually innocent of the crimes for which they were convicted. Thus, "in an extraordinary case, where a constitutional violation has probably resulted in the conviction of one who is actually innocent, a federal habeas court may grant the writ even in the absence of a showing of cause for the procedural default." Carrier, 477 U.S. at 496, 106 S.Ct. 2639. Because credible claims of actual innocence are "extremely rare," federal court adjudication of constitutional challenges by petitioners who may be actually innocent prevents miscarriages of justice, but does not threaten state interests in finality. Schlup, 513 U.S. at 321-22, 115 S.Ct. 851. Accordingly, a petitioner may use his claim of actual innocence as a "gateway," or a means of excusing his procedural default, that enables him to obtain review of his constitutional challenges to his conviction. See id. at 315-17, 115 S.Ct. 851. This remains the case after AEDPA, which codifies the actual innocence doctrine, with modifications not relevant here, as it applies to successive petitions and state procedural defaults. See 28 U.S.C. §§ 2244(b); 2254(e)(2).

Because AEDPA's one-year limitations period protects the same values of comity and finality as do the procedural limits on successive and defaulted petitions, see Acosta v. Artuz, 221 F.3d 117, 122-23 (2d Cir. 2000), it is not surprising that several habeas petitioners have attempted to argue that considerations of justice mandate adopting the actual innocence gateway as a means of tolling AEDPA's limitations period. See, e.g., Whitley v. Senkowski, 317 F.3d 223, 225-26 (2d Cir. 2003). We have not yet decided this question, reasoning that we should decide whether the Constitution requires tolling for innocence only in a case in which the petitioner can show that, because he can demonstrate his actual innocence, he would be injured if not entitled to tolling on this basis. See Lucidore, 209 F.3d at 113-14. Thus, we have instructed district courts faced with untimely petitions in which the petitioner asserts his or her actual innocence to determine, in each case, whether the petitioner has presented a credible claim of actual innocence before ruling on the legal issues of whether such a showing provides a basis for equitable tolling and whether the petitioner must also demonstrate that he or she pursued his or her claim with reasonable diligence. See Whitley, 317 F.3d at 225.

Because the interests that must be balanced in creating an exception to the statute of limitations are identical to those implicated in the procedural default context, we see no reason not to apply the Schlup standard in the tolling context. Indeed, we have previously assumed that if actual innocence does provide a basis for tolling the limitations period, Schlup's delineation of the evidentiary showing necessary to demonstrate actual innocence would apply in evaluating whether the petitioner had made a credible showing of actual innocence. See Lucidore, 209 F.3d at 114 (applying the Schlup standard and holding that petitioner had not demonstrated actual innocence).

The Schlup Court carefully limited the type of evidence on which an actual innocence claim may be based and crafted a demanding standard that petitioners must meet in order to take advantage of the gateway. The petitioner must support his claim "with new reliable evidence — whether it be exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physical evidence — that was not presented at trial." Schlup, 513 U.S. at 324, 115 S.Ct. 851. Because Schlup explicitly states that the proffered evidence must be reliable, the habeas court must determine whether the new evidence is trustworthy by considering it both on its own merits and, where appropriate, in light of the pre-existing evidence in the record. See id. at 327-28, 115 S.Ct. 851.

Once it has been determined that the new evidence is reliable, Schlup unequivocally requires that reviewing courts consider a petitioner's claim in light of the evidence in the record as a whole, including evidence that might have been inadmissible at trial:

[This] standard is intended to focus the inquiry on actual innocence. In assessing the adequacy of petitioner's showing . . . the district court is not bound by the rules of admissibility that would govern at trial. Instead, the emphasis on "actual innocence" allows the reviewing tribunal also to consider the probative force of relevant evidence that was excluded or unavailable. . . . [W]e believe that Judge Friendly's description of the inquiry is appropriate: The habeas court must make its determination concerning the petitioner's innocence "in light of all the evidence, including that alleged to have been illegally admitted (but with due regard to any unreliability of it) and evidence tenably claimed to have been wrongly excluded or to have become available only after the trial."

Id. (quoting Henry Friendly, Is Innocence Irrelevant? Collateral Attack on Criminal Judgments, 38 U. Chi. L.Rev. 142, 160 (1970)). Our dissenting colleague asserts that under Bousley v. United States, 523 U.S. 614, 118 S.Ct. 1604, 140 L.Ed.2d 828 (1998), a court may consider only admissible evidence when assessing the reliability of a claim of actual innocence. We disagree. As Schlup makes clear, the issue before such a court is not legal innocence but factual innocence. Integral to Schlup's holding is an express command that reviewing courts consider all evidence without regard to its admissibility, while the portion of Bousley relied upon by the dissent is mere dictum.

The Bousley petitioner had procedurally defaulted his habeas claim by failing to challenge the validity of his plea. 523 U.S. at 621, 118 S.Ct. 1604. The Court held that he had neither satisfied the exception to the procedural default rule nor overcome the rule by showing cause or prejudice. Id. at 622-23, 118 S.Ct. 1604. In dicta, the Court suggested that petitioner's habeas claim might survive on remand if he could establish a claim of actual innocence. Id. at 623, 118 S.Ct. 1604. After a terse summary of the actual innocence inquiry, the Court noted that "the Government is not limited to the existing record to rebut any showing that petitioner might make. Rather, on remand, the Government should be permitted to present any admissible evidence of petitioner's guilt even if that evidence was not presented during petitioner's plea colloquy." 523 U.S. at 624, 118 S.Ct. 1604. The dissent argues that Bousley has therefore "clarified the type of evidence that can be considered in assessing actual innocence." However, we decline to treat this dictum as overruling Schlup's square holding that reviewing courts must evaluate all the evidence. Schlup, 513 U.S. at 327-28, 115 S.Ct. 851. We also observe that Bousley's summary of the actual innocence inquiry cites to that precise portion of Schlup discussing the importance of reviewing courts' consideration of both admissible and inadmissible evidence. Bousley, 523 U.S. at 623, 118 S.Ct. 1604 (citing Schlup, 513 U.S. at 327-28, 115 S.Ct. 851). These conflicting signals — Bousley's favorable citation to this fraction of Schlup, juxtaposed against the single word of dictum stressed by the dissent — reinforce our refusal to interpret Bousley as overruling Schlup.

Only after examining all evidence is the court able to determine whether new evidence truly throws the petitioner's conviction into doubt, or whether it is so overwhelmed by the weight of other evidence that it is insufficient to raise a question as to a petitioner's factual innocence. If the court then concludes that, in light of all the evidence, it is "more likely than not that no reasonable juror would have found petitioner guilty beyond a reasonable doubt," a petitioner may invoke the actual innocence gateway and obtain review of the merits of his claims. Schlup, 513 U.S. at 327, 115 S.Ct. 851.

The actual innocence inquiry therefore has a hybrid structure: On the one hand, the court must make its own evaluation of the evidence, free of the strictures of the rules governing a trial, but on the other hand, the standard according to which the court must evaluate that evidence requires the court to determine the verdict that reasonable jurors would probably reach. Because in rare cases a finding of guilt beyond a reasonable doubt may not be synonymous with factual guilt, Schlup recognizes that the reviewing court must do more than reenact a trial of the petitioner; it must be free to evaluate independently all of the evidence, old and new, to determine whether that evidence may show that the petitioner is factually innocent. Because our legal system has no means of defining innocence independently of the finding of reasonable doubt, however, "the analysis must incorporate the understanding that proof beyond a reasonable doubt marks the legal boundary between guilt and innocence." Schlup, 513 U.S. at 328, 115 S.Ct. 851. This two-step process allows the court to analyze the petitioner's potential innocence in light of the fact that the petitioner is essentially claiming that the criminal justice process has reached the wrong factual result, whether after a trial or a guilty plea.

Thus, taking into account "the probative force of relevant evidence," the court must consider how reasonable jurors, fairly examining all of the evidence presented, would assess the petitioner's guilt or innocence. Id. at 327-29, 115 S.Ct. 851. In evaluating the record as a whole, the habeas court may make its own credibility determinations as to both the new evidence and the evidence already in the record that may be thrown into doubt by the new material. See id. at 330, 115 S.Ct. 851. With respect to the ultimate issue of innocence, however, Schlup does not permit the court to make an independent judgment of whether reasonable doubt exists, but instead requires a "probabilistic determination about what reasonable, properly instructed jurors would do." Id. at 329, 115 S.Ct. 851. This probabilistic analysis must determine "not merely [whether] . . . a reasonable doubt exists in the light of the new evidence," but rather whether it is more likely than not "that no reasonable juror would have found the defendant guilty." Id.; see also id. at 333, 115 S.Ct. 851 (O'Connor, J., concurring) ("[A] petitioner does not pass through the [innocence] gateway . . . if the district court believes it more likely than not that there is any juror who, acting reasonably, would have found the petitioner guilty beyond a reasonable doubt." (emphasis added; internal citation omitted)).

Because the determination as to whether no reasonable juror would find a petitioner guilty beyond a reasonable doubt is a mixed question of law and fact, we review the district court's ultimate finding of actual innocence de novo. See United States ex rel. Bell v. Pierson, 267 F.3d 544, 551-52 (7th Cir. 2001) (noting that district court must make factual findings with respect to new evidence, but concluding that district court is no better placed than appellate court to make probabilistic determination as to what reasonable juror would find and concluding that review is therefore de novo); cf. Pham v. United States, 317 F.3d 178, 182 (2d Cir. 2003) (stating that mixed questions of law and fact presented in a habeas petition are reviewed de novo).

We review the district court's underlying findings of fact under the more deferential clearly erroneous standard. See Fed.R.Civ.P. 52(a); Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985). Clear error review permits only limited reexamination of factual findings where "the district court's account of the evidence is plausible in light of the record viewed in its entirety." Anderson, 470 U.S. at 573-74, 105 S.Ct. 1504. The reviewing court may reverse, however, when, although there is evidence to support the finding, "on the entire evidence [the court] is left with the definite and firm conviction that a mistake has been committed." Id. at 573, 105 S.Ct. 1504 (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 92 L.Ed. 746 (1948)).

By the same token, although Rule 52(a) "demands even greater deference to the trial court's" credibility determinations, the district court may not "insulate [its] findings from review by denominating them credibility determinations, for factors other than demeanor and inflection go into the decision whether or not to believe a witness." Id. at 575, 105 S.Ct. 1504. Thus, reviewing for clear error allows an appellate court to examine the district court's credibility determinations in light of the evidence in the record as a whole, in order to determine whether the credibility assessment can be reconciled with other evidence: "Documents or objective evidence may contradict the witness' story; or the story itself may be so internally inconsistent or implausible on its face that a reasonable factfinder would not credit it." Id. Where, however, two witnesses have each told a "coherent and facially plausible story that is not contradicted by extrinsic evidence," the trial court's decision to credit one witness over the other, "if not internally inconsistent, can virtually never be clear error." Id. at 575-76, 105 S.Ct. 1504.

We have found a district court's factual findings to be clearly erroneous where the court has failed to synthesize the evidence in a manner that accounts for conflicting evidence or the gaps in a party's evidentiary presentation. Thus, we have reversed factual findings where the trial court incorrectly assessed the probative value of various pieces of evidence, leading it to rely on speculation, see United States v. Rizzo, 349 F.3d 94, 100-02 (2d Cir. 2003), and where the court failed to weigh all of the relevant evidence before making its factual findings, see Ortega v. Duncan, 333 F.3d 102, 107 (2d Cir. 2003). Moreover, we have found the trial court's credibility determinations vulnerable where they were founded on factual inferences that the evidence did not permit. See Krizek v. Cigna Group Ins., 345 F.3d 91, 100-01 (2d Cir. 2003). Accord Jiminez v. Mary Washington College, 57 F.3d 369, 380-81 (4th Cir. 1995) (reversing district court's discrediting of five semesters of students' assertions against a professor denied tenure as tainted by racial animus and finding that the district court's conclusion of taint was based on speculation and was inconsistent with the other evidence in the record); Hayes v. Invesco, Inc., 907 F.2d 853, 856-57 (8th Cir. 1990) (holding that fact findings were clearly erroneous where the district court misinterpreted some evidence and failed to consider some testimony).

A. The District Court's Factual Findings

We hold that the district court's finding that the hearing testimony of Doe and Edwin constitutes reliable evidence of Doe's actual innocence was clearly erroneous. While we recognize that district courts are generally best placed to evaluate testimony in light of the witnesses' demeanor, credibility determinations are not composed of demeanor evaluations alone. Even where a court finds that a witness appears to be telling the truth, it must, as Anderson recognizes, evaluate the testimony in light of the substance of other evidence, considering the potential motives to be untruthful that the witness may possess, corroboration or the lack thereof, internal consistency, and the inferences or assumptions that crediting particular testimony would require. See Anderson, 470 U.S. at 575, 105 S.Ct. 1504.

This is particularly true in the context of an actual innocence determination, as Schlup requires the habeas court to determine whether the new evidence on which the actual innocence claim is based is reliable. See Schlup, 513 U.S. at 324, 115 S.Ct. 851. In order to make this assessment where, as here, the new evidence consists entirely of testimony that challenges the facts on which the prosecution relied in obtaining the conviction, the court must carefully consider the nature of the testimony in light of the existing record to determine whether it can be considered reliable. See id. at 327-38, 115 S.Ct. 851; Anderson, 470 U.S. at 575, 105 S.Ct. 1504. The court's conclusion that it believes a witness's testimony at an evidentiary hearing is only one element of the determination that the testimony constitutes new reliable evidence. The court must then evaluate whether its subjective impression of the testimony can be sustained in light of the record as a whole.

Having reviewed the testimony of Doe and Edwin in light of Doe's prior statements and admissions of guilt and the other evidence in the record, we are left with the "definite and firm conviction" that the district court committed clear error in crediting the testimony. See Anderson, 470 U.S. at 573, 105 S.Ct. 1504. We therefore find that neither Doe's nor Edwin's testimony constitutes new reliable evidence.

1. Doe

Doe testified that he never attempted to have any sexual contact with Edwin and that his relationship with Edwin consisted solely of taking him to arcades and fast food restaurants at Edwin's behest. The district court's reasons for finding that Doe's testimony had the "ring of truth" and crediting it in its entirety were deeply flawed. We therefore find that Doe's testimony does not constitute reliable evidence of his innocence.

As an initial matter, the district court relied on a misinterpretation of the evidence in concluding that the truthfulness of Doe's hearing testimony was buttressed by his conduct around the time of his guilty plea. The court found that just before Doe pled guilty, he fired his lawyer, Murray Richman, because Richman suggested that Doe plead guilty despite his innocence. The court then inferred that Richman's firing was consistent with Doe's claim of innocence. Both Doe and Richman testified unequivocally, however, that Doe fired Richman because Richman had incorrectly advised Doe that his 1993 use immunity agreement would prevent all state authorities, including the DA's Office, from prosecuting him. The district court acknowledged and credited Doe's testimony on this point, rendering its fact findings internally inconsistent.

A review of the record leads to the conclusion that none of Doe's conduct during the investigation and around the time of his guilty plea is sufficient to permit the inference that Doe was actually innocent. The district court therefore should have determined whether it could credit Doe's recent protestations of innocence, without more, over his admission during the 1993 federal proffer session that he had sex with Edwin. This analysis requires an examination of the substance of both statements, as well as Doe's motives during each proceeding, and the persuasiveness of his current explanation of his 1993 admission.

The use immunity provided to Doe as an incident of his participation in the September 1993 proffer session does not preclude us from considering the substance of Doe's admissions as evidence against him. Notably, Doe himself does not argue that we may not consider the admissions. Although Richman testified that a representative of the DA's Office was present at the proffer session, the agreement itself is not signed by any representative of the DA's Office, and does not purport to preclude that Office from arguing, as it does, that Doe's admissions during the session should be used against him. See Pelletier, 898 F.2d at 302. Moreover, even assuming that the DA's Office was somehow bound by the immunity agreement, the agreement itself allows Doe's statements to be used to rebut any evidence that Doe might offer.

In the federal proffer session, Doe gave a detailed account of the activities of fellow Boy Lovers O'Rourke, Bagarozy, and Federowicz, as well as their various "main squeezes" and shared victims. In this context, Doe stated that he had had sex with Edwin, who was also one of O'Rourke's boys, and that the sex "included oral copulation and masturbation" and took place at Doe's apartment in July or August of 1993. He later testified under oath at the Bagarozy trial that all of his statements at the proffer session were true. In contrast, Doe asserted at the hearing that these statements were false, because he had "panicked" when asked about Edwin during the proffer session and had felt compelled to state, untruthfully, that he had had sex with Edwin. The district court found this explanation "somewhat strained," an assessment with which we agree. Prior to being asked about Edwin, Doe detailed his fantasies about various boys in the choir that he led, but was capable of denying attempting to seduce or molest them. Doe also described traveling to Maryland to have sex with young boys and provided graphic details about various sexual encounters with boys, but denied participating, along with other Boy Lovers, in the filming of certain boys. Thus, it is evident that Doe admitted extensive criminal conduct but, under the same or similar circumstances, was completely capable of denying certain other criminal acts, and did not feel compelled to implicate himself in a wider range of misconduct.

Edwin's hearing testimony that he never had sex with Doe also does not provide a basis for discounting Doe's 1993 admission of guilt. Because Doe does not dispute that he admitted in 1993 having sex with Edwin, Edwin's testimony that no sex act took place does not explain why Doe would admit the sex act to the authorities even though it had assertedly not actually occurred. For the same reason, Edwin's testimony does not make Doe's explanation that he panicked at the proffer session any more persuasive. Moreover, the potential for Edwin's testimony to corroborate Doe's hearing testimony and render Doe's prior admission of guilt less likely to be true depends on the credibility of Edwin's hearing testimony, and as is discussed below, Edwin's testimony on this issue is not reliable.

Doe had little motive to lie about his contact with Edwin in the proffer session, because the focus of the session was on Doe's knowledge of the other Boy Lovers' activities, and, at the time, Doe was facing a potential prosecution on federal child pornography charges, but not, as far as he knew, on sexual conduct charges. Doe's own conduct with Edwin or any other boys was at most an ancillary issue that, given the wealth of information that he supplied about several other Boy Lovers during the session, would have had little bearing on the authorities' impression of Doe's overall knowledge and potential usefulness as a cooperator. Doe therefore had no motive to attempt to incriminate himself by asserting falsely that he had had sex with Edwin. Indeed, given that the use immunity provided as an incident to the proffer session would not prevent the authorities from using Doe's statements against him in cross-examination or substantively to rebut his testimony, Doe had no reason to exaggerate his personal criminal conduct.

In contrast, Doe had ample motive to lie at his habeas hearing. Doe failed to file a state collateral challenge to his state conviction or assert his innocence until it appeared that the conviction would adversely affect his federal sentencing. Now he admittedly seeks to avoid the onerous requirements imposed by New York's Sex Offender Registration Act, N.Y. Correct. L. § 168 et seq., by having this felony sodomy conviction vacated. Doe has every reason to maintain that he never had sex with Edwin.

For these reasons, a petitioner's own testimony that he did not commit the crime for which he was convicted, absent some type of corroboration, may rarely constitute reliable evidence of actual innocence.

Although this motive to lie might not be sufficient in itself to discredit Doe's hearing testimony, that testimony presents a substantially different account of his involvement with the Boy Lovers and his relationship with Edwin than his testimony at the Bagarozy/Federowicz trial and the 1993 proffer session, leading ineluctably to the conclusion that Doe's hearing testimony is not reliable. At the hearing, Doe repeatedly testified that although he was introduced to Edwin by Edward R., he associated with Edwin not because Edwin was one of the Boy Lovers' victims, but because Doe enjoyed taking him — and other boys who happened to be victims of the Boy Lovers — to places they enjoyed. Doe also attempted to distance himself from the Boy Lovers' activities, defining "boy lover" as someone who "prefers to associate" socially with boys, and maintaining that he was not very familiar with Bagarozy or his activities. Specifically, Doe asserted that he was not aware that Bagarozy was actually having sex with Edward R. and other boys, and that he himself had never used Edward R. to meet other potential victims. Thus, Doe attempted to portray his relationship with Edwin as fairly innocent, and to characterize it as only tangentially related to the Boy Lovers by common friends rather than by actual knowledge or a network of shared sex partners.

Doe's hearing testimony stands in stark contrast to his testimony at the Bagarozy/Federowicz trial and his other statements. With respect to his relationship with Edwin, Doe not only admitted at the proffer session to having sex with him, but also stated that Edwin was having sex with Bagarozy and O'Rourke. Moreover, he acknowledged at the Bagarozy trial that he was well aware, by 1993, that Edward R. and his brother were sleeping with Bagarozy, and perhaps other Boy Lovers as well. Thus, Doe's introduction to Edwin did not take place in an innocent context disconnected from the Boy Lovers; rather, it was arranged by a boy through whom Doe and other Boy Lovers had routinely met new victims. These statements are inconsistent with Doe's attempt at the hearing to present his relationship with Edwin as disconnected from the circle of Boy Lovers and their victims.

With respect to his connection to the other Boy Lovers, in prior testimony Doe twice defined "boy lover" as someone who has sex with boys who are under the age of consent, demonstrating both his awareness of the circle's activities and his own reason for joining the group. Doe also explicitly defined each Boy Lover's age of preference, leaving no doubt that Doe was aware that he was associating with a circle of pedophiles. Doe stated that in 1984, Doe and Bagarozy both had a sexual relationship with a boy named Tony, and that the two fought over the boy. Finally, Doe asserted that he observed Bagarozy coaching Edward R. on how to deceive investigators in 1991, and that he and Bagarozy took Edward R. to a nude beach in New Jersey. According to his trial testimony, therefore, Doe was both enmeshed in the Boy Lovers' activities and was close friends with Bagarozy, with whom he shared underage sex partners, exchanged child pornography, and discussed methods of seducing young boys.

In sum, a comparison of all of Doe's testimonial statements demonstrates that at the hearing Doe attempted to present his relationship with Edwin in a substantially more innocent light, and himself as substantially less involved with the Boy Lovers, than he had in prior testimony. Because Doe does not argue that he lied at the Bagarozy trial, or at the 1993 proffer session, in any respect other than his admission that he had sex with Edwin, the inconsistencies in his testimony can only reasonably indicate that Doe was attempting to present himself in a favorable light at the hearing. Given the stark discrepancies in the testimony, Doe's blunt admission that he had sex with Edwin before Edwin had even been identified by the state authorities, his obviously extensive knowledge of, and participation in, the Boy Lovers' activities, and his earlier testimony as to Edwin's involvement in those activities, Doe's hearing testimony is simply not reliable.

In ruling that Doe's testimony was credible in its entirety, the district court clearly erred. First, its failure to weigh the hearing testimony against the 1993 admissions, in the face of the court's recognition that Doe had no persuasive explanation of those admissions, was internally inconsistent. Second, the court failed to consider all of the evidence in the record when it neglected to attempt to reconcile Doe's hearing testimony with his Bagarozy/Federowicz trial testimony, and therefore remained unapprised of Doe's consistent attempt through his hearing testimony to present himself in a flattering light. The court also failed to consider Storino's testimony that Edwin told Doe that he would not give the authorities any incriminating information about Doe, which suggests that the relationship between Edwin and Doe was less than innocent. When Doe's hearing testimony is weighed against his 1993 admissions and other evidence in the record, bearing in mind his probable motives at both proceedings and his evident willingness to bend the truth at the hearing, his hearing testimony cannot be credited as reliable evidence of innocence.

We also find that the district court's discounting of Doe's admission of guilt at his plea allocution was clearly erroneous. The court credited Doe's assertion that he pled guilty only because he was afraid that the DA's Office would bring more serious charges, and because he did not want to endanger his cooperation against Bagarozy and Federowicz. It therefore concluded that Doe decided "to say whatever he needed to say" in order to plead guilty and secure a sentence of probation. While we have no reason to doubt the sincerity of Doe's explanations of the forces that led him to plead guilty, these explanations are not, viewed in light of the record as a whole, sufficient to support the district court's finding that he falsely allocuted to the crime. Defendants may be reluctant to plead guilty in the absence of clear strategic reasons for so doing not only because they are innocent, but for any number of reasons that have nothing to do with factual innocence, such as their belief that they would fare well at a trial, or their reluctance to be seen as "admitting" guilt. In light of our determination that the 1993 proffer session admissions are reliable, but Doe's hearing testimony that he never had sex with Edwin is not, we find that Doe has not established that he lied under oath at his guilty plea. Doe's plea admission that he had sexual contact with Edwin may therefore be considered as evidence of Doe's guilt. See Rosario v. United States, 164 F.3d 729, 734 (2d Cir. 1998) (considering petitioners' inculpatory statements made during plea allocution in determining whether no reasonable juror would find petitioners guilty).

2. Edwin

Edwin testified that he never had any sexual contact with Doe, and that when asked by Cronin whether Doe had ever made sexual advances toward him, he denied any improprieties and insisted that Doe was simply a "friend figure" and a positive influence. In crediting this testimony, the district court stated that it found Edwin "forthright [and] direct." While this assessment of Edwin's demeanor on the witness stand is uniquely within the province of the district court, the demeanor finding alone cannot be determinative of the reliability of Edwin's testimony. As discussed above, the actual innocence determination and Anderson require the court to consider the substance of the testimony in light of the record as a whole, as well as the witness's demeanor, when deciding whether to credit the witness's account. See Schlup, 513 U.S. at 327, 115 S.Ct. 851; Anderson, 470 U.S. at 575, 105 S.Ct. 1504 (noting that "factors other than demeanor and inflection go into the decision whether or not to believe a witness"). Here, the district court ignored considerable evidence in the record that directly contradicts or undermines Edwin's testimony. For this reason, we find that the court committed clear error in accepting Edwin's version of events, and that Edwin's testimony is not reliable evidence of Doe's actual innocence.

First, Edwin's testimony that Doe never molested him, while consistent with Doe's hearing testimony, is of dubious veracity for much the same reasons as Doe's. Doe himself admitted under oath at his plea allocution that he had had sex with Edwin, and has not provided a credible basis on which to discount that admission. Edwin's testimony must therefore be weighed against Doe's admission and the other evidence in the record. We find that Edwin's testimony conflicts with that evidence in a manner that leads to the conclusion that he was attempting to minimize his involvement with the Boy Lovers at the hearing.

At the hearing, Edwin testified that he knew Bagarozy only slightly, through his participation in a boys' basketball team run by Bagarozy, but that he did not know that Bagarozy was a Boy Lover. Edwin also stated that he knew of O'Rourke, but never had contact with him, and denied making a habit of spending time with single older men, either alone or in the company of other boys. At the same time, however, Edwin testified that he knew that Angel R. "was O'Rourke's boy," indicating that he had more extensive knowledge of the Boy Lovers than he otherwise admitted. The thrust of Edwin's testimony, therefore, is that he was only tangentially connected to the circle of Boy Lovers, and that he was unaware of the nature of their activities, or even of the men's preference for young boys.

These assertions are belied both by Doe's 1993 statements that O'Rourke and Bagarozy were sexually involved with Edwin, and by Investigator Storino's testimony. Doe testified at the evidentiary hearing that he, O'Rourke, and Edwin had gone to Playland Park in 1994 to discuss the DA's Office's investigation, thus contradicting Edwin's assertion that he never met O'Rourke. Storino also provided evidence of Edwin's close association with Doe and other Boy Lovers by testifying that, in January 1994, before Edwin was interviewed by the DA's Office, Doe informed federal agent Kyle that O'Rourke had met with Edwin and coached him on lying to the authorities. Moreover, Edwin attempted to contact Doe in order to assure him that he would say nothing incriminating. Storino's testimony is corroborated by the notes that he took during his conversation with Agent Kyle, which state that "Edwin [is] reaching out thru Bill O'R[ourke] and others to see [Doe]. Edwin and Edward [R.] spoke with O'R[ourke] and were coached on what to say to thwart the investigation. Edwin is supposedly not going to say anything incrim[inating]. . . . Edwin and Edward [R. are] spending time with O'R[ourke]." This evidence contradicts Edwin's assertions that he had never met O'Rourke, that he did not realize that Doe and his adult friends were pedophiles, and also strongly suggests that Edwin himself was involved in sexual activities, and therefore had "incriminating" information. Moreover, Edwin's stated desire to inform Doe in particular that he would not divulge anything to the authorities raises the inference that Doe would have reason to fear Edwin's communicating with investigators.

It is worth noting that both Storino's testimony on this issue and his notes rest on the hearsay statements of O'Rourke and Kyle. We need not consider whether this evidence would be admissible at a trial, however, because Schlup allows us to consider all evidence in the record. See Schlup, 513 U.S. at 327, 115 S.Ct. 851. We find Storino's testimony and notes to be reliable, because both O'Rourke's and Doe's statements were against their penal interest, Kyle and Storino are law enforcement professionals who were performing their duties, and Storino contemporaneously recorded the substance of his conversation with Kyle. Moreover, neither party has proffered any evidence to suggest that either Kyle or Storino fabricated the evidence.

Thus, the bulk of the evidence in the record casts doubt on the veracity of Edwin's testimony as a whole. Storino's testimony and notes, which suggest a particular connection between Doe and Edwin and corroborate Doe's proffer session statement that Edwin was close to O'Rourke, demonstrate that Edwin was attempting to distance himself from the Boy Lovers in the hearing, and that the version of events that he presented was not entirely accurate. The district court, however, ignored this evidence in assessing Edwin's testimony. Given the indications that Edwin lied about the closeness of his relationship with Doe and others, it is impossible to credit Edwin's testimony over the other evidence, including Doe's own admission, that Doe assaulted Edwin.

Second, Edwin's assertion that he denied being victimized by Doe when interviewed by the DA's Office is contradicted by testimony and by other evidence. Most obviously, ADA Cronin, who conducted the interviews, testified that Edwin detailed two instances of sexual advances by Doe, one of which resulted in actual sexual contact and formed the basis of the sodomy charge. The district court discounted her testimony, however, finding that her "recollection [was] based primarily on her notes," and that Cronin "conceded that she was not sure if the notes were" taken during the interviews. This is a mischaracterization of Cronin's testimony.

Cronin testified that, of seven pages of handwritten notes, four pages were dated March 18, 1994, and were taken during the second interview with Edwin. These are the pages that purport to record Edwin's account of the sexual assault. Of the remaining three pages, Cronin testified that two were taken during the first interview, although they are not dated, and the last page may have been taken while she was preparing for a later proceeding. Thus, the record is clear that the four pages of notes on which Cronin relied for her account of Edwin's accusations were taken during an interview, and the fact that Cronin refreshed her recollection with these notes therefore provides no basis for discounting her testimony. In addition, the district court's assumption that Cronin did not remember her interviews with Edwin because "[t]his case was but one of many sordid situations involving children dealt with by Ms. Cronin" is pure speculation, as Cronin did not suggest that her memory of the interviews was impaired for any reason, and there was no testimony that Cronin routinely dealt with similar accusations. The district court therefore discounted Cronin's testimony for reasons that are not valid.

Cronin's account of the interviews and Edwin's statement that Doe assaulted him was corroborated in full by the testimony of ADA Bernhard. The district court failed to consider this testimony in substance, however, and made no credibility determination, adverse or otherwise, with respect to Bernhard. Bernhard, who was present at the interviews, testified that he had not reviewed Cronin's notes because he did not need to refresh his recollection, and gave an account of Edwin's accusations and description of his relationship with Doe that is identical to Cronin's testimony and her notes in every respect.

In order to credit Edwin's testimony over Cronin's and Bernhard's, therefore, the factfinder would have to conclude that Cronin and Bernhard were part of a conspiracy to fabricate their account of Edwin's statement and the notes that recorded the statement. The district court explicitly refused to find that Cronin was lying or had fabricated her notes, however, perhaps because there is no evidence in the record that suggests the possibility of prosecutorial misconduct. Yet the court's conclusion that Cronin did not remember the interviews, and might not have taken her notes during the interviews, does not go far enough to explain the inconsistency between her testimony and Edwin's, as crediting Edwin's testimony requires the factfinder to conclude not that Cronin's recollection of what was said at the interviews differed from Edwin's, but that Cronin affirmatively fabricated the account of the sexual assault that is recorded in her notes. In this respect, the court's findings are internally inconsistent, as the court credited Edwin's testimony in full, but did not make the further finding that Cronin had lied at the hearing and committed misconduct during the prosecution. Furthermore, by ignoring Bernhard's testimony, the district court failed to acknowledge that because Cronin was not the only prosecutor involved in prosecuting Doe, it would have to discredit both prosecutors, and find that a conspiracy existed in the DA's Office in 1994, in order to credit Edwin's account. We find, therefore, that the district court's crediting of Edwin's testimony in light of the clearly contradicting evidence and in light of Doe's 1993 admission and other incriminating statements was clearly erroneous.

Our conclusion does not conflict with our holding in Hemstreet v. Greiner, 367 F.3d 135 (2d Cir. 2004), in which we upheld the district court's finding, on the basis of a sparse record, that agents of the prosecution had threatened a defense witness. Id. at 139-40. There, the state did not argue that the misconduct had not occurred, and we therefore found that the district court's inference of misconduct, in the absence of any evidence to the contrary, was not clearly erroneous. See id. at 140. Here, of course, the state has presented credible evidence that the employees of the DA's Office acted properly.

We do not hold that prosecutorial misconduct may never be proven absent direct evidence of misconduct or that an alleged victim's statement that he never accused the defendant of wrongdoing is always insufficient to raise an inference of misconduct. Courts observe a presumption that prosecutors act properly in furtherance of their duties, see United States v. Mezzanatto, 513 U.S. 196, 210, 115 S.Ct. 797, 130 L.Ed.2d 697 (1995) ("Rather, tradition and experience justify our belief that the great majority of prosecutors will be faithful to their duty.") (quoting Town of Newton v. Rumery, 480 U.S. 386, 397, 107 S.Ct. 1187, 94 L.Ed.2d 405 (1987) (plurality opinion)), but that presumption is of course rebuttable by competent evidence. We simply hold here that Edwin's testimony alone is insufficient, in light of his lack of candor about his involvement with the Boy Lovers and the other evidence that the attack actually occurred, to raise an inference of prosecutorial misconduct. B. Application of Schlup to Doe's Proffered New Evidence

For much the same reasons, Edwin's less detailed affidavit, submitted as part of Doe's habeas petition, does not constitute reliable evidence of Doe's innocence.

We find that Doe presented no new reliable evidence of his factual innocence at the March 2003 evidentiary hearing or in his habeas submissions, as is required by Schlup. See Schlup, 513 U.S. at 324, 115 S.Ct. 851. Because Doe has not presented any new reliable evidence, it is unnecessary to determine whether no reasonable juror would convict in light of Doe's newly proffered evidence.

It is worth noting, however, that the evidence of Doe's guilt is overwhelming. Doe admitted guilt twice, at the proffer session and his plea allocution, and Edwin described the assault to Cronin.

The district court's crediting of Doe's and Edwin's testimony, and its consequent finding of actual innocence, resulted in the dilution of Schlup's directives. Schlup dictates that the habeas court faced with purported new evidence of a petitioner's actual innocence make its own credibility determinations based on its assessment of the totality of evidence in the record, both "old" and "new." Because Schlup also requires that any new evidence of actual innocence be reliable, the habeas court must analyze not only whether the new evidence throws the pre-existing evidence into doubt, but whether the new evidence itself may be considered reliable in light of the pre-existing evidence. Thus, the habeas court may not merely examine the new evidence and, if it believes the new material credible without regard to the pre-existing record, find that the petitioner is actually innocent. That is how the district court proceeded here, and its disregard for the bulk of the evidence in the record allowed it to circumvent Schlup's instruction that the proffered new evidence must be reliable. To allow this mode of analysis would eviscerate the Schlup standard.

In light of our conclusion that it is not necessary to apply the Schlup"no reasonable juror" standard because Doe has not presented any new reliable evidence, we need not review the district court's application of that standard. The district court's misstatement of the Schlup standard in its June 2003 order merits comment, however, in order to clarify our standing instruction to district courts faced with actual innocence claims in the context of AEDPA tolling. See Whitley, 317 F.3d at 225-26. The district court stated that Schlup requires only that "innocence . . . be demonstrated to be more likely than unlikely," thereby substantially reducing the showing that Doe was required to make in order to establish his actual innocence. This statement of the standard essentially allowed the district court to act as a jury, determining in the first instance whether Doe had demonstrated by a preponderance of the evidence that he was innocent. Thus, the court concluded that "petitioner has established his innocence by a preponderance of the credible evidence."

As discussed above, however, the Schlup standard requires the district court to find, by a preponderance of the evidence, that no reasonable juror would find the petitioner guilty. The preponderance showing is not in itself a high burden of proof, as the district court noted, but the fact that the showing must establish that no reasonable juror would convict substantially increases that burden. Even if the court, as one reasonable factfinder, would vote to acquit, the court must step back and consider whether the petitioner's evidentiary showing most likely places a finding of guilt beyond a reasonable doubt outside of the range of potential conclusions that any reasonable juror would reach. See Schlup, 513 U.S. at 333, 115 S.Ct. 851 (O'Connor, J., concurring). If the court finds that even one juror might reasonably vote to convict, the court must find that the petitioner has failed to establish his actual innocence.

The district court also misapprehended our direction in Doe I that the court should "decide as a matter of fact whether Doe has presented a credible claim of actual innocence." Doe I, 49 Fed.Appx. at 342. We later extended this instruction to all district courts faced with claims of actual innocence in the context of AEDPA tolling in Whitley, 317 F.3d at 225-26, in which we held that district courts should determine whether a petitioner presents a "credible claim of actual innocence" before ruling on the legal issue of whether the limitations period could be tolled. The district court understood our reference to a " credible claim of actual innocence" to represent a "lesser standard" than that delineated by Schlup, and noted that, in its view, Doe had satisfied both the higher Schlup standard and the lesser credible claim standard. The reference to a credible claim arises directly from Schlup itself, however, which states that "[t]o be credible, [an actual innocence] claim requires petitioner to support his allegations of constitutional error with new reliable evidence." Schlup, 513 U.S. at 324, 115 S.Ct. 851. Thus, our instruction in this case, and our extension of that instruction in Whitley, do not purport to relax the Schlup standard or suggest that the burden of showing of actual innocence is lessened in the equitable tolling context, but simply refer to the fact that any showing of actual innocence must be based on new reliable evidence.

The issue in this case is not one of the relative credibility of witnesses but of reliability of new evidence in light of all the evidence. The actual innocence standard is a demanding one. Failing to account for Doe's conduct inconsistent with his claim of actual innocence, our dissenting colleague would have us overturn the weighty presumptions favoring the veracity of a defendant's sworn plea of guilty and the propriety of prosecutorial conduct. While a recanting victim presents a sympathetic scenario for a claim of actual innocence, we must not flinch from our duty to scrutinize all the evidence to assure ourselves of that recantation's reliability. Here, the evidence proffered by Doe fails the strict requirements of Schlup. C. AEDPA Tolling for Actual Innocence

In light of our holding, we need not review the district court's conclusion that a credible claim of actual innocence does not toll AEDPA's limitations period. We have previously stated that "[i]n a proper case, we will decide whether constitutional considerations require an actual innocence exception to the AEDPA's statute of limitations." Whitley, 317 F.3d at 225. Such a proper case will arise, of course, where a petitioner is able to make a credible showing of actual innocence based on new evidence, thereby demonstrating that equitably tolling the limitations period could prevent a miscarriage of justice. Doe's petition clearly does not present such a case.

It is worth noting, moreover, that the Supreme Court has recently counseled lower courts to avoid extending the actual innocence exception to new contexts if there are "other grounds for cause to excuse the procedural default." Dretke v. Haley, 541 U.S. 386, ___, 124 S.Ct. 1847, 1852, ___ L.Ed.2d ___, ___ (2004). The Court's statement that "[b]ecause the various exceptions to the procedural default doctrine are judge-made rules[,] . . . courts as their stewards must exercise restraint, adding to or expanding them only when necessary," id. at 1853, counsels restraint in the judge-created area of equitable tolling as well. Thus, in order to avoid "the unhappy effect of prolonging the pendency of federal habeas applications," id., district courts confronted with allegations of actual innocence as a ground for tolling the limitations period should first consider any other grounds for tolling, and only after having ruled out any other possible means of tolling, consider petitioner's actual innocence claim.

This is fully consistent with our decision in Whitley, which instructs district courts on the manner in which to create a record that is sufficient for this Court to use to determine whether a given petition presents the proper case in which to decide whether actual innocence is a ground for equitable tolling. See Whitley, 317 F.3d at 225-26 (laying out questions that district court should investigate with respect to actual innocence claims). We simply hold here that, in accordance with Haley, district courts should address the actual innocence question only after fully exploring all other avenues of relief.

II. Tolling for Attorney Incompetence

Doe next argues that the limitations period should be tolled between February 1999 and February 2000 because Patrick Wall, the attorney who represented him during that time period, incompetently failed to file Doe's state post-conviction motion. The district court dismissed this argument in its February 2002 order, finding that Wall's alleged incompetence did not rise to the level of extraordinary circumstances necessary to invoke equitable tolling, see Smaldone v. Senkowski, 273 F.3d 133, 138 (2d Cir. 2001). The court also found that Doe did not exercise reasonable diligence during this period, because he should have realized that Wall was incapable of filing the petition at least ten months before he finally fired him and hired new counsel. We declined to review these conclusions in Doe I, and Doe now argues that Baldayaque v. United States, 338 F.3d 145 (2d Cir. 2003), which clarified that egregious misconduct on the part of an attorney can, in some circumstances, be sufficiently extraordinary to toll the limitations period, dictates that Wall's conduct is sufficient to entitle Doe to tolling. We need not reach this argument, however, because Doe has not established that he acted with reasonable diligence in attempting to file his federal habeas petition during the period that he seeks to toll.

If the limitations period were tolled during this period, then the clock would have run at least until May 10, 2000, when Doe filed his state post-conviction motion. The filing of that motion would have tolled the limitations period, pursuant to 28 U.S.C. § 2244(d)(2), until November 14, 2001, when the denial of his collateral attack became final. Doe's federal habeas petition, filed on November 30, 2001, after only 189 days had run, would have been timely.

In order to equitably toll AEDPA's limitations period, a petitioner must show that "extraordinary circumstances prevented him from filing his petition on time." Smith v. McGinnis, 208 F.3d 13, 17 (2d Cir. 2000). This formulation "requires the petitioner to demonstrate a causal relationship between the extraordinary circumstances . . . and the lateness of his filing." Valverde v. Stinson, 224 F.3d 129, 133 (2d Cir. 2000). As one component of the causation showing, the petitioner must establish that he diligently pursued his application during the time that he seeks to have tolled. See id. at 134. If the petitioner is unable to establish that he diligently attempted to file his petition, the extraordinary circumstances on which his tolling claim is based cannot be said to have caused the lateness of his petition. See Hizbullahankhamon v. Walker, 255 F.3d 65, 75 (2d Cir. 2001).

Even where the extraordinary circumstances on which the petitioner rests his claim involve attorney incompetence, the petitioner must still demonstrate that he himself made reasonably diligent attempts to ensure that his petition was filed on time. See Baldayaque, 338 F.3d at 153. In other words, the act of retaining an attorney does not absolve the petitioner of his responsibility for overseeing the attorney's conduct or the preparation of the petition. Particularly because petitioners often are fully capable of preparing and filing their habeas petitions pro se, and pro se status does not in itself constitute an extraordinary circumstance meriting tolling, see Smith, 208 F.3d at 18, it would be inequitable to require less diligence from petitioners who are able to hire attorneys than from those who are forced to proceed pro se. In the attorney incompetence context, therefore, the reasonable diligence inquiry focuses on the purpose for which the petitioner retained the lawyer, his ability to evaluate the lawyer's performance, his financial and logistical ability to consult other lawyers or obtain new representation, and his ability to comprehend legal materials and file the petition on his own. See Baldayaque, 338 F.3d at 153 (listing factors relevant to reasonable diligence that district court should consider on remand).

The district court issued its decision with respect to the attorney incompetence issue before Baldayaque was decided, and therefore did not have the opportunity to consider the egregiousness of Wall's conduct in light of Baldayaque's holding that "at some point, an attorney's behavior may be so outrageous or so incompetent as to render it extraordinary," id. at 152. The court ruled not only that Wall's conduct was not sufficiently outrageous, but also that Doe was not reasonably diligent. In so doing, the district court nevertheless considered the factors that Baldayaque establishes as relevant to the reasonable diligence determination. After determining that Doe should have been aware of Wall's incompetence as early as April 1999, had both the financial wherewithal to hire another lawyer and his brother's logistical assistance, and could have filed pro se at any point, the court concluded that "[t]he facts submitted, viewed most favorably to Petitioner, do not suggest a finding of . . . prevention by . . . misconduct which is at the core of most equitable tolling cases." In light of the district court's consideration of the Baldayaque factors, as well as Doe's opportunity to present evidence of his asserted diligence both in his initial submissions to the district court and during the hearing on remand, we find no need to remand to the district court once again to consider Doe's claim in light of Baldayaque. Rather, we find that as a matter of law, Doe cannot establish that he acted with reasonable diligence during the period in which he was represented by Wall.

Although the mandate instructed the district court to conduct factfinding as to Doe's reasonable diligence with respect to his actual innocence claim, the court allowed both Doe and his brother to testify at length regarding Wall's incompetence, for the stated purpose of creating a record for our consideration of the attorney competence issue on appeal.

An examination of the circumstances under which Doe hired Wall, the scope of the representation, Doe's instructions to Wall, and Doe's capability of filing his habeas petition himself, indicates that Doe did not exercise reasonable diligence. The district court found, and the record supports, that Doe became interested in collaterally challenging his conviction shortly after he was arrested on federal child pornography charges in November 1998, when he realized that his second-degree sodomy conviction would mandate a significantly higher sentencing range under the federal Sentencing Guidelines and the relevant child pornography statute. To this end, Doe hired Patrick Wall for the stated purposes of filing a state post-conviction motion pursuant to New York Penal Law § 440.10, and representing Doe with respect to any sentencing issues that might arise in connection with the federal charges.

Doe was once again represented by Murray Richman in connection with his defense of the federal charges, but Richman declined to represent him with respect to the § 440 motion because he would potentially have to testify with respect to Doe's claim that the DA's Office violated his federal immunity agreement.

The evidence demonstrates that the scope of Wall's representation was limited to vacating the sodomy conviction for the purpose of favorably impacting Doe's federal sentencing. After Doe pled guilty to the federal charges on April 9, 1999, his sentencing was initially scheduled for January 7, 2000. Although the AEDPA limitations period was to expire in November 1999, Doe and Wall treated their operative deadline for filing the § 440 motion as January 2000, as Wall promised to have the motion prepared by "New Years' Day 2000." The November 19 expiration of the statute of limitations came and went without notice, apparently, as Doe and his brother made a payment of $10,000 in legal fees to Wall around December 15, 1999. Moreover, Doe's account of his conversations with Wall demonstrates that Doe was concerned only with his sentencing. Wall assured Doe that he would file the § 440 motion well before the sentencing, but neither mentioned the approaching November 19 AEDPA deadline. Even more tellingly, Doe testified that he understood Wall's desire to postpone filing the § 440 motion until they received the federal pre-sentence report as based at least in part on the assumption that if the pre-sentence report did not award criminal history points for the sodomy conviction, it would not be worth filing the § 440 motion at all. The fact that Doe understood this to be Wall's thinking on the issue, and did not correct him or ask him to file the § 440 motion regardless of Doe's federal sentencing outlook, indicates that Doe himself was concerned only about his sentencing, and Wall acted accordingly.

Thus, although Wall might have accomplished the filing of the § 440 motion in time to toll the AEDPA limitations period if not for his alleged incompetence, it cannot be said that Wall's incompetence prevented Doe from filing a federal petition or filing the § 440 motion before November 19. Wall was not retained for the purpose of filing a federal habeas petition or exhausting Doe's claims in preparation for the federal petition, and apparently was not instructed to preserve Doe's federal habeas rights. Rather, Doe's focus on his federal sentencing and the goal of avoiding heightened sentencing exposure led him to overlook the expiration of the AEDPA statute of limitations, and to define the scope of his agent's duties accordingly. Given that we expect pro se petitioners to know when the limitations period expires and to understand the need to file a state post-conviction motion within that limitations period, see Smith, 208 F.3d at 18, such inadvertence on Doe's part cannot constitute reasonable diligence.

Although Doe asserted in his affidavit that he realized that AEDPA provided a "limited time" in which to file his federal habeas petition, his interactions with Wall indicate that he was not concerned with the expiration of the limitations period. As discussed below, Doe offers no excuse for his failure, in light of his asserted awareness of the need to file a timely petition, to ensure that his habeas rights were preserved, either by supervising his attorney or by filing pro se.

Moreover, although Doe testified that in April 1999 he asked Wall to file the § 440 motion as soon as possible, Doe did nothing even as it should have become evident that Wall was not actually working on the motion. During the spring of 1999, Doe, a self-professed lay student of the law, repeatedly asked Wall why it was necessary to wait for the federal pre-sentence report before filing the § 440 motion, but Wall refused to answer him. Given Doe's extensive experience with legal representation and the legal system, this unresponsiveness in the face of a client's questioning should have been ample evidence of Wall's dereliction. Even earlier, however, there were indications that Wall was not actually maintaining a viable law practice, as Wall asked Doe to wire the retainer fee to his wife's bank account, had no proper office, and repeatedly remained unreachable for extended periods of time. Although Doe states that he had no way of knowing about the deterioration of Wall's practice because he was incarcerated, Doe's brother was in constant contact with Wall, and was fully capable of evaluating the status of Wall's practice for himself.

Had Doe been concerned with meeting the AEDPA deadline, he would have realized earlier that Wall's unresponsiveness was jeopardizing his ability to file the § 440 motion in time to toll the AEDPA limitations period. Doe then would have had a number of choices as to how to proceed. See Villanueva v. United States, 346 F.3d 55, 63 (2d Cir. 2003) (rejecting petitioner's claim that he detrimentally relied on a potential attorney, where petitioner was aware of the AEDPA deadline and capable of taking other steps to meet it). He could have filed a federal habeas petition or the § 440 motion pro se at any point during the summer and fall of 1999. As discussed below, the purported inadequacies in the federal prison library should not have prevented Doe from so doing. Doe was obviously fully capable of acting pro se, as his personal submissions to the court are well-written, articulate, and clearly the product of someone both fluent in English and highly educated. Doe's testimony in this and other proceedings evidences his legal sophistication, and there is no reason that he could not have harnessed his legal abilities and experience in filing a challenge to his conviction. Alternatively, Doe could have hired another lawyer, as he proved he was capable of doing in February 2000. Doe was at the time represented in connection with the federal charges by Murray Richman, who could have helped him obtain another lawyer, and Doe's brother was clearly willing and able to provide assistance as well.

Doe's exclusive focus on obtaining the most favorable federal sentence possible, and his consequent inattention to the AEDPA deadline and Wall's incompetence, compel the conclusion that he failed to act with reasonable diligence. It is important to note, however, that because the standard is reasonable diligence, the amount of diligence required necessarily varies according to the petitioner's abilities and circumstances. See Baldayaque, 338 F.3d at 153 (noting that reasonable diligence is considered in light of the petitioner's circumstances). Despite his incarceration, Doe was in a privileged position relative to most habeas petitioners, as he possessed the education, skills, and financial and logistical ability capably to evaluate his lawyer's performance and act to remedy any problems. In contrast, many prisoners may be forced to rely on incompetent lawyers because their lack of English skills, education, funds, or familiarity with the legal system, as well as their inability to obtain access to other lawyers or sources of assistance, leaves them unable to evaluate their lawyer's performance or hire new counsel. Despite these advantages, Doe did not demonstrate any desire to file the § 440 motion before the AEDPA deadline, did not instruct Wall to preserve his habeas rights, and failed to fire Wall after it became clear that his performance was incompetent.

This failure is inexcusable in light of Doe's abilities, and renders his situation distinguishable from that at issue in Baldayaque. There, we found that a remand was warranted to allow the district court to consider whether petitioner had exercised reasonable diligence, on the basis of ample evidence suggesting that Baldayaque had no choice but to rely on his attorney. The evidence indicated that Baldayaque had hired his attorney specifically to file his § 2255 motion, and he and his wife spoke little or no English, had little education, and were incapable of raising the funds necessary to retain another lawyer. See id. (suggesting that Baldayaque's difficulties might lead to a finding of reasonable diligence). Although we do not suggest that a petitioner's situation must be as extreme as Baldayaque's in order to demonstrate reasonable diligence, we also do not hold that all petitioners alleging attorney incompetence should be held to the qualitative standard of reasonable diligence to which we hold Doe.

III. Tolling for Inadequate Prison Library Materials

Finally, Doe argues that he was unable to file his § 440 motion pro se because he was incarcerated in federal prison during the operative time period, and consequently had no access to New York State case law. Although this Circuit has not yet determined whether the deprivation of access to legal materials can constitute the extraordinary circumstances necessary to toll AEDPA's limitations period, see Hizbullahankhamon, 255 F.3d at 75-76, there is no need to address this issue here, because Doe's argument is patently meritless. Doe's § 440 motion was based primarily on federal law, as he argued that the DA's Office withheld Brady material, broke its promise not to prosecute him, and coerced his guilty plea, and that he received ineffective assistance of counsel preceding his plea. Doe therefore could have prepared the motion using the federal case law that he concedes was available where he was incarcerated. Moreover, as discussed above, Doe has not proffered any evidence that despite the availability of federal case law, he could not have filed the § 440 motion pro se, or suggested that any of the arguments raised in his eventual § 440 motion could not have been raised without relying on New York case law. We therefore affirm the district court's denial of equitable tolling on this theory.

CONCLUSION

For the foregoing reasons, the judgment of the district court is AFFIRMED on the grounds stated in this opinion.


POOLER, dissenting.

I respectfully dissent because I believe we should affirm the district court's finding that Doe presented a credible claim of actual innocence. This panel should not second guess the trial judge's credibility findings where no compelling extrinsic or other evidence compels a different conclusion. In holding otherwise, the majority supplants the district court's reasoned factual findings with its own speculations and conjectures. More importantly, the majority glosses over the most unusual and disconcerting aspect of this case: here, the alleged victim of a violent sexual assault has come forward and testified that the crime never occurred. It is my view that no reasonable juror, faced with a witness, found highly credible by the district court, who emphatically testifies that he was never victimized, could find beyond a reasonable doubt that the defendant was guilty of that same crime. This is therefore one of the few cases "where a constitutional violation has probably resulted in the conviction of one who is actually innocent[.]" Murray v. Carrier, 477 U.S. 478, 496, 106 S.Ct. 2639 (1986).

I. STANDARD OF REVIEW

In order to demonstrate actual innocence in a collateral proceeding, a petitioner must present "new reliable evidence" that was not presented at trial and show "that it is more likely than not that no reasonable juror would have found [him] guilty beyond a reasonable doubt." Schlup v. Delo, 513 U.S. 298, 324, 327, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995) (quotation marks omitted). Where the defendant pleaded guilty, the standard nevertheless remains the same. See Bousley v. United States, 523 U.S. 614, 623, 118 S.Ct. 1604, 140 L.Ed.2d 828 (1998).

Here, the district court found that Edwin and Doe credibly testified that the crime for which Doe was convicted never occurred. It then found that Doe "exercise[d] reasonable diligence in pursuing his actual innocence claim once he became able to pursue it at all." Doe v. Menefee, No. 01 Civ. 10782, slip op. at 22 (S.D.N.Y. June 23, 2003). "In reviewing factual findings, we owe particularly strong deference where the district court premises its findings on credibility determinations." United States v. Monzon, 359 F.3d 110, 119 (2d Cir. 2004) ( citing Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 64 (2d Cir. 1992)) (internal quotation marks omitted); see also Anderson v. City of Bessemer City, 470 U.S. 564, 575, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985). The "clearly erroneous" standard, applied to factual determinations made by the district court, is derived from Fed.R.Civ.P. 52(a), which provides, in pertinent part, that "[f]indings of fact, whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge the credibility of the witnesses." Id. (emphasis added). In Anderson, the Supreme Court held that the circuit court overstepped its bounds in finding that the district court's factual conclusions were clearly erroneous. In so holding, the Court reiterated that this standard sets an exceedingly high threshold:

If the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Where there are two permissible views of the evidence, the fact finder's choice between them cannot be clearly erroneous.

Anderson, 470 U.S. at 573-74, 105 S.Ct. 1504 (citations omitted). Recognizing the superior position that trial courts occupy in making factual determinations, the Court noted that trial courts are particularly able to assess the credibility of testifying witnesses:

When findings are based on determinations regarding the credibility of witnesses, Rule 52(a) demands even greater deference to the trial court's findings; for only the trial judge can be aware of the variations in demeanor and tone of voice that bear so heavily on the listener's understanding of and belief in what is said.

Id. 470 U.S. at 575, 105 S.Ct. 1504 (emphasis added, citations omitted).

This circuit has repeatedly relied on Anderson in finding that a "district court's factual findings based on the testimony of witnesses are distinctly the province of the district court, and we will not lightly overturn such assessments." United States v. Beverly, 5 F.3d 633, 642 (2d Cir. 1993) (citations omitted); see also United States v. Morrison, 153 F.3d 34, 52 (2d Cir. 1998) ("We grant particular deference to the district court when its factual findings are based upon the court's observation of testimony.").

Although the majority notes that, under certain circumstances, this Court has "reversed factual findings where the trial court incorrectly assessed the probative value of various pieces of evidence[,]" Op. at 165, the circumstances illustrated by these cases do not involve situations where the evidence at issue was witness testimony. Instead, the majority cites to United States v. Rizzo, where this Court primarily evaluated documentary evidence before the district court, and the district court's conclusions with respect to this evidence. 349 F.3d 94, 100-02 (2d Cir. 2003) (concluding that the district court clearly erred in finding "that the documents used by [the defendant] were obtained through theft[.]"). Rizzo did not comment on the reliability or credibility of any testifying witness. Similarly, in Jiminez v. Mary Washington College, the Fourth Circuit held that the district court drew unreasonable conclusions with respect to documentary evidence, specifically student evaluations. 57 F.3d 369, 380 (4th Cir. 1995). Finally, in Ortega v. Duncan, this Court found that the lower court erred in drawing unreasonable conclusions based on testimony which the lower court had found to be credible. 333 F.3d 102, 107 (2d Cir. 2003). In so holding, we did not disturb the lower court's finding that the prior testimony of the witness, which was the recantation of a previous statement, was not credible, and in fact, we specifically noted that we gave "deference to [the lower court's] evaluation of the credibility of his recantation." Id.

None of the cases cited by the majority involve an appellate court subverting the lower court's assessment of a witness's testimony and I was unable to find any case where this Court refused to abide by a district court's credibility findings. As the following discussion makes clear, the facts of this case are not such that a departure from this established line of cases is warranted.

II. STANDARD FOR DETERMINING ACTUAL INNOCENCE

A. Standard for Determining Actual Innocence

I agree with the majority that the district court misapprehended the standard for determining actual innocence. Op. at 161. The district court incorrectly held that it was "satisfied on the totality of the record before [it] that [Doe] has established his innocence by a preponderance of the credible evidence. He has also met the lesser standard mentioned in the mandate, of having presented a credible claim of actual innocence. On the evidence presented, a reasonable juror would have to find a reasonable doubt as to [Doe's] guilt." Doe, No. 01 Civ. 10782, at 20. The district court based this conclusion on the erroneous assumption that "[i]nnocence need only be demonstrated to be more likely than unlikely." Id. at 5.

Our inquiry, however, is not whether Doe established his innocence by a preponderance, but rather, whether the new evidence, when considered in conjunction with the entire record, establishes that it is "more likely than not that no reasonable juror would have found petitioner guilty beyond a reasonable doubt." Schlup, 513 U.S. at 327, 115 S.Ct. 851.

B. Evidence That Can Be Considered In Determining Actual Innocence
Schlup makes clear that, in assessing actual innocence, the district court must reach its decision in "light of all the evidence, including that alleged to have been illegally admitted (but with due regard to any unreliability of it and evidence tenably claimed to have been wrongly excluded or to have become available only after the trial.)" Id. at 328, 115 S.Ct. 851 (quoting Henry Friendly, "Is Innocence Irrelevant? Collateral Attack on Criminal Judgments," 38 U. CHI. L.REV. 142, 160 (1970)). More recently, however, the Court further specified the type of evidence that should be considered in assessing actual innocence. In its directive to the district court on how to assess an actual innocence claim that may arise on remand, the Supreme Court in Bousley stated that "the Government is not limited to the existing record to rebut any [actual innocence] showing that petitioner might make. Rather the Government should be permitted to present any admissible evidence of petitioner's guilt even if that evidence was not presented during petitioner's plea colloquy." 523 U.S. at 624, 118 S.Ct. 1604 (emphasis added); see also Fountain v. United States, 357 F.3d 250, 255 (2d Cir. 2004) (reciting the Bousley"admissible evidence" standard for determining actual innocence in a case involving conspiracy to launder proceeds of wire fraud). Thus, the only evidence that this Court should consider in determining actual innocence is evidence that would have been admissible at trial, and not patently inadmissible evidence such as hearsay testimony or highly prejudicial evidence with little probative value. See Rosario v. United States, 164 F.3d 729, 733-34 (2d Cir. 1998) (finding that defendant failed to establish actual innocence under Bousley because the evidence in the record, minus the hearsay testimony by the confidential informant, established defendant's guilt). The majority dismisses Bousley's directive that a court may consider only admissible evidence as "mere dictum," Op. at 163. While it is true that the Court did not strictly have to reach the evidentiary issue because the actual innocence claim was not directly before it, we should not ignore such an explicit directive from the highest court. Moreover, a contrary reading would cause an inconsistency with the actual innocence standard, which uses as its benchmark the reasonable juror. This hypothetical juror would plainly not have access to inadmissible evidence. Accordingly, as intimated by Bousley, in determining whether the reasonable juror would have found the petitioner to be actually innocent, this Court may not rely on evidence that would have been plainly unavailable to this hypothetical juror. The majority's implicit suggestion that there is a two-step process that neatly separates the evidentiary analysis from the reasonable juror analysis is not supported by case law.

III. APPLICATION OF LAW TO FACTS

I respectfully find that the majority improperly substituted its own view of the evidence over that of the district court's when it concluded that "the district court's finding that the hearing testimony of Doe and Edwin constitutes reliable evidence of Doe's actual innocence was clearly erroneous." Op. at 165. Instead, I find that the district court's assessment of Doe's and Edwin's testimony was reasonable, and its "account of the evidence is plausible in light of the record viewed in its entirety." Anderson, 470 U.S. at 574, 105 S.Ct. 1504. Thus, this Court's reversal of the district court on this issue is not appropriate.

(a) No Physical Evidence Supported The Allegation That Edwin Had Been Assaulted

The majority disregards the district court's role as the finder-of-facts when it sets aside the lower court's assessment of Edwin's testimony. In discounting his testimony, the majority relies not on solid extrinsic evidence, but instead, on testimony by other witnesses. Such testimony was oftentimes hearsay, conjectural, or directly refuted by other testimony. Thus, the alleged victim's testimony was not contradicted by any extrinsic evidence; instead, the majority chooses to credit the testimony by other witnesses over the testimony by the victim. This choice, however, is not for us to make. As held by the Supreme Court in Anderson,"when a trial judge's finding is based on his decision to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, that finding, if not internally inconsistent, can virtually never be clear error." 470 U.S. at 575, 105 S.Ct. 1504.

Here, there was no forensic evidence or trustworthy eyewitness accounts establishing that Doe had in fact assaulted Edwin. Indeed, there was no admissible evidence that corroborated the testimony by former prosecutors in the Westchester District Attorneys Office ("D.A.'s Office"), aside from handwritten notes created by the prosecutor herself. Therefore, pursuant to Anderson, the district court's ruling should not be disturbed unless there is extraordinary evidence that directly undermines the district court's assessment of the victim's testimony. No such evidence exists in this case.

(b) Edwin's Testimony

At the hearing before the district court, Edwin testified that he was 22 years old, worked for Wal-Mart, had one son, and resided in upstate New York (Arkville). Regarding the events at issue, Edwin stated that he never engaged in sexual relations with Doe, forcible or otherwise, and that he was never alone with Doe when he went to see him. He also stated that Doe never refused to take him home or put his hands on him. He testified unequivocally that, although "I was young at the time . . . I knew right from wrong. And I wasn't going to let a man or any other man try to touch me." Tr. at 16. He also testified that he never told investigators from the D.A.'s Office that he had been assaulted by Doe. He said that he told investigators details about Doe's apartment, but that was because he had been inside the apartment on other occasions with his friend Edward and his cousin Abamelay.

On cross examination, Edwin testified that he did not previously know that Doe had a preference for young men. He instead explained that Doe had been a positive role model for him, and wanted him to stay in school and out of trouble. He also responded on cross-examination that Doe never gave him any presents, not even on his birthday.


Question: . . . Your family didn't have a lot of money?

Edwin: No, sir.

Question: What did [Doe, Bill O'Rourke, Rich Federowicz] do for you?

Edwin: They didn't do — they didn't do anything. It wasn't like I went out with them andI — I'm supposed to like let them touch me or anything like that. No. It's just I met them, you know. And ever since I met [Doe], he's been like a . . . friend figure, because he's always . . . cared about like the things I've done, you know. Like he didn't want me to get into no trouble. He wanted me to stay in school.

Tr. at 29.

The district court found Edwin to be forthright, direct and responsive in answering the questions posed. No motivation appears from the record that would justify an inference that he was lying. He left the stand with a strong impression on the part of this Court that he was a truthful witness, that he had socialized with [Doe] and some of his friends, but had not personally engaged in any sexual activity with [Doe] or any other man.

Doe, No. 01 Civ. 10782, at 10.

This assessment of Edwin's testimony was not clear error. It is without dispute that Edwin had no discernable motive to lie. This is, of course, notwithstanding respondents's best efforts to disparage Edwin's character and motives. Respondents suggested that Edwin was testifying because he was homophobic, and was uncomfortable with the insinuation that he was the victim of a homosexual attack. To this Edwin responded that homosexuality does not bother him, and that his brother was in fact gay. Respondents likewise suggested that Edwin had a lengthy criminal history. However, Edwin's criminal record consisted of a conviction for a domestic dispute, an arrest for not paying the fare when riding the subway, and an arrest for possessing a paint-ball gun. Respondents accused Edwin of testifying out of loyalty to Doe, because in the past Doe had been a friend to him. To this Edwin responded that he did feel loyal to Doe, because "he never did anything wrong to me." Tr. at 30. Finally, Edwin testified that he had not been promised anything nor was he getting any benefit in exchange for his testimony. Instead, he stated simply that "I feel that the truth needs to be told, and I'm going to tell the truth. That's all." Tr. at 52-53.

The majority's rejection of Edwin's testimony rests on an adverse factual inference, contrary to that permissibly drawn by the trial judge. It discredits Edwin's testimony based on its speculation that Edwin misrepresented his relationship with the "Boy Lovers" group and that this relationship was actually a close one. The majority anchors this conclusion solely on the fact that Edwin responded, "yeah" when asked whether he knew that Angel was "Bill O'Rourke's boy?" This answer does not indicate when Edwin first learned of this information or whether he was aware of the implications of his response. Additionally, the majority relies on the fact that investigator Pat Storino testified that he received a telephone call from FBI Agent Kyle on or about February 10th or 12th, 1994 indicating that "Edwin had been reaching out, thr[ough] Bill O'R[ourke] and others, to see [Doe]. Edwin and Edward [R.] spoke with O'R[ourke] and were coached on what to say to thwart the investigation." Op. at 171. This factual contention, however, had a great potential to discredit Edwin's testimony. It seems strange that respondents did not directly question Edwin or Doe about this significant issue.

Moreover, the majority admits that "Storino's testimony on this issue and his notes rest on the hearsay statements of O'Rourke and Kyle." Op. at 170 n. 20. Relying on Schlup, the majority dismisses this concern, finding that it may properly consider Storino's testimony in assessing Doe's claims of actual innocence. The majority, however, loses sight of the fact that the standard for determining actual innocence is whether a reasonable juror would have found it beyond a reasonable doubt that the petitioner was actually innocent. Schlup, 513 U.S. at 327, 115 S.Ct. 851. As discussed supra, Section II, under Bousley, evidence that would be inadmissible at trial should not be considered in an actual innocence determination. Accordingly, since Storino's testimony regarding what O'Rourke told FBI Agent Kyle would not have been admissible, see Fed.R.Evid. 802, this court should not use this evidence as a basis for discounting Edwin's testimony.

Finally, a potential to confuse the boys involved was great. There were numerous similarly aged boys involved with the "Boy Lover" group, and one of the main boys' name was Edward R., who was also known as Eddie R., and who lived across the street from Edwin. Coincidentally, both Edwin's and Edward's mother was named Gladys, a fact that took respondents by surprise at the hearing.

Respondents were also confused regarding the identity of other boys. For example, at one point during the hearing, respondents confused Edwin and Edwin's brother, Xavier. On cross examination, respondents asked Edwin whether he ever represented himself to be Xavier, and Edwin responded: "No. That's my brother." Respondents then asked, "That's your brother?" Edwin replied, "Yes, sir." Tr. at 24.

I am not convinced that the two events cited by the majority are sufficient to render Edwin's testimony "internally inconsistent." See Anderson, 470 U.S. at 575, 105 S.Ct. 1504. Accordingly, it is not appropriate for this Court to second guess the trial judge's assessment of Edwin's testimony and to substitute its view of the evidence "as a whole" for that of the lower court's.

(c) Doe's Testimony

Doe testified that he did not have any physical or sexual contact with Edwin. He testified that, contrary to Edwin's testimony, he was alone with Edwin once at his apartment and the two were there for less than 10 minutes before they left for an arcade. He testified that, while there, he did not attempt to forcibly have sex with Edwin. The district court observed that "Doe's emphatic denial of guilt in connection with [Edwin] presents a difficult call for this [c]ourt." The court then concluded that Doe's testimony, that he had never assaulted Edwin, "has the ring of truth." Doe, No. 01 Civ. 10782, at 19.

The majority disbelieves Doe's protestations of innocence, finding his testimony to be self-serving and lacking in credibility. As support for its position, the majority points to three discrepancies in Doe's testimony. First, it points to the fact that Doe attempted to minimize his pedophilic tendencies during his testimony before the district court. According to the majority, Doe testified that he socialized with Edwin and other boys merely because he enjoyed doing so, and not because he was sexually involved with them. The majority notes that, although Doe testified that "he was not aware that Bagarozy was actually having sex with Edward R. and other boys," in truth, he was "enmeshed in the Boy Lovers' activities and was close friends with Bagarozy, with whom he shared underage sex partners, exchanged child pornography, and discussed methods of seducing young boys." Op. at [35-36]. It is, however, without dispute that Doe was a pedophile. The majority's observation that Doe may have been involved with other boys is not evidence that Doe necessarily assaulted Edwin in the summer of 1993, the crime for which he was convicted.

Second, the majority contends that Doe's testimony was unreliable because it contravened statements made by him during the September 15, 1993 proffer session. There is, however, some evidence to suggest that Doe may have exaggerated his relationship to Edwin and Edwin's alleged relationships with other "Boy Lovers" during the proffer session. Although the notes memorializing this meeting are extensive, Doe only mentions Edwin in one sentence where he claims to have also had sex with Edwin "at his residence just in the last couple [of] months and the sex included oral copulation and masturbation." Appendix for the Respondent-Appellee ("RA") 267. Significantly, in subsequent federal proffer sessions, Doe was not questioned about Edwin. See Laura Kaplan Aff. ¶ 5 (June 8, 2000) ("[a]t no time during the October 4, 1993 proffer session did the defendant discuss his relationship with Edwin ."); see also James A. Kyle Aff. ¶ 6 (June 16, 2000) (the Oct. 4, 1993 proffer session "did not include any discussion of any relationship between [Doe] and a boy named `Edwin.'").

Although the majority held that there is no indication in the record of what Doe stated in the October proffer session, see Op. at [9 n. 8], affidavits submitted by AUSA Laura Kaplan and investigator Kyle clearly indicate that Edwin was not discussed at the second proffer session.

Moreover, even though Doe supposedly claimed at the September 15, 1993 proffer session that there were rumors that Edwin was having sex with O'Rourke, and that Edwin was also "one of Richard Bagarozy's `main boys,'" RA 267. Laura J. Kaplan, the assistant United States Attorney ("AUSA") for the District of New Jersey who was the lead prosecutor in the Bagarozy/Federowicz case, acknowledged that "Edwin . . . was not one of the children involved in the case being investigated as part of the Bagarozy/Federowicz case." Kaplan Aff. ¶ 5 (June 8, 2000). Further, Edwin was not mentioned in Doe's testimony in the Bargarozy trial even though Doe discussed several boys that he, Bagarozy and Federowicz had assaulted. Thus, the fact that Edwin was never again mentioned, aside from the one mention at the first proffer session, indicates that the prosecution found Doe's initial assertions regarding his relationship with Edwin, and Edwin's relationship with other Boy Lovers, to be tenuous. Thus, it is equally plausible that Doe exaggerated his relationship with Edwin at the initial proffer session in order to gain the benefits of cooperation.

The majority further argues that Doe's proffer statements must have been truthful because Doe had no motive to lie about having sex with Edwin. The focus of this meeting was to ascertain Doe's knowledge of the other Boy Lovers's activities. Doe's personal conduct "was at most an ancillary issue." Op. at 167. The majority concludes that Doe must have mentioned having sex with Edwin because it was true. The majority's speculation regarding Doe's supposed motive for discussing Edwin is dubious. There is no evidence to suggest that Doe did not feel the need to implicate himself. To the contrary, notes from the proffer session were replete with self-incriminating information, such as Doe's statement that he paid two young boys to pose naked for photographs, that he traveled to Baltimore, Maryland in 1989 and 1990 to engage in sex with young boys, and that he took boys to a certain hotel frequented by Boy Lovers. Indeed, it is equally likely that he mentioned these events as an attempt to exaggerate his experience as a Boy Lover, which thereby increased his usefulness to the prosecution, rather than volunteered information because of some desire to confess all of his past deeds.

Third, the majority contends that Doe's testimony was unreliable because it contravened statements made by him during his guilty plea allocution in February 1995. The fact that Doe admitted to sodomizing Edwin in order to obtain the benefits of a plea bargain is not conclusive evidence of guilt. The district court correctly acknowledged that, in the real world, "people yield their honor and principles to overwhelming force and will say, even under oath, what satisfies their instinct for survival." Doe, No. 01 Civ. 10782, at 17. As noted by the district court, in this case, if Doe accepted the plea, he would have been sentenced to a non-custodial sentence of probation. On the other hand, if he did not accept the plea, or if his plea was rejected for failure to allocute, he could not testify in the federal Newark case and he would have been charged and tried in Westchester for the more serious charge of forcible sodomy, a Class B violent felony, which would subject him to a prison sentence of up to 25 years. The likelihood that he would be convicted if he failed to plead was great. He was allegedly informed by Elizabeth Cronin that she had a written statement from the victim and that the victim intended to testify against him. Murray Richman, Doe's attorney during these events, testified that Cronin "definitely" represented to him that she had a signed statement from the alleged victim. Tr. at 18-19. Cronin denies that this representation was made. Further, Richman most likely informed Doe that a child's testimony regarding a sexual assault can be devastating and irrebutable. Further, individuals convicted of child molestation have a difficult time in prison. Thus, as found by the district court, "[f]aced with the alternatives known to him, and as might be expected, [Doe] decided to accept the highly favorable plea bargain and to say whatever he needed to say in order to do so." Doe, No. 01 Civ. 10782, at 18.

Elizabeth Cronin is currently the Director of Legal Affairs for our Court of Appeals. In my memorandum of December 22, 2003 to the Special Committee on Conflicts, I expressed my unease over this Court's retention of jurisdiction over this case because, in considering Doe's petition, we are required to rule on the propriety of past prosecutorial conduct by Cronin, a senior member of our staff. RSP MEM. (Dec. 22, 2003); Op. at [45-46]. We are also required to assess the credibility of her recent testimony regarding these events before the district court. These circumstances may appear to create a conflict of interest. The committee considered at length these concerns and concluded that the conflict was not so severe as to necessitate our recusal. REPORT OF THE AD HOC COMMITTEE ON CONSIDERATION of Doe v. Menefee No. 03-2432 (Feb. 4, 2004).

Moreover, Doe persuasively argues that AUSA Kaplan pressured him to take the plea. Doe testified that he informed Kaplan that he was averse to the idea of pleading guilty because, "not only hadn't I committed the underlying crime, but I have never in my life done anything like take somebody and throw them down on a bed or have any violent connection with them involving . . . sex, ever." Tr. at 124. He claims that AUSA Kaplan responded, "You have a problem. You told them in the federal proffer that you had had some contact, and you, in that case, will have to either go back to that or you will have to take the Fifth Amendment on the stand. If you take the Fifth Amendment on the witness stand during my trial . . . I will not be a happy camper." Tr. at 124. Thus, based on these considerations, Doe entered the plea on February 6, 1995, one day before he was originally scheduled to testify in the federal case. Accordingly, the district court's decision to credit Doe's testimony was not so unreasonable that this Court is compelled to reverse.

(d) The Respondents's Case

The majority holds that it was patently unreasonable for the district court to have credited Edwin's and Doe's testimony because their testimony was contradicted by the testimony of former employees of the D.A.'s Office, Elizabeth Cronin, Greg Bernhard, and Pat Storino. There are, however, four reasons why this evidence was not so compelling that the district court's decision to credit Edwin's and Doe's testimony was not clear error.

First, the D.A.'s Office's case against Doe was poorly developed. The only evidence supporting Doe's guilty plea, aside from Doe's own statements, was the recollection of members of the D.A.'s office regarding what a 12-year-old boy had told them during one or two interviews. Cronin's interview with Edwin was not tape recorded or otherwise memorialized except for her handwritten notes. Further, the same alleged victim now testifies that he never made these statements to investigators and that the underlying crime never occurred. Thus, this is certainly not the case where the district court, in choosing to credit the victim's testimony, turned its back on a wealth of damming evidence.

Second, when Cronin prosecuted this case, she had not sufficiently corroborated Edwin's alleged statement that he had been assaulted by Doe. On direct examination, she was asked whether she took any steps to corroborate the details of what Edwin told her. She responded that Edwin had also given her information such as "what [Doe's] apartment looked like, what [his] car looked like, and . . . what [Doe] looked like." Tr. at 157. These facts, however, establish very little. It is without question that Edwin was familiar with Doe and had been inside his car and apartment on several occasions. The fact that Edwin knew what Doe looked like hardly establishes the issue of whether Doe sexually assaulted him.

On cross-examination, when again asked "what other independent evidence did you have at the time to substantiate or corroborate the statements of Edwin," Cronin responded "Well, I wouldn't have had to corroborate his statements, because he was old enough to testify." Tr. at 182. She then responded that, aside from Edwin's description of Doe, Doe's apartment, and Doe's car, she had wiretaps of conversations by "men in this group about Edwin." Tr. at 183. Significantly, the wire tap evidence primarily consisted of conversations by other individuals involving what Doe had said, or what they speculated was Doe's relationship with Edwin. Again, much of this evidence consisted of hearsay testimony that would not have been admissible at trial. Thus, a reasonable juror could not consider this evidence in evaluating Doe's claim of actual innocence.

As a final matter, certain aspects of Cronin's testimony regarding what Edwin allegedly told her during the two interviews are inconsistent. She claims that Edwin told her that he had been violently assaulted by Doe on two occasions at his apartment. The first time, Doe forcibly threw Edwin onto his bed, Edwin allegedly struggled against him, and shouted "I want to go, I want to go." Tr. at 151-53. Doe responded, "You have no other way to get home, because I brought you here." Id. at 153. Approximately one or two months later, Edwin returned to Doe's apartment. On this second occasion, Doe supposedly attacked Edwin again, and this time, put his mouth on Edwin's penis. It is unclear, however, why Edwin would return to Doe's apartment or even continue associating with Doe after the first violent attack occurred. Indeed, even after the second attack, Edwin continued to go to places with Doe such as the arcade. When asked on cross-examination whether Cronin thought it was odd that, after Edwin had been violently attacked, he would continue to meet with Doe, Cronin responded, "I wouldn't say I considered it odd, but I did consider it something that I needed to talk to him about." Tr. at 183.

Third, the time line of the alleged events is also suspicious. Cronin testified that Edwin met Doe "sometime in May [1993], and then the sexual stuff didn't actually happen until July or August." Tr. at 185. Edwin's introduction to Doe in May 1993 was corroborated by Cronin's notes. Thus, Edwin had associated with Doe for a short amount of time before the alleged assault occurred. Moreover, during July and early August of 1993, Doe was traveling with his chorus group in Europe, a fact attested to in the Bagarozy/Federowicz trial in 1995, and again at the 2003 hearing before the district court. This unappreciated fact has two serious implications: First, prior to the alleged assaults, Edwin had not associated with Doe for a continuous four-month period. Given the brief amount of time that Edwin had known Doe, it seems unlikely that Edwin would have developed such a strong sense of loyalty or comfort level with Doe that he would have returned to Doe's apartment soon after he was initially attacked, would meet with Doe at an arcade after the second attack, and then would lie on the witness stand for Doe several years later. Second, and more importantly, the fact that Doe was away from New York in July and early August of 1993 is concerning because he plead guilty to assaulting Edwin "on or about and between July 1st of 1993 and August 13th of 1993." Op. at [11]. Thus, the fact that Doe may have been out of the country when the alleged attacks occurred further supports Doe's claims of innocence.

Fourth, the district court simply did not find Cronin's testimony to be credible. It stated:

Without making any adverse credibility finding (as requested by [Doe]) I conclude that where [Edwin's] testimony differs as to material facts, his version is entitled to greater weight.

Doe, No. 01 Civ. 10782, at 11. While the majority discounts Doe's testimony as self-serving, at the same time, it ignores the fact that Cronin's testimony was likewise tinged with self-interest. Doe accuses Cronin of committing misconduct in the prosecution of his case. He claims that she actively misled his attorney into believing that there was a signed victim's statement when there was none. He also claims that the D.A.'s Office made assurances that if he cooperated with Westchester and federal law enforcement agents, he would not be prosecuted by the D.A.'s Office. Cronin therefore had an incentive to defend the choices that she made while prosecuting this case. Because her testimony could likewise be viewed as self-serving, it is not for this Court to decide which testimony is more believable on this basis.

The district court further found that this "case was but one of many sordid situations involving children dealt with by Ms. Cronin during her lengthy public service in the office of the District Attorney. Her recollection is based primarily on her notes, at least some of which are not contemporaneous records." Id. The majority dismisses this observation as "pure speculation," although it admits that only four out of seven pages of notes were dated. Op. at 172. It contends that "there was no testimony that Cronin routinely dealt with similar accusations." Op. at 172. Cronin, however, testified that she was employed at the D.A.'s Office for 14 years, and for 12 of those years, she was assigned to the "Special Prosecutions Division." She describes her position as the "Assistant D.A. in that office handling domestic child abuse cases." Accordingly, because it is clear that she routinely dealt with child abuse cases, the majority's discounting of the district court's view of Cronin's testimony on this basis was likewise improper.

The majority further suggests that the district court erred in failing to acknowledge the fact that Cronin's testimony was "corroborated in full by the testimony of ADA Bernhard." Op. at 172-73. There was, however, a seeming discrepancy between Cronin's and Bernhard's testimony. Bernhard, who was present at Cronin's interviews with Edwin, testified about a single assault, while according to Cronin's testimony and her notes, Edwin was allegedly assaulted twice. When asked on direct examination whether Bernhard remembered "how the coerced aspect of the conduct first came to be?" he responded that Edwin told him and Cronin that Doe took Edwin to his apartment, began "grabbing at [Edwin's] legs . . . [and that Doe] eventually pulled [Edwin] off of the chair and threw him on a bed and pulled down his pants and put his mouth on Edwin's penis." Tr. at 198-99. The majority attempts to explain this discrepancy by stating that "Bernhard was asked only about Edwin's second visit to the apartment, from which the sodomy charge arose." Op. at 158 n. 14. This characterization of the question posed to Bernhard is inaccurate. Bernhard first discussed the assault, and then a few questions later, respondents returned to this line of questioning and made clear that Bernard was describing the second attack. Unfortunately, neither party sought to clarify this ambiguity.


Question: Okay. Do you remember how the coerced aspect of the conduct first came to be? Was it suggested by you or Ms. Cronin or stated by the boy for the first time?

Bernhard: It was Edwin.

Question: He stated it himself, before —

Bernhard: He didn't use the word force. But Edwin was the one who was telling us that [Doe] invited him into his apartment to use a computer game, and that, while he was on the computer, [Doe] was grabbing at his legs and trying to turn him away from the computer game, and that he eventually pulled him off of the chair and threw him on a bed and pulled down his pants and put his mouth on Edwin's penis.

Question: Okay. For clarification's sake, is it true that you're basing your testimony here today on a combination of your own recollection and your review of notes prepared that you understand to be prepared by Ms. Cronin?

. . .

Question: So, in essence, just tell us what you understand — you recall Edwin to have said in regard to the second time that he went to [Doe's apartment]?

Bernhard: That he, in fact, was invited up to use the video game, to see a computer game, actually, and that [Doe] kept grabbing him, and he told him to cut it out, "Stop grabbing my leg.". . . that [Doe] put his mouth on his penis after [Doe] pulled [Edwin's] pants down.

Tr. at 198-99.

* * * * * *

I conclude that the district court did not err in crediting Edwin's and Doe's testimony. Adopting the district court's credibility findings, I find that Doe has presented new reliable evidence of actual innocence and that it is more likely than not "that no reasonable juror would have found [Doe] guilty." Schlup, 513 U.S. at 329, 115 S.Ct. 851. No reasonable juror would have found Doe guilty beyond a reasonable doubt where Edwin, the alleged victim, credibly testified that no criminal act occurred, and no damning, physical evidence or eyewitness testimony exists to contradict this testimony.

IV. AEDPA TOLLING

Had I been able to convince the majority that this case presented a credible claim of actual innocence, this panel would then have taken up the issue of whether AEDPA's one year statute of limitations is tolled. Because the majority does not agree that this case does present a credible claim of actual innocence, I decline to consider that issue as a panel of one, as tempting as that opportunity may be, and leave for another day the resolution of this difficult issue.

V. ATTORNEY INCOMPETENCE

I do not dispute the majority's affirmance of the district court's conclusions with respect to the attorney incompetence issue. For the reasons stated by the majority, I likewise find that Doe did not exercise reasonable diligence during the period he seeks to toll.

CONCLUSION

While there are conflicts between Doe's and Edwin's testimony on one hand, and Cronin's, Storino's, and Bernard's testimony on the other, I believe the district court did not clearly err in finding Doe's and Edward's testimony to be more credible. It is inappropriate in all but the most extraordinary cases for this Court to second-guess a district court's credibility findings. Nothing in the record convinces me that Cronin's, Bernhard's and Storino's testimony was irrefutably more credible. The majority's dissection of the district court's decision departs from our precedents and wrongly supplants the lower court's assessment of the evidence with its own factual inferences, never having seen or heard any of the testimony that it now seeks to discredit. Edwin's testimony was not so inconsistent with other evidence in the record that this Court is compelled to discredit it. Accordingly, I must dissent.


An alternative to Lexis that does not break the bank.

Casetext does more than Lexis for less than $65 per month.