COMMONWEALTH OF MASSACHUSETTS HAMPDEN, ss. COMMONWEALTH v. GEORGE PERROT SUPERIOR COURT DEPARTMENT OF THE TRIAL COURT ) ) ) ) ) ) ) Docket Nos. 85-5415, 16, 18, 20, and 25 MEMORANDUM IN SUPPORT OF DEFENDANT’S MOTION FOR A NEW TRIAL TABLE OF CONTENTS SUMMARY OF ARGUMENT........................................................................................................ STATEMENT OF THE FACTS....................................................................................................... I. THE EVENTS OF NOVEMBER 30, 1985............................................................................. A. The Lichwala Break-in..................................................................................................... B. The Prekop Break-in And Sexual Assault......................................................................... II. PERROT’S ARREST, INTERROGATION, AND THE RECOVERY OF OTHER EVIDENCE.............................................................................................................................. A. Perrot Is Interrogated For 12 Hours Having Barely Slept The Previous Night................ B. Perrot Purportedly Signs A Partial “Confession” To The Lichwala And Prekop Break-ins........................................................................................................................... C. None Of The Victims Identifies Perrot During A Highly Suggestive Investigative Lineup............................................................................................................................... D. Other Evidence Is Gathered By The Police, But Is Flawed And Subject To Conflicting Testimony Regarding Its Origin And Relevance......................................... III. PERROT’S TRIAL................................................................................................................ A. Prekop Explains Why Perrot Was Not Her Attacker...................................................... B. Wayne M. Oakes’s Microscopic Hair Analysis.............................................................. C. William G. Eubank’s Serological Protein Profiling........................................................ PROCEDURAL HISTORY........................................................................................................... I. PERROT’S FIRST TRIAL..................................................................................................... A. Perrot’s First Conviction Is Reversed By The Massachusetts SJC................................. B. The District Attorney Fabricates And Forges A Statement Falsely Attributed To Perrot, Attempting To Tie Perrot To The Serial Rapes Apparently Committed By A Single Perpetrator............................................................................................................ II. PERROT’S SECOND TRIAL AND POST-CONVICTION LITIGATION.......................... A. Perrot Is Convicted Again, But The Superior Court Orders Another New Trial............ B. The Court Of Appeals Reverses And Vacates The Order For A New Trial.................... C. The Superior Court Grants DNA Testing........................................................................ ARGUMENT................................................................................................................................. I. THE FBI’S ADMISSION DEMONSTRATES THAT PERROT WAS CONVICTED USING FALSE AND MISLEADING FORENSIC TESTIMONY THAT HE WAS THE SOURCE OF THE CRIME-SCENE HAIR.................................................................. A. The FBI Has Admitted That Its Hair Microscopy Testimony Exceeded The Limits Of Science, Reversing Its Decades-Long Practice Of Testifying That Criminal Defendants Were The Source Of Evidentiary Hairs....................................................... 1 B. The FBI’s Admission Demonstrates That Oakes’s Testimony And The Prosecutor’s Arguments Linking Perrot To The Evidentiary Hair Exceeded The Limits Of Science........................................................................................................... 1. Type I Errors.......................................................................................................... 2. Type II Errors......................................................................................................... 3. Type III Errors........................................................................................................ 4. Other Errors In Oakes’s Testimony........................................................................ 5. Oakes Repeatedly And Deliberately Undermined His “Disclaimer” That Hair Is Not A Unique Identifier............................................................................. II. PERROT IS ENTITLED TO A NEW TRIAL FREE FROM THE NEWLY DISCREDITED HAIR MICROSCOPY TESTIMONY THAT WAS NEVERTHELESS PRESENTED TO THE JURY AS IRREFUTABLE “SCIENTIFIC” EVIDENCE OF GUILT.................................................................................................................................... A. The NAS Report And FBI Review Of Hair Testimony Are Newly Discovered Evidence Providing Objective Scientific Standards For Hair Microscopy That Was Previously Unavailable........................................................................................... B. The Introduction Of The FBI’s False And Misleading Testimony Violated Perrot’s Due Process Rights......................................................................................................... C. Perrot’s Counsel Provided Ineffective Assistance By Failing To Adequately Investigate and Challenge The Commonwealth’s Hair Microscopy Testimony, And By Failing To Properly Argue The Voluntariness of Perrot’s Statement On The Night Of His Arrest.................................................................................................. D. The Erroneous Hair Testimony Was Material To Perrot’s Conviction........................... E. 1. Scientific Evidence Is Particularly Difficult For A Jury To Evaluate.................... 2. The Prosecution’s Evidence Against Perrot Was Insufficient To Sustain A Conviction.............................................................................................................. 3. The Police And Prosecutorial Misconduct Evident In The Record Call Into Question The Chain Of Custody For All Of The Evidence In Perrot’s Case......... The Prosecution’s Error-Ridden Closing Argument, Repeatedly Referring To Oakes’s Hair Testimony, Sealed Perrot’s Fate................................................................ CONCLUSION.............................................................................................................................. 2 TABLE OF AUTHORITIES Page(s) CASES Alcorta v. Texas, 355 U.S. 28 (1957)................................................................................................................... Commonwealth v. Adams, 374 Mass. 722 (1978).............................................................................................................. Commonwealth v. Alvarez, 433 Mass. 93 (2000)................................................................................................................ Commonwealth v. Amirault, 424 Mass 618 (1997)............................................................................................................... Commonwealth v. Baker, 440 Mass. 519 (2006).............................................................................................................. Commonwealth v. Bennett, 43 Mass. App. Ct. 154 (1997).................................................................................................. Commonwealth v. Buck, 64 Mass. App. Ct. 760 (2005)......................................................................................43, Commonwealth v. Burkett, 396 Mass. 509 (1986).............................................................................................................. Commonwealth v. Ciampa, 406 Mass. 257 (1989).............................................................................................................. Commonwealth v. Conaghan, 433 Mass. 105 (2000).............................................................................................................. Commonwealth v. DiGambattista, 813 N.E.2d 516 (Mass. 2004).................................................................................................. Commonwealth v. Farley, 432 Mass. 153 (2000).............................................................................................................. Commonwealth v. Gilday, 382 Mass. 166 (1980).............................................................................................................. Commonwealth v. Grace, 397 Mass. 303 (1986)..................................................................................................42, Commonwealth v. Griffith, 45 Mass. App. Ct. 784 (1998).................................................................................................. 3 Commonwealth v. Haggerty, 400 Mass. 437 (1987).............................................................................................................. Commonwealth v. Hoose, 467 Mass. 395 (2014).............................................................................................................. Commonwealth v. Hurst, 364 Mass. 604 (1974)........................................................................................................ Commonwealth v. Kane, 19 Mass. App. Ct. 129 (1984).................................................................................................. Commonwealth v. LeFave, 430 Mass. 169 (1999).............................................................................................................. Commonwealth v. Lykus, 451 Mass. 310 (2008)............................................................................................43, 44, Commonwealth v. Mosby, 11 Mass. App. Ct. 1 (1980)...................................................................................................... Commonwealth v. Ortiz, 84 Mass. App. Ct. 258 (2013).................................................................................................. Commonwealth v. Perrot, 38 Mass. App. Ct. 478 (1995)............................................................................................ Commonwealth v. Perrot, 407 Mass. 539 (1990)....................................................................................................... Commonwealth v. Perrot, 441 Mass. 1104 (2004)............................................................................................................. Commonwealth v. Perrot, 58 Mass. App. Ct. 1102 (2003)..............................................................................24, 42, Commonwealth v. Perrot, No. 85-5415, slip op. (Hampden Cnty. Super. Ct., Sept. 11, 2001)................................. Commonwealth v. Rembiszewski, 391 Mass. 123 (1984).............................................................................................................. Commonwealth v. Rondeau, 378 Mass. 408 (1979).............................................................................................................. Commonwealth v. Saferian, 366 Mass. 89 (1974).......................................................................................................... 4 Commonwealth v. St. Germain, 381 Mass. 256 (1980).............................................................................................................. Commonwealth v. Sullivan, 410 Mass. 521 (1991).............................................................................................................. Commonwealth v. Watson, 409 Mass. 110 (1991)............................................................................................................... Daubert v. Merrell Dow Pharm., 509 U.S. 579 (1993)................................................................................................................. DeChristoforo v. Donnelly, 473 F.2d 1236 (1973), rev’d on other grounds, 416 U.S. 637 (1974)..................................... Estelle v. McGuire, 502 U.S. 62 (1991)................................................................................................................... Fisher v. Gibson, 282 F.3d 1283 (10th Cir. 2002)................................................................................................ Giglio v. United States, 405 U.S. 150 (1972)................................................................................................................. In re Pers. Restraint Trapp, 165 Wash. App. 1003 (2011).............................................................................................. In re Winship, 397 U.S. 358 (1970)................................................................................................................. Kyles v. Whitley, 514 U.S. 419 (1995)................................................................................................................. Manning v. Mississippi, 2013-DR-00491-SCT (Miss. S. Ct.)........................................................................................ Matter of Bloom, 9 Mass. Att’y Discipline Rep. 23 (1993)................................................................................. Napue v. Illinois, 360 U.S. 264 (1959)................................................................................................................. Pavel v. Hollins, 261 F.3d 210 (2d Cir. 2001)..................................................................................................... Poole v. Woods, No. 08-cv-12955, 2011 WL 4502372 (E.D. Mich. Aug. 9, 2011)........................................... 5 Sivak v. Hardison, 658 F.3d 898 (9th Cir. 2011).................................................................................................... Smith v. Commonwealth, 331 Mass. 585 (1954).............................................................................................................. State v. Young, 35 So. 3d 1042 (La. 2010)....................................................................................................... Strickland v. Washington, 466 U.S. 668 (1984)................................................................................................................. United States ex rel. Fergerson v. Atchison, No. 09 C 2105, 2012 WL 581163 (N.D. Ill. Feb. 22, 2012).................................................... United States v. Addison, 498 F.2d 741 (D.C. Cir. 1974)................................................................................................. United States v. Bagley, 473 U.S. 667 (1985)................................................................................................................. United States v. Cronic, 466 U.S. 648 (1984)................................................................................................................. United States v. Frazier, 387 F.3d 1244 (11th Cir. 2004)................................................................................................ United States v. Lester, 254 F. Supp. 2d 602 (E.D. Va. 2003)....................................................................................... United States v. Willock, 696 F. Supp. 2d 536 (D. Md. 2010)......................................................................................... Wilson v. Beard, 426 F.3d 653 (3d Cir. 2005)..................................................................................................... STATUTES AND RULES 28 U.S.C. § 2255...................................................................................................................... Mass. R. Crim. P. 30(b).................................................................................................................. Mass. Gen. Law Ch. 278A............................................................................................................. OTHER AUTHORITIES Yamil Berard, Forensic Science Commission to Review Convictions Based on Hair Samples, Star-Telegram, Aug. 11, 2013................................................................................... 6 John M. Butler, Fundamentals of Forensic DNA Typing (2010)................................................... Department of Commerce Economic and Statistics Administration Bureau of the Census, 1990 Census of Population General Population Characteristics Massachusetts (April 22, 1992), http://www.census.gov/prod/cen1990/cp1/cp-123.pdf....................................................................................................................................... Keith A. Findley, Innocents at Risk: Adversary Imbalance, Forensic Science, and the Search for Truth, 38 Seton Hall L. Rev. 893 (2008)................................................................ Forensic Science Communications, Part 1: Hair Evidence (July 2000), available at http://www.fbi.gov/about-us/lab/forensic-sciencecommunications/fsc/july2000/deedric1.htm............................................................................ B.D. Gaudette, Some further thoughts on probabilities and human hair comparisons, 23 J. of Forensic Sciences 758 (1978).......................................................................................... Mark A. Godsey & Marie Alao, She Blinded Me with Science: Wrongful Convictions and the “Reverse CSI Effect,” 17 Tex. Wesleyan L. Rev. 481 (2011)............................................ M.M. Houck & B. Budowle, Correlation of Microscopic and Mitochondrial DNA Hair Comparisons, 47 J. Forensic Sci. 1 (2002)........................................................................ Spenser S. Hsu, Convicted Defendants left uninformed of forensic flaws found by Justice Department, Wash. Post, April 16, 2012............................................................................ Innocence Project, Innocence Project and NACDL Announce Historic Partnership with the FBI and Department of Justice on Microscopic Hair Analysis Cases (July 18, 2013)........................................................................................................................................ Saul M. Kassin et al., The Forensic Confirmation Bias: Problems, Perspectives and Proposed Solutions, 2 J. of Applied Research in Memory & Cognition 42 (2013)................. Helayne Lightstone, Judge cuts bail for teenager charged with rapes of elderly, The Morning Union, Jan. 1, 1986................................................................................................. Matter of Bloom, 9 Mass. Att’y Discipline Rep. 23 (Dec. 16 1993)............................................. Elizabeth Norton, Identifying the Brain’s Own Facial Recognition System, Science, Oct. 23, 2012.................................................................................................................................... Cary T. Oien, Forensic Hair Comparison: Background Information for Interpretation, 11 Forensic Sci. Comm’ns (2009).......................................................................................... Nancy Peccin, Life sentence given in rape, murder plea, The Morning Union, April 7, 1987............................................................................................................................................ Norman L. Reimer, The Hair Microscopy Review Project, The Champion, July 2013................ 7 Tom R. Tyler, Viewing CSI and the Threshold of Guilt: Managing Truth and Justice in Reality and Fiction, 115 Yale L.J. 1050 (2006)....................................................................... Richard H. Underwood, Evaluating Scientific and Forensic Evidence, 24 Am. J. Trial Advoc. 149 (2000)................................................................................................................... M.S. Verma et al., Hair-MAP: A prototype automated system for forensic hair comparison and analysis, 129 Forensic Sci. Int’l 168 (2002)................................................. 8 SUMMARY OF ARGUMENT George Perrot has spent nearly 29 years in prison for crimes that he did not commit, convicted because the prosecution and an FBI examiner misled the jury into believing that he was the source of a single hair recovered from the crime scene. Perrot has sought at every juncture to clear his name through whatever means were available to him. The Superior Court has already granted Perrot a new trial because of a host of serious prosecutorial errors, but the Court of Appeals overturned the granting of a new trial because it wrongly believed that microscopic hair evidence corroborated Perrot’s guilt. In 2013, however, a critical breakthrough in Perrot’s case occurred when the FBI publicly conceded for the first time that its microscopic hair analysts had no scientific basis to testify that crime-scene hairs came from or likely came from particular criminal defendants. Because hair microscopy testimony was critical in securing and upholding Perrot’s wrongful convictions, he now moves for a new trial under Massachusetts Rule of Criminal Procedure 30(b) to rectify the grave injustice of incarcerating an innocent man based on outdated and unreliable forensics. Hair microscopy filled a critical gap in the prosecution’s weak case against Perrot, providing the only physical evidence directly linking him to the crime scene of a rape perpetrated against an elderly woman. The victim insisted to police and prosecutors throughout the investigation and at trial that Perrot—with his shaggy hair, full mustache, and goatee—looked nothing like her attacker. In fact, this rape was one of a series of rapes perpetrated against elderly women in the Springfield area that investigators believed to be committed by the same person, and yet each elderly victim of the serial rapist refused to identify Perrot as her attacker. Apart from hardly conclusive serological evidence that put Perrot in a group of at least 14,000 white men in Hampden County alone who could have left the blood found at the crime scene, the primary evidence against Perrot was a partial, coerced “admission” that he had been to the crime 1 scene; Perrot did not admit to the rape. The circumstances under which the “admission” was elicited were ripe for false confession. Seventeen-year-old Perrot, unaccompanied by parents or counsel, has no recollection of signing this “confession” after 12 hours of nearly continuous interrogation, having slept no more than an hour or two in the prior 36 hours, suffering from extreme amounts of drug and alcohol consumption the previous evening, and having been beaten and harassed by police at the time of his arrest and in the car ride to the police station. Even more troubling is the grave prosecutorial misconduct by an Assistant District Attorney (“ADA”) in fabricating and forging an affidavit attributed to Perrot, actions that speak directly to the willingness of Perrot’s interrogation team to resort to unconstitutional tactics to coerce “confessions.” Perrot’s case is not an aberration. At least 74 exonerations have involved wrongful convictions attributable to flawed microscopic hair comparisons. FBI examiners for decades linked crime-scene hairs to individual defendants by testifying that “matching” hairs either came from or were highly likely to have come from individual defendants. In recent years, however, a growing chorus of scientific authorities—including a seminal 2009 report by the National Academy of Sciences—has criticized these attempts to peg a defendant as the source of a crimescene hair. Following the NAS Report, several exonerations of defendants who had been convicted based on the FBI’s microscopic hair analysis testimony compelled the FBI to publicly acknowledge for the first time in its history that its agents lacked any scientific basis when they routinely testified that an individual was or was likely the source of a crime-scene hair. The FBI now concedes that the very testimony that was used to secure Perrot’s conviction has no scientific basis, and by the FBI’s admission, if Perrot were tried today, the FBI agent’s hair microscopy testimony would not be admissible. 2 The FBI is now in the midst of an historic review of every conviction at both the state and federal level that included testimony by an FBI analyst linking a crime-scene hair and a criminal defendant. Thousands of defendants, including Perrot, are within the scope of this review, but its slow pace means it may be years before a recommendation is issued in Perrot’s case. Given the FBI’s admission that its examiners routinely provided false and misleading testimony and the conclusions of leading academic authorities, this Court should proactively examine the trial record and order a new trial for three independent reasons: First, the FBI’s acknowledgment that its examiners provided scientifically unsupported testimony is newly discovered evidence justifying a new trial. Second, the Commonwealth violated Perrot’s due process rights by presenting false and misleading testimony that the FBI knew or should have known lacked a scientific foundation. Third, Perrot’s defense counsel was constitutionally ineffective in failing to investigate and challenge the Commonwealth’s hair evidence and in failing to properly argue the voluntariness of Perrot’s alleged “confession.” STATEMENT OF THE FACTS1 In 1984 and 1985, a series of rapes targeted at least three women over the age of 60 in the Springfield area: 78-year old Mary Prekop, 82-year-old Mae Marchand, and 64-year old Emily Lichawala. The similarities between these crimes make it likely that they were the work of a single perpetrator. Prekop was raped three times in the span of a year and a half. The first time Prekop was raped in May 1984, Perrot was in juvenile custody at the Westfield Detention Center. 1 Citations to transcripts from Perrot’s first trial in December 1987 are denoted “First Trial Tr. __, __.” The transcripts from Perrot’s second trial in January 1992 are denoted “Tr. __, __.” The transcript of the December 13, 1985 Grand Jury proceedings is denoted “GJ, __.” The transcript of the September 2, 1986 Motion to Suppress hearing is denoted “MTS, __.” The transcript of the February 25, 1991 Motion to Dismiss hearing is denoted “MTD, __.” 3 See Ex. A, George Perrot Juvenile Record.2 And after the second and third rapes, Prekop told police that Perrot was not the man who attacked her. See, e.g., Ex. B, Olive Edwards Aff., Jan. 23, 2013. Nevertheless, Perrot was tried and convicted for the last of the serial rapes perpetrated against Prekop. Emily Lichwala was raped on October 24, 1985, GJ at 28, but she did not identify Perrot as her attacker in a police lineup, and a palm print found on Lichwala’s window also did not match Perrot, see Ex. C, Perrot Police Lineup Summ., Dec. 7, 1985; Tr. III, 461–62 (ADA telling the Court at side bar that there is “a palm print that’s recovered that they’re unable to match to anybody”). Mae Marchand was raped in the same general timeframe, and a semen stain removed from Marchand’s bed “exculpated Perrot as the donor of the stain.” MTD 29–30, 66 (statement of ADA Francis Bloom). The two earlier rapes of Prekop and the rapes of Marchand and Lichwala have apparently never been solved. Rapes of elderly women in the area did not end with Perrot’s arrest, despite the District Attorney’s (“DA”) contemporaneous pronouncements that he expected they would. On January 1, 1986, weeks after Perrot’s arrest and in response to public concern about the unsolved crimes, then-DA Matthew Ryan stated that “[w]e feel that this arrest [of Perrot] may very well result in the lessening of these types of offenses.” See Ex. D, Helayne Lightstone, Judge cuts bail for teenager charged with rapes of elderly, The Morning Union, Jan. 1, 1986. In reality, however, that same day the same newspaper reported that an 80-year-old woman, Ann O’Connor, had been raped and murdered in Springfield. See Ex. D, Helayne Lightstone, Slaying suspect to undergo tests, The Morning Union, Jan. 1, 1986. James Jones, who later pleaded guilty to the O’Connor crimes, was 22 years old at the time, see Ex. E, Nancy Peccin, Life sentence given in rape, murder plea, The Morning Union, April 7, 1987, placing him within the age range of 20 to 30 2 Based on a May 1, 1984 report of Sgt. David Louden, the perpetrator also left a 10.5 inch sneaker print on Prekop’s door, but Perrot wore only a size eight shoe at the time. 4 years old that Prekop believed her attacker to be, see Tr. I, 146. The DA’s Office nonetheless proceeded with its prosecution of Perrot for the Prekop rape of November 30, 1985. I THE EVENTS OF NOVEMBER 30, 1985 A. The Lichwala Break-in Lichwala lived alone at 33 Covel Street in Springfield, Massachusetts in a one-story house with a basement. Tr. I, 119–21. Sometime between 3:00 and 3:30 a.m. on November 30, 1985, Lichwala awoke to the sound of someone trying to break into her kitchen through a locked door leading to her basement. Id. at 120–21. She then heard footsteps down the stairs back to the basement, but she convinced herself that the sounds were just her imagination. Id. at 122. After five minutes of silence, Lichwala heard glass shatter near a breezeway door that also led to the kitchen. Id. Lichwala fled out the front door, never seeing the perpetrator, id., although she believed it was the same individual who raped her on a previous occasion, First Trial Tr. II, 251. B. The Prekop Break-in And Sexual Assault A little later that morning, Prekop woke in the early hours to the sound of her cocker spaniel barking. Tr. I, 129. She entered the kitchen and listened for a few minutes before returning to her bedroom. Id. at 132. Prekop heard the sound of glass breaking and went back into the kitchen to investigate, taking a stick for protection. Id. at 134. She turned on the light and saw a man, who she later described as clean-shaven with short, dark hair. Id. at 135, 172. She hit the intruder on the head with the stick. Id. at 135–36. The intruder grabbed Prekop and pushed her into the bedroom. Id. at 136. As the perpetrator pushed her from the lighted kitchen into the lighted bedroom, Prekop and her assailant were face-to-face. Id. at 136–37. Prekop had a clear opportunity to see her assailant. Id. at 137. In the bedroom, the assailant knocked the lamp lighting the room onto the floor and the bulb broke. Id. The perpetrator knocked Prekop to the floor, gave her a black eye, and raped her. Id. at 136–39. He kicked Prekop as he left her 5 bedroom. Id. at 141. After the attack, at around 4:00 a.m., Prekop telephoned the police. Id. at 137. Officers Robert Cheetham, Eugene Dean, and James Murphy were the first police officers to arrive at the scene, and they called an ambulance for Prekop. Id. at 143, 178. Prekop recalled speaking to two officers before the ambulance came. Id. at 143. Officer Thomas Kelly arrived approximately 45 minutes later, at around 4:30 or 5:00 in the morning. Id. at 187. At the hospital, Prekop received a sexual assault examination, and afterward police officer Marianne Popko drove Prekop to the police station and took her statement. Id. at 143–46. In her statement, Prekop provided a clear and detailed description of her attacker: I can describe the man who broke into my home as a white male about 20–30 years old. He had dark, wavy hair, and he was clean shaven. He was average height and weight. He was wearing a blue jacket, it looked like nylon and sneakers, which were white. He was wearing a dark color pair of pants that were neither jeans nor corduroy material. Ex. F, Prekop Statement to Police, Nov. 30, 1985; Tr. I, 172 (statement read to jury). II. PERROT’S ARREST, INTERROGATION, AND THE RECOVERY OF OTHER EVIDENCE On December 6, 1985, Perrot worked all day with his brother-in-law at a construction job. Tr. III, 508. That evening he went out with his girlfriend and two friends. Beginning at around 7:30 p.m., Perrot drank approximately 12 beers, used a gram of cocaine, and consumed eight valium pills. Id. at 508–09; MTS, 53–55. Later that night, he broke into the house of Joseph McNabb, running out when he realized McNabb was home, and he stole a purse from a woman outside a Denny’s restaurant. Tr. III, 510–11. Perrot went home and passed out around 2:30 a.m. MTS, 56. A few hours later, Perrot was awakened by his sister, Nancy Westcott, and several police officers shaking him. Id.; Tr. III, 538. Intoxicated and disoriented, he implored the young woman who he believed was his girlfriend to contact his sister, not realizing that he 6 was already talking to her. Id. Westcott reported that Perrot smelled like he had been drinking and was “out of it.” Id. As Perrot was being escorted to the police cruiser, he made a futile attempt to escape down the street. MTS, 56–58. The police beat and harassed Perrot upon apprehending him and in the car ride to the police station. Id. at 58–59. A. Perrot Is Interrogated For 12 Hours Having Barely Slept The Previous Night When he arrived at the police station, Perrot was taken to a cell for about 45 minutes. MTS, 62. Perrot was exhausted, but the officer on guard rattled his baton on the cell bars to prevent Perrot from sleeping. Id. Perrot, only 17 years old, had not slept more than an hour or two the previous night and was suffering the after-effects of heavy drug and alcohol abuse, but the police took him into an interrogation room and began accusing him of committing a series of rapes. Id. at 62–63. They interrogated Perrot on and off for the next 12 hours. See Tr. III, 436, 438, 456, 471–72, 474, 481, 484; see also Ex. G, Commonwealth v. Perrot, No. 85-5415, slip op. at 37 (Hampden Cnty. Super. Ct., Sept. 11, 2001) (“There was evidence that the Defendant had ingested drugs and alcohol shortly before his arrest; that he was either not permitted to sleep or unable to sleep during the many, many hours of questioning, and that he was interrogated intermittently during a period of at least 12 hours commencing as early as 5:05 a.m.” (citations omitted)). The police began by questioning Perrot about the McNabb break-in and the Denny’s purse-snatch. MTS, 64. Perrot recalls giving an oral statement admitting these crimes. Id. at 65. In typing up Perrot’s admission, however, the police wrote that the purse-snatch occurred before the McNabb break-in, when the events in fact occurred in the opposite order. Id. at 64–66. Perrot signed the statement despite its undisputed inaccuracy. Id. B. Perrot Purportedly Signs A Partial “Confession” To The Lichwala And Prekop Break-ins 7 Perrot’s arrest for the McNabb burglary and the Denny’s purse-snatch provided the Springfield Police with an obvious suspect for the string of burglaries and rapes of elderly women that had been unsolved for more than a year and a half. After Perrot’s arrest, the DA called a press conference “to ease the fears of neighborhood residents, who were frightened by 18 incidents of rape and assaults” in the community. See Ex. D, Judge cuts bail for teenager charged with rapes of elderly, The Morning Union, Jan. 1, 1986. At the conference, the DA told the press that he was “hoping we can tie [Perrot] to all of them, somehow” and stated that Perrot “was the prime suspect in several other rapes and sexual assaults.” Id. The police were also familiar with Perrot, having arrested him more than ten times previously for other offenses, none of which were ever sexual crimes. Commonwealth v. Perrot, 407 Mass. 539, 540 (1990). The police officers responsible for investigating the rapes arrived at the station soon after Perrot’s interrogation began, and they immediately sought a confession to the rapes. Officer Kelly, who was knowledgeable of the details of the Lichwala and Prekop break-ins, arrived at 4:30 a.m. to interrogate Perrot, despite having worked until midnight the night before. Tr. II, 388–89. In response to the police questioning, Perrot explained that, on November 30, he had been at a bar in Brattleboro, Vermont, and he had taken two purple mescaline pills. MTS, 97–98. Perrot denied any involvement in the assaults, and freely consented to provide blood and hair samples for comparison against the evidence found at the rape crime scenes. Tr. III, 478; MTS, 73–74. Nonetheless, Officer Kelly and later Officer Thomas Jarvis accused Perrot of breaking into the Lichwala and Prekop homes when he returned from Brattleboro. MTS, 99–100. After 36 hours without sleep and 12 hours into the interrogation, Perrot finally began to break down. He was “hysterical,” “emotional,” “a mess,” “very fatigued,” began weeping, and 8 periodically lapsed into incoherence. Perrot, 407 Mass. at 542; Tr. III, 544; see also id. at 541. Although Perrot has no recollection of doing so, after nearly 12 hours of questioning, police investigators claim that Perrot signed a partial “admission” stating that he was responsible for the Lichwala and Prekop break-ins.3 Tr. III, 484–85; Ex. H, Alleged George Perrot Statement to Police, Dec. 7, 1985. Perrot was then placed on suicide watch. The statement—the phrasing of which sounds more in police code than in the words of an exhausted 17-year-old—states that on the night of November 30, 1985, Perrot took two purple mescaline pills; returned from Brattleboro; was wearing his sister’s wool gloves; “entered an unlocked rear breezeway door” at Lichwala’s home by “kick[ing] the rear inner door”; took a handbag and “left through the rear door” and left the pocketbook a few houses down the street; and entered Prekop’s home through “the breezeway,” when a little black dog started barking, causing Perrot to run. Tr. III, 489–90; see also Ex. H, Alleged Perrot Statement to Police. The statement did not admit that Perrot encountered Prekop or raped her. Id. C. None Of The Victims Identifies Perrot During A Highly Suggestive Investigative Lineup Shortly after Perrot purportedly signed the partial “admission,” at approximately 4:30 p.m., and one week after the Lichwala and Prekop break-ins, the police asked Prekop, Lichwala, and Marchand to view a lineup of seven individuals at the police station.4 Thus, the police and prosecutor had also immediately linked the crimes to the same perpetrator. Tr. I, 147; Ex. C, Perrot Police Lineup Summ. Perrot stood out in the lineup as the only teen, the only man wearing dirty socks, with shaggy hair, a mustache and goatee, and the only one appearing 3 Curiously, no time is listed on the “confession” document to verify the police testimony that it was signed at 4:00 p.m., an omission notable for a document of its significance. See Ex. H, Alleged Perrot Statement to Police. 4 When shown the picture of the lineup, Prekop testified that she was not shown a total of seven men, but instead was shown a lineup of just four men. Tr. I, 157. 9 disheveled from lack of sleep and the previous night’s drug abuse. Ex. I, Photograph of George Perrot Police Lineup. Five of the individuals selected for the lineup are easily identifiable as police officers because of their physique, clean cut hairstyles, and groomed mustaches. See id.; see also Ex. C, Perrot Police Lineup Summ. (confirming that five police officers were in the lineup). The police included only one other prisoner in the lineup, who is also easily distinguishable because of his shoulder-length hair. Id. Despite the highly suggestive lineup, none of the victims identified Perrot as her attacker. Ex. C, Perrot Police Lineup Summ. 2. Prekop’s refusal to identify Perrot was based on objectively verifiable facts. The perpetrator was clean shaven. Perrot had a “rough beard,” thick mustache, and a full goatee. The perpetrator had wavy black hair, above the neckline, and he looked like someone “who went often . . . to the barber shop.” Tr. I, 155–56. Perrot had “real wild,” shaggy, long, brown hair, both at the time of his arrest and at the time of the incident. See Tr. III, 500; Ex. J, George Perrot Booking Photographs. These photographs are consistent with the uncontroverted testimony by both the prosecution’s and defense’s witnesses as to Perrot’s appearance at the time of his arrest. Tr. III, 540 (Westcott testifying that Perrot had a mustache and beard for weeks or months before his arrest); id. at 500 (Detective Jarvis testifying that Perrot “had a rough beard, mustache and [his] hair was real wild”). As this court previously concluded, “[b]y inference and by uncontroverted direct evidence, [Perrot] had this facial hair on November 30, 1985,” the night that a clean-shaven man raped Prekop. Ex. G, Perrot, No. 85-5415, slip op. at 3. Despite Prekop’s unequivocal statement that her attacker was clean shaven, all the men in the police lineup had facial hair. Id. Believing Prekop would select Perrot, despite the irreconcilable differences between him and the perpetrator, Officer Kelly testified that it never occurred to him to put an individual matching Prekop’s description in the lineup. Tr. II, 230. 10 Prekop’s refusal to identify Perrot is all the more significant because she was raped three times, the first of which took place while Perrot was in juvenile detention. See Ex. A, George Perrot Juvenile Record. Nor did Lichwala or Marchand identify Perrot as their attacker. Ex. C, Perrot Police Lineup Summ. 2. Lichwala, though she did not see the perpetrator on November 30, 1985, was raped one month earlier on October 24, 1985, and she believed the person who broke into her house on November 30, 1985 to be the same person. First Tr. II, 251. And Marchand, who was also previously raped, did not identify Perrot, corroborating the forensic evidence that excluded Perrot as the donor of a semen stain taken from the Marchand crime scene. MTD, 66. D. Other Evidence Gathered By The Police Was Flawed And Subject To Conflicting Testimony Regarding Its Origin And Relevance The refusal by each of the three victims to identify Perrot left the Commonwealth with a highly circumstantial case reliant on a partial “admission” extracted after relentless questioning of a 17-year-old, sleep-deprived, drug-dependent suspect, only after he had an emotional breakdown. To shore up that case, police attempted to use forensic evidence to link Perrot to the crime scene. The origins and relevance of each piece of forensic evidence, however, were hotly disputed. First, the FBI purportedly found blood on a bed sheet and a pair of gloves recovered from the crime scene. There was, however, no evidence that the perpetrator bled on the bed or anywhere else. Prekop disputed that any blood from the attack could be on the bed because the sexual assault occurred on the floor, not the bed. 5 Tr. I, 163 (“[H]ow could his blood be on my sheets?”). Rather, she believed the stains on the sheets were from her uncle, who lived with 5 Prekop said that the police tried to convince her that the blood and the hair on the bed were the rapist’s, which she did not believe was possible because the rapist was never on the bed. Ex. B, Edwards Aff. 11 Prekop through 1980 and often coughed up blood on the sheets due to a respiratory illness. Id. at 163–65. Prekop also testified that there was no fresh blood on the sheets following the attack, nor did she see blood anywhere else in the house—including on her nightgown, the floor where the assault occurred, or on the stick she used to hit the attacker. Id. at 169. Prekop’s account is consistent with the findings of police officers who initially investigated the crime scene—they did not report seeing any blood in the house, on the bedsheet, or on the floor. Id. at 182–83 (Officer Robert Cheetham testified that he did not observe any bloodstains on the stick when he collected it); id. at 194 (Officer Kelly testified that the bedsheet “was soiled, appeared to have some fibers on it . . . [and] was flowered,” but he did not mention any bloodstains). The Commonwealth also introduced a pair of women’s gloves, the police testimony about which was plainly contradictory. Officer Kelly had purportedly recovered the gloves from the crime scene and then brought them to Perrot’s sister, Nancy Westcott, to ask if they were hers. Tr. II, 406–07. Westcott told Officer Kelly that the gloves were not hers, but that they looked like a pair that belonged to Perrot’s girlfriend. Tr. III, 546. At trial, however, Officer Kelly testified that Westcott had actually told him that the gloves were hers. Tr. II, 406–07. But, the record demonstrates that after Officer Kelly spoke to Westcott, Detective Jarvis took the gloves to Perrot’s girlfriend’s home to see if she owned them. MTS, 156–57; see also Tr. III, 498–99. Detective Jarvis, however, concluded that “they didn’t belong to [Perrot’s girlfriend].” MTS, 156–57; see also Tr. III, 498–99. Detective Jarvis further testified that there was a “good possibility” that Perrot’s girlfriend showed him a pair of gloves that resembled the gloves found at the crime scene and that she still had in her possession. Id. at 497–99. Thus, Officer Kelly’s testimony that the gloves were Westcott’s was contradicted by Detective Jarvis’s testimony that corroborated Westcott’s version of events—that Westcott in fact told Kelly that they were not 12 hers and looked like Perrot’s girlfriend’s. If Perrot’s sister had actually claimed the crime-scene gloves as her own, as Officer Kelly claimed, there would have been no need for Detective Jarvis to later visit Perrot’s girlfriend to see if they belonged to her (which they did not). There was also conflicting testimony about how the gloves were discovered and who discovered them. According to Prekop, after giving a statement to Officer Popko at the police station, she and her sister returned home and found the gloves at the foot of the bed while making the bed. Tr. I, 163. Yet her statement in its final form included a statement about discovering the gloves: “When the police arrived I noticed a pair of gloves on the bed. They are not mine. I never saw them before.” Id. at 173. When presented with this statement at trial, Prekop immediately noticed the inconsistency; she acknowledged that the statement in general was hers, but stated: I didn’t give this, any statement about any gloves. . . . They got it here about some gloves. I don’t know when they put that in. I didn’t see the gloves until two or three hours until after I got home and I went to make up my bed and there was the gloves and the stick was still on the foot of the bed and the gloves too. Id. at 144. Officer Kelly testified at Perrot’s first trial that he found both gloves on the floor at the foot of Prekop’s bed, First Tr. III, 195–97, but during the second trial, Officer Kelly told the jury that he found one glove on the bed and one glove on the floor. Tr. II, 215. The change in Officer Kelly’s story allowed the prosecution to suggest that the bloodstain on the bed had come from the gloves. Also inexplicably, two evidence tags exist for the gloves.6 The inconsistencies 6 The first was purportedly submitted by Kelly on November 30, 1985. Ex. EE, S.P.D. Evid. Tag 49449. The second was submitted by Kelly and Sgt. Louden under circumstances that appear to indicate it was submitted at a later date, but was made to appear as though it was submitted on November 30th with the rest of the evidence: The second gloves evidence tag is dated November 30, 1985, but the date is inserted to the right of what appears to be a crossed out “12,” indicating the writer initially wrote “12” for December and then backdated the tag. Ex. FF, S.P.D. Evid. Tag 49492. Kelly was never asked and it was never explained why there were two evidence tags for the gloves. 13 about who recovered the gloves and how were not resolved at trial. Finally, after gathering the remaining physical evidence, the police collected a second sample of Perrot’s blood to conduct serological testing on December 10, 1985, after his arraignment.7 Police claimed that, during that collection, Perrot allegedly disclosed the location of Lichwala’s pocketbook. Perrot strenuously denies giving the statement. The pocketbook was allegedly found later that day, though the police testimony about the circumstances under which it was found strains credibility. Because the questioning of Perrot outside the presence of his counsel violated his due process rights, Perrot’s purported statement was suppressed at both trials. In addition, the Supreme Judicial Court (“SJC”) vacated his first conviction after finding that the introduction of the pocketbook itself was fruit of the poisonous tree and therefore improper. See Perrot, 407 Mass. at 544–45. Neither Perrot’s alleged statement nor the pocketbook were evidence at his second trial. III. PERROT’S TRIAL The only evidence introduced at trial that the prosecutor could claim unequivocally placed Perrot at the crime scene was the now discredited hair microscopy testimony. FBI Agent Wayne M. Oakes provided damning, scientifically invalid testimony that he had “match[ed]” a hair removed from the bedsheet with a known sample from Perrot, and he repeatedly claimed that the possibility that the hair might “match” someone else was so remote as to be irrelevant. Although FBI Agent William G. Eubanks performed a serological analysis that he claimed was also consistent with Oakes’s finding, it was significantly less conclusive because it only narrowed the pool of possible perpetrator’s to eight percent of the white population. The 7 The Commonwealth’s chain of custody for Perrot’s blood samples was also baffling. Inexplicably, ADA Bloom—who later went so far as to forge Perrot’s signature on a fabricated confession imputed to Perrot, see infra—personally delivered these samples to the FBI office in Washington D.C., a 10–12 hour drive from Springfield. Tr. II, 235. 14 Commonwealth’s remaining evidence—testimony regarding gloves allegedly belonging to Perrot’s sister—was filled with inconsistencies. A. Prekop Explains Why Perrot Was Not Her Attacker Although the prosecutor was convinced of Perrot’s guilt, the victim—the only person to have actually seen the perpetrator, and the person with the most interest in seeing this man prosecuted—was not. At trial, Prekop explained exactly why Perrot did not resemble the man who attacked her. Her attacker had dark, wavy hair, and was clean shaven, without a mustache or beard. Tr. I, 141, 154. His hair “wasn’t too long” and he “looked like somebody went—he went often, went to a barber shop. [His hair] wasn’t hanging way down over his neck.” Id. at 155–56. He was young, “between 20 and 30 at the most.” Id. at 146. Perrot’s appearance when he was arrested a week later did not accord with Prekop’s observations. Prekop testified that the men she was shown in the lineup did not have “that kind of color hair,” looked “like they needed a hair cut,” and appeared as though their hair had “never been combed.” Id. at 147. To her, these individuals “looked like some kind of a tramp.” Id. She noted that “some of these people they were showing me at the courthouse [sic] they all had longer hair and pretty stringy and pretty sick looking.” Id. at 156. Defense counsel presented Prekop with a picture of the lineup, identified Perrot, and asked Prekop whether he looked like her attacker. She responded: “How can I say it was when this man has a mustache and a beard? This fellow [who attacked me] didn’t have any beard. He didn’t have any mustache.” Id. at 158. She then pointed to another individual in the lineup (presumably a police officer) and said, “It could be this one, more this length.” Id. But of the individuals she was shown, “nobody had hair on that order. Everybody’s hair was hanging down their neck and looked like they needed a hair wash.” Id. at 159. She once again reiterated why Perrot was not her attacker: “The man who assaulted me didn’t have a mustache or a beard. . . . He was clean shaven[,] the man who 15 came into my house.” Id. at 161–62. Defense counsel asked Prekop whether Perrot, who was clean shaven and sitting in the courtroom, looked like her attacker. Id. at 160. She indicated that Perrot did not. Id. B. Wayne M. Oakes’s Microscopic Hair Analysis Faced with the victim’s insistence that Perrot was not the perpetrator, the prosecution relied heavily on forensic evidence that today would be inadmissible, since it has been rejected by the very agency that proffered the evidence in the first instance. The evidence was improperly used to link Perrot not just to the crime scene but to place him in the victim’s bed. Wayne M. Oakes, a Supervisory Special Agent at the FBI, testified for the prosecution as an expert in the area of forensic microscopic analysis of hairs and fibers, and claimed to be a “scientist” with expertise in “hairs and textile fibers”. Tr. II, 334–35. Oakes’s hair analysis was based entirely on a single brown hair that was removed from the bedsheet or brown paper packaging in which the bedsheet was shipped. Id. at 351–52. Oakes testified, and the prosecutor drove the point home, that in all likelihood this hair came from Perrot, and that any possibility that the hair came from another source was so insignificant and remote as to be meaningless. Oakes accomplished this by boasting—without empirical evidence—of near infallibility when comparing hairs. Instead, in what the FBI has identified as invalid testimony, see infra, he used his experience to argue that hair evidence was highly probative. Oakes testified that out of “literally hundreds of thousands of hair comparisons,” he had only failed to discriminate between two hairs in a handful of cases. Tr. II, 354. He further claimed that “it’s my experience in 10 years [that] it’s extremely rare I will have known hair samples from two different people I can’t tell apart.” Id. at 349. Oakes estimated that he examined hairs for between 3,000 and 5,000 cases over the course of this ten-year period, at a rate of between 300 to 500 cases per year. Id. 16 at 335. He later testified that “[l]iterally it would be very hard to figure out” how many hair examinations he had conducted. Id. at 354. He estimated that he had “conducted thousands of examinations in thousands of cases. Many of those cases involved literally hundreds of thousands of hair comparisons or examinations so it would be a very large number.” Id. Oakes telegraphed to the jury that even if his conclusion did not mean with 100 percent certainty that the hair found in the victim’s bed came from Perrot, the probability that it was someone else was so remote as to be irrelevant, and he falsely described the “match” as “[the] basis of strong association.” Id. at 350. Oakes drove this point home by using a highly misleading analogy, comparing hair analysis to a person being able to recognize the face of someone that she knows. Oakes told the jury that for someone of his expertise comparing two hairs is analogous to comparing two faces: [W]hen you look at a person’s face you look at their eyes, their ears and mouth. Most people have those characteristics, but when you walk into a room that’s crowded you can spot and recognize someone you know even though everyone has two eyes, two ears and a nose and mouth because you know that person. You can detect subtle differences. You can pick out—oh, yeah, that’s the person I’m pointing out. Id. at 341–42 (emphasis added); see also id. at 339 (“I’m looking at all of the microscopic characteristics that go to make that hair somewhat unique. It would be just like if you asked me to look at an individual’s face.”). Perrot’s defense counsel objected, but the court overruled the objection because the information was simply “background” information. Id. at 343. The prosecution then echoed this improper testimony, impressing it yet again on the jury: “Sir, you were indicating that an experienced hair examiner would look at each of these various parts of the hair that we might look at various parts of the face.” Id. Defense counsel failed to call an expert to correct Oakes’s false testimony, relying only on an unsuccessful attempt to cross-examine Oakes. For example, defense counsel attempted to 17 elicit the only scientifically valid conclusion concerning the putative probative value of hair evidence, asking if it was correct to say that Oakes’s results showed only that the hair was “sufficiently similar to one or some of the hairs that you received from Perrot so that you could not excuse Perrot as having been the source of that hair.” Id. at 371. But Oakes refused to agree to that proper characterization of the weight of the evidence. Id. He also denied that there could be any inconsistency between his conclusions and the conclusions of another examiner unless the other examiner “had a lesser amount of training” than he did. Id. at 374–75. Rather than acknowledging the inherent uncertainty in hair analysis, Oakes misled the jury into thinking that hair analysis was all but certain evidence of Perrot’s guilt. The jury was falsely told by an FBI-trained and certified “scientist” that hair examination was just shy of a unique basis of identification. Oakes provided no context for his analysis, never explaining, for example, how frequently given characteristics of hair appear in the population, because no such data exists. Deprived of the opportunity to meaningfully consider the risk of a false positive association—as evidenced by at least 74 wrongful convictions implicating hair microscopy—the jury was unable to ascribe the proper weight to Oakes’s testimony. As scientifically false and misleading as Oakes’s testimony was, it was made much worse by the prosecutor’s opening and closing statements. In his opening, the prosecutor claimed outright that the evidentiary hair was not simply similar to samples from Perrot, but that it actually belonged to Perrot. Tr. I, 107. The prosecution dramatized this unsupported assertion by falsely telling the jury that in order for Perrot to be innocent, the police would have had to have planted his hair on the scene of the crime. Tr. IV, 627 (arguing that in order for Perrot to be innocent there would have to be a “conspiracy you can’t believe” where the police “[p]ull out one of George’s hairs. . . . And the FBI finds that single hair. That’s preposterous. Its [sic] 18 preposterous.”). After constructing this straw man, the prosecution blew it down by arguing that in order for Perrot to be innocent, the FBI agents would have to “get together and risk their reputations and careers to create evidence to reach conclusions not supported.” Id. at 626. Today, we now know that Oakes did just that by assigning a high probability that the evidentiary hair came from Perrot. The prosecutor capitalized on the false testimony in his closing as well, emphasizing that Oakes had testified that “it was very rare, very rare to ever find two hairs from different people that he couldn’t tell apart,” to claim without foundation that there was a very strong likelihood that the hair belonged to Perrot. Tr. IV, 620. These statements were not only improper opening and closing arguments (as both this Court and the Court of Appeals already concluded), but they also flagrantly disregarded the limits of science by arguing that the evidentiary hair came from or was at least very likely to have come from Perrot. C. William G. Eubank’s Serological Protein Profiling The Commonwealth also presented serological evidence showing that bloodstains recovered contained enzymes and proteins consistent with Perrot’s blood and at least eight percent of the white population, which equated to more than 14,000 white males in Hampden County alone.8 Eager to test this evidence with a more discriminating forensic test to demonstrate his innocence, Perrot recently moved this Court for DNA testing on the evidence, but could only test a few very small bloodstains remaining on the sheet and gloves; this evidence failed to produce a quantifiable level of DNA. The Commonwealth called William G. Eubanks, also a Supervisory Special Agent at the FBI, at trial as an expert in the area of forensic serology. Tr. II, 263. Eubanks had performed 8 Department of Commerce Economic and Statistics Administration Bureau of the Census, 1990 Census of Population General Population Characteristics Massachusetts (April 22, 1992), http://www.census.gov/prod/cen1990/cp1/cp-1-23.pdf. Perrot’s counsel failed to research this or raise this statistic at trial. 19 serological protein profiling on the gloves and bedsheet. Id. at 194–95, 270–71. Eubanks found blood at the base of the fingers of the gloves. He concluded that the blood contained the PGM 1+ genetic marker, but that it could not be characterized for any of the remaining 12–14 possible testable enzymes or proteins. Id. at 287. He testified that the protein found on the gloves was found in 41 percent of the Caucasian population, and that neither Prekop nor Perrot could be excluded as the source of the blood. Id. at 299–300. And, as previously described, the circumstances in which the gloves were found and their ownership was the subject of conflicting prosecution testimony at trial. Although there was no evidence that either the perpetrator or Prekop bled or were on the bed during the incident, Eubanks tested two bloodstains on the bedsheet. Id. at 292. One bloodstain was identified as human blood, but could not be characterized further. Id. at 295. The other contained four genetic markers: PGM 1+, EAP BA, HP 2-1, and Tf C. Id. at 298. Tests for ABO blood type and eight to ten other testable genetic markers were inconclusive. Id. at 292. Eubanks testified that Perrot’s blood contained the same four genetic markers as the blood on the sheet. Id. at 299–300. Prekop was excluded as the source of the blood. Id. at 299. Eubanks further testified that eight percent of the white population had blood matching the stain on the sheet, a figure he calculated by multiplying the population frequency of each of the four blood markers as reported in national studies.9 Id. at 301. Eubanks did not detect blood or semen on the pillowcase. Id. at 289. As with the hair testimony, the prosecutors amplified and exaggerated the serological results in opening and closing statements. Tr. I, 107 (prosecution asserting in 9 Eubanks explained that this method of multiplying the four different frequencies to reach a final percentage—known as the “product rule”—is based on the law of independent events and the genetic theory that “each of these factors [is] inherited independent[ly] of one another.” Tr. II, 249–50. Notably, at Perrot’s first trial the court prohibited Eubanks from testifying about the product rule and his “final [statistical] calculation” because, in the court’s view, “the use of statistics [in this manner] would be prejudicial.” First Trial Tr. IV, 3–4. 20 opening statements that the blood was left behind by George Perrot); Tr. IV, at 627 (prosecution arguing that in order for Perrot to be innocent, police officers would have had to have placed blood on the evidence). Defense counsel failed to retain an expert to challenge Eubank’s testimony. Instead, he relied upon testimony from the victim to try to prove that the bloodstains did not come from the attacker, but rather from Prekop’s uncle. Tr. I, 163–65 (noting blood on sheets likely came from her uncle, who suffered from a respiratory illness). Nor did Prekop see any blood elsewhere in the house; there was no blood on the nightgown, on the floor, on the stick she used to defend herself, or anywhere else in the house. Id. at 169. Likewise, the police officers who investigated the crime scene did not see any blood in the house and did not report finding any blood on the bedsheet when they collected it. See id. at 182–83. But this collateral attack on the origins of the blood evidence did little to counter the prosecution’s amplification and exaggeration of Eubank’s serological results in opening and closing statements. Id. at 107; Tr. IV, at 627. PROCEDURAL HISTORY Because of gross prosecutorial misconduct, Fourth and Sixth Amendment violations, highly questionable circumstantial evidence, and now the discovery of the use of false “scientific” evidence, Perrot’s case has a long and tortuous procedural history involving two trials, review by Massachusetts SJC, and even post-conviction relief granted by this Court. Perrot was first tried and convicted in 1987, but his conviction was reversed by the SJC. While preparing for a second trial, the DA’s office uncovered serious misconduct by a former prosecutor who had forged an affidavit attributed to Perrot. But the Superior Court nevertheless allowed the prosecution to proceed with a second trial. Perrot was convicted again in 1992, but was granted a new trial by the Superior Court in 2001, which found that there was a substantial risk of a miscarriage of justice. Perrot was released pending the new trial, but the new trial order 21 was reversed and vacated by the Court of Appeals and Perrot was incarcerated. Both this Court and the SJC have recognized serious deficiencies in the case against Perrot, demonstrating how difficult it has been for the prosecutor to obtain a conviction on such flawed, circumstantial evidence. This is only underlined by the fact that the other rapes, which were extremely similar and of which the police initially suspected Perrot, were never charged or solved. I PERROT’S FIRST TRIAL D. Perrot’s First Conviction Is Reversed By The Massachusetts Supreme Judicial Court In December 1987, Perrot was convicted of unarmed robbery, indecent assault and battery, aggravated rape, burglary and assault on an occupant, and breaking and entering with intent to commit a felony for the crimes of November 30, 1985. He was sentenced to life imprisonment. On appeal, the SJC vacated Perrot’s convictions and remanded for a new trial. Perrot, 407 Mass. 539. The SJC held that the Superior Court erred in admitting an illegally obtained handbag belonging to Lichwala that the prosecution claimed the police found after Perrot described its location. Id. In so holding, the SJC emphasized that there “were no identifications of the defendant by the victims” and the “fact and the validity” of Perrot’s statements to the police were “strenuously denied by the defendant [during] trial.” Id. at 549. The Court also noted that Perrot has “denied throughout that he was involved in any sexual attacks on women.” Id. E. The District Attorney Fabricates And Forges A Statement Falsely Attributed To Perrot, Attempting To Tie Perrot To The Serial Rapes Apparently Committed By A Single Perpetrator Following Perrot’s successful appeal to the SJC, a newly assigned ADA began preparing for Perrot’s second trial. During the course of her preparations, she discovered that the former ADA 22 assigned to the case, Francis W. Bloom, had fabricated and forged a statement attributed to Perrot. The statement was dated December 16, 1987, a few days after Perrot’s first conviction. See Commonwealth v. Perrot, 38 Mass. App. Ct. 478, 479 (1995); Ex. K, Fabricated and Forged Statement Attributed to George Perrot, Dec. 16, 1987. The forged statement indicates that Perrot raped Prekop and Lichwala on the previous occasions, and that he raped and stole $1,000 from a third elderly woman, presumably Marchand, with the assistance of two other individuals. Perrot’s signature appeared on the bottom of the affidavit because either ADA Bloom or a police officer traced it from another document to make it appear as though Perrot had signed the forged document. MTD, 39. Bloom used the forgery during interrogations of the two other individuals in an effort to extract confessions from them. Id. at 59–62. Neither admitted to the crimes. Bloom left the fabricated and forged confession in Perrot’s file before leaving the DA’s Office for private practice. The new ADA found the document, which she “naturally took to be legitimate,” and predicted that retrying the case “should be relatively easy,” given that she believed the prosecution could use the forged confession at trial. Perrot, 38 Mass. App. Ct. at 480. Springfield Police Officer Kelly then identified the confession as a forgery to the ADA. The Superior Court held an evidentiary hearing on the forged affidavit, at which Officer Kelly denied any involvement and testified that ADA Bloom had admitted to him that he had created the forged document. MTD, 36–38, 41. ADA Bloom claimed that he could not remember the details of who typed and signed the forgery, but he did acknowledge that he had “approved” the fraudulent document. Id. at 49; see also id. at 55. The Court ultimately found that ADA Bloom was responsible. Perrot, 38 Mass. App. Ct. at 479–80. For his misconduct, ADA Bloom 23 received a public censor. See Ex. L, Matter of Bloom, 9 Mass. Att’y Discipline Rep. 23 (Dec. 16 1993). The “confession” was discovered to be a forgery before Perrot’s second trial. Perrot moved to dismiss the indictments, arguing that the fraudulent statement made it more likely that his purported “admission” regarding the events of November 30, 1985 had also been forged. The Superior Court denied Perrot’s motion and granted the Commonwealth’s Motion in Limine for suppression of evidence of the forged statement during the second trial, finding that the prosecutorial misconduct was irrelevant. Tr. II, 386–87. IV. PERROT’S SECOND TRIAL AND POST-CONVICTION LITIGATION A. Perrot Is Convicted Again, But The Superior Court Orders Another New Trial On January 9, 1992, Perrot was again convicted of the same five offenses and received another life sentence. On April 25, 2000, Perrot filed a Motion for a New Trial, arguing that counsel provided ineffective assistance by failing to object to and appeal the prosecution’s improper closing arguments. On September 11, 2001, Superior Court Justice Wernick—who had not presided over the trial—issued an order granting Perrot a new trial. Ex. G, Perrot, No. 855415, slip op. The Superior Court held that during closing remarks the prosecutor misstated the presumption of innocence, “with the result that the jury was potentially misled into believing that the presumption disappeared once the presentation of evidence began.” Id. at 20–23. This error was exacerbated by the judge’s instructions, which similarly misstated the law. Id. In addition, the prosecutor improperly implied that the jurors could acquit only if they believed that the government’s witnesses were lying and vouched for the credibility of government’s witnesses, including FBI Agents Wayne Oakes and William Eubanks, who provided critical (false) forensic testimony. Id. at 24–33. The prosecutor also referred to evidence not in the record and excluded 24 evidence, and enlisted the jurors on the side of the prosecution by making appeals that placed the prosecutor and the jury “on the same footing.” Id. The Superior Court concluded that this misconduct—which was plainly part of a pattern of misconduct surrounding the effort to convict Perrot of this crime—created a substantial risk of a miscarriage of justice given the highly circumstantial case against Perrot, leaving the Court with “serious doubt that the Defendant’s guilt was fairly adjudicated.” Id. at 39. B. The Court Of Appeals Reverses And Vacates The Order For A New Trial Perrot was released on bond in late December of 2002—after 17 years in prison—and began to put his life back together. He started dating a girlfriend and started a business that occupied most of his time as he tried to pay back the money he borrowed from friends and family to make bail. On May 12, 2003, however, the Court of Appeals, relying on the now discredited hair comparison evidence, reversed Justice Wernick’s Order in an unpublished decision. Commonwealth v. Perrot, 58 Mass. App. Ct. 1102 (2003). Although finding the motion court’s concerns well-founded, critical to the Court of Appeals’ ruling reinstating the conviction was that the cumulative effect of the errors was not sufficiently significant because there was corroborating hair evidence. Id. at *5, 8. The SJC denied Further Appellate Review. Commonwealth v. Perrot, 441 Mass. 1104 (2004). Perrot’s two subsequent pro se motions for a 25 new trial were denied. None of Perrot’s motions for a new trial involved the false scientific testimony addressed in this motion.10 C. The Superior Court Grants DNA Testing On May 17, 2012, Perrot through undersigned counsel filed a Motion for Post-Trial Discovery and DNA Testing under the newly enacted post-conviction DNA testing law. Mass. Gen. Law Ch. 278A. On September 21, 2012, the Court granted DNA testing on the forensic items in possession of the Superior Court Clerk of Court (a bedsheet and a pair of women’s gloves), but denied Perrot discovery to assist in locating the other forensic evidence not in possession of the Clerk’s office. This other evidence includes samples taken from the sheet and gloves by the FBI for forensic testing, the hair that was the subject of Oakes’s expert testimony, and samples from a sexual assault forensic kit taken from the victim after the November 30th rape. On April 24, 2014, Bode Technology (the agreed-upon laboratory to conduct the testing) notified the parties and Court that the DNA testing had failed to produce measurable levels of DNA from several remaining bloodstains found on the sheet. Because of the low likelihood of obtaining DNA results in these circumstances, Perrot is not moving forward with DNA testing on these items at this time. ARGUMENT I THE FBI’S ADMISSION DEMONSTRATES THAT PERROT WAS CONVICTED USING FALSE AND MISLEADING FORENSIC TESTIMONY THAT HE WAS 10 On November 29, 2006, Perrot filed pro se a second motion for a new trial, arguing that he had been denied effective assistance of his trial and appellate counsel for a variety of reasons, including counsel’s failure to test the root of the questioned hair found at the crime scene, failure to call an expert witness to oppose the prosecution’s hair expert, failure to object to certain hearsay evidence and the prosecution’s closing statement, failure to request a consciousness of guilt instruction, and failure to investigate certain facts and corroborating witnesses. This Court denied the motion on December 5, 2006. On June 6, 2008, Perrot filed pro se a third motion for a new trial, asserting that he had been denied due process because the prosecution had knowingly presented false testimony of Officer Kelly, and claiming that his counsel had been constitutionally ineffective. On August 21, 2008, this Court denied the motion. 26 THE SOURCE OF THE CRIME-SCENE HAIR A The FBI Has Admitted That Its Hair Microscopy Testimony Exceeded The Limits Of Science, Reversing Its Decades-Long Practice Of Testifying That Criminal Defendants Were The Source Of Evidentiary Hairs In 2013, the FBI—the very agency that proffered the “scientific” testimony in Perrot’s trial—publicly conceded that its microscopic hair analysts’ testimony was scientifically invalid. And since the time of Perrot’s trial, when not a single defendant had been exonerated on the basis of invalid hair testimony, there have been 74 exonerations involving hair evidence. This newly discovered evidence establishes that the FBI’s hair testimony in Perrot’s trial would be inadmissible today and the prosecutor would not have been able to argue in his opening and closing that Perrot’s hair was found at the crime scene. For decades, FBI analysts linked criminal suspects to crime scenes through hair microscopy. Simply put, hair microscopy is a fairly crude forensic technique,11 first developed in the 1860s for comparing the microscopic characteristics of a hair of an unknown origin with hairs from known sources and attempting to make an association. Despite the scientific jargon used by the FBI analysts when testifying, hair microscopy involves little more than mounting the hairs on slides and examining them under a microscope, comparing various features of the hairs. The examiner may find these features to be similar or dissimilar, and there is no scientific consensus for how many features must be similar before a match is declared, or how rare or common those features are in the general population. Nevertheless, for decades, when FBI examiners concluded that an evidentiary hair was microscopically similar to a suspect, they testified that the hair came from or likely came from the suspect. In the most extreme cases, the 11 With the introduction of mitochondrial DNA testing in the mid to late nineties, hair microscopy became obsolete. Prior to the development of that technology, however, the FBI performed hair microscopy on behalf of federal and state investigators in thousands of cases and trained state forensic units in the technique. 27 examiner testified that an evidentiary hair can only have come from the defendant to the exclusion of all others. But examiners oftentimes also made probabilistic claims about the likelihood that the hair came from a particular source, offering testimony that suggested or implied that the hair was likely to have come from a particular defendant, or testified that, in their personal experience, they have only rarely seen undifferentiated hairs in thousands of examinations. Prior to Perrot’s trial, while there was scientific debate about the value of hair microscopy, researchers did believe it could play a role in identification of suspects in the right circumstances. Although the academic research agreed that “hair is not generally a basis for positive personal identification,” researchers suggested that hair microscopy can “lead to a more positive conclusion” than simply stating that the hairs shared microscopic characteristics. Ex. M, B.D. Gaudette, Some further thoughts on probabilities and human hair comparisons, 23 J. of Forensic Sciences 758, 761 (1978). In other words, the researchers at that time believed that although hair microscopy cannot uniquely identify the source of the hair, under the right circumstances there could be grounds to believe that a suspect was likely the source of a hair. In 2002, when the cracks in hair microscopy started to appear, the FBI still held fast to the claim that it was reliable. At that time, a study of nine hair samples from nine individuals found that the examiners had a false positive rate of 12 percent (hairs of different sources were deemed microscopically similar), and a false negative identification rate of 44 percent (hairs of the same source deemed microscopically distinguishable). Ex. N, M.S. Verma et al., Hair-MAP: A prototype automated system for forensic hair comparison and analysis, 129 Forensic Sci. Int’l 168 (2002). Another study, authorized by an FBI agent after the advent of mitochondrial DNA testing, studied 80 hair microscopy matches identified by the FBI and found that nine of them 28 were excluded by mitochondrial DNA testing—a false positive rate of 11 percent. Ex. O, M.M. Houck & B. Budowle, Correlation of Microscopic and Mitochondrial DNA Hair Comparisons, 47 J. Forensic Sci. 1, 3 (2002). This error rate is even more significant when considering that these samples were not random, but were for suspects or victims whose hair had been submitted to the FBI for comparison. Despite these developments, even as recently as 2009, the FBI published a report stating that, although hair microscopy is not a means of “positive identification,” it could provide “substantial information because of the variation in hair among individuals.” Ex. P, Cary T. Oien, Forensic Hair Comparison: Background Information for Interpretation, 11 Forensic Sci. Comm’ns (2009). The FBI insisted that the significance of shared characteristics may be impressed on a jury “qualitatively or semiqualitatively” and that the only “limitation on the science is that there is always the possibility of a coincidental match.” Id. In 2009, however, NAS issued its seminal forensic report calling attention to the scientific invalidity of attempts to individualize results to particular sources.12 See Ex. Q, Excerpts from Nat’l Academy of Sciences, Nat’l Research Council, Strengthening Forensic Science in the United States A Path Forward (2009) (hereinafter “NAS Report”) at 155–61. NAS concluded that, at best, a “match” between two hair samples “means only that the hair could have come from any person whose hair exhibited—within some levels of measurement uncertainties —the same microscopic characteristics,” a statement that Oakes specifically rejected in Perrot’s trial. See id. at 156 (emphasis added). The NAS Report, like others, found that hair microscopy 12 Congress authorized and funded “the National Academy of Sciences to conduct a study on forensic science” and directed that the committee tasked with assessing and identifying the needs of the forensic sciences in this country to “include members of the forensics community representing operational crime laboratories, medical examiners, and coroners; legal experts; and other scientists as determined appropriate.” Ex. Q, NAS Report at 1. 29 cannot uniquely identify one person as the source of a hair. Id. But NAS went even further, making clear that even probabilistic estimates of the likelihood that an evidentiary hair came from a particular source are impermissible. This is true for two reasons. First, there is no consensus “on the number of features on which hairs must agree before an examiner may declare a ‘match.’” Id. at 160. Thus, while some examiners may look to seven characteristics and declare a match, others may look to 15. Second, there are no statistics on the distribution of particular hair characteristics in the population. Id. For example, hair pigmentation may be characterized as streaked, clumpy, or patchy, but there are no known samples for what percentage of the population has each of these characteristics. Because of these limitations, NAS concluded “that testimony linking microscopic hair analysis with particular defendants is highly unreliable,” and that evidence of a match “must be confirmed using mtDNA analysis.” Id. at 161. By contrast, previous research had insisted that hair microscopy had independent value irrespective of mtDNA results. Ex. O, M.M. Houck & B. Budowle, Correlation of Microscopic and Mitochondrial DNA Hair Comparisons, 47 J. Forensic Sci. 1 (2002). Following the NAS Report, a series of news articles in the Washington Post revealed that erroneous testimony by FBI hair examiners was “widespread and could affect potentially thousands of cases in federal, state and local courts.” Ex. R, Spenser S. Hsu, Convicted defendants left uninformed of forensic flaws found by Justice Dept., Wash. Post, April 16, 2012. The Post reported that, despite claims by FBI analysts such as Oakes that it was virtually impossible not to be able to distinguish between two human hairs from different sources, in one instance mitochondrial DNA testing revealed that two “FBI-trained analysts . . . could not even distinguish human hairs from canine hairs.” Id. “As a result, hundreds of defendants nationwide remain in prison or on parole for crimes that might merit exoneration, a retrial or a retesting of 30 evidence using DNA because FBI hair and fiber experts may have misidentified them as suspects.” Id. Based on this changing understanding, three men in the Washington, D.C. area— Donald Gates (2009), Santae Tribble (2012), Kirk Odom (2012)—were exonerated after mitochondrial DNA contradicted hair microscopy testimony proffered by FBI examiners. Following their exoneration, John Huffington was granted a new trial in Maryland state court after mitochondrial DNA demonstrated that the FBI had erroneously testified to a hair match in his trial. To date, more than 70 exonerations have involved problems with hair microscopy. See Ex. S, Yamil Berard, Forensic Science Commission to Review Convictions Based on Hair Samples, Star-Telegram, Aug. 11, 2013. The NAS Report, the exonerations of Gates, Tribble, and Odom, and intense media scrutiny eventually persuaded the FBI to reexamine the thousands of cases between 1985 and 2000 where, as in Perrot’s case, its agents testified to a positive association between a defendant’s hair and a questioned hair collected from a crime scene.13 See Ex. T, Norman L. Reimer, The Hair Microscopy Review Project, The Champion, July 2013, at 16. In doing so, the FBI has finally acknowledged for the first time that hair microscopy is limited “in that the size of the pool of people who could be included as a possible source of a specific hair is unknown.” Ex. V, FBI, Microscopic Hair Comparison Analysis Agreement, Nov. 9, 2012. Therefore, an examiner may not apply “probabilities to a particular inclusion of someone as a source of a hair of unknown origin.” Id. Instead, the strongest conclusion one can draw is that the suspect could be included in a class of people of unknown size that could have been a possible source of the evidentiary hair. Id. 13 The review was announced in July 2013, and is being administered in conjunction with the Innocence Project and the National Association of Criminal Defense Lawyers. See Ex. U, Innocence Project, Innocence Project and NACDL Announce Historic Partnership with the FBI and Department of Justice on Microscopic Hair Analysis Cases (July 18, 2013). 31 The FBI has identified three particular errors that exceed the limits of science: Error Type 1: “The examiner stated or implied that the evidentiary hair could be associated with a specific individual to the exclusion of all others.” Error Type 2: “The examiner assigned to the positive association a statistical weight or probability or provided a likelihood that the questioned hair originated from a particular source, or an opinion as to the likelihood or rareness of the positive association that could lead the jury to believe that valid statistical weight can be assigned to a microscopic hair association.” Error Type 3: “The examiner cites the number of cases or hair analyses worked in the lab and the number of samples from different individuals that could not be distinguished from one another as a predictive value to bolster the conclusion that a hair belongs to a specific individual.” Id. In the FBI’s ongoing review, DOJ has said that it will waive any statute of limitations or procedural bars under 28 U.S.C. § 2255, in federal post-conviction cases where error is found, due to the manifest unfairness of punishing a criminal defendant for the FBI’s protracted campaign to advance the prosecution’s case without regard to the limitations of hair microscopy. D. The FBI’s Admission Demonstrates That Oakes’s Testimony And The Prosecutor’s Arguments Linking Perrot To The Evidentiary Hair Exceeded The Limits Of Science The parameters of the FBI review reveal the falsity of the forensic testimony presented against Perrot, casting grave doubt on the justice of Perrot’s conviction. Each of the three errors for which the FBI is reviewing is manifestly present in Oakes’s testimony, infecting the jury’s ability to rationally consider the evidence and leaving them with the impression of nearirrefutable evidence of Perrot’s guilt. 1. Type I Errors The human face is widely understood to be unique. Every individual’s is different. This popular perception was exploited by Oakes to suggest that hair microscopy was capable of individualization, that a trained hair examiner looking through a microscope at a questioned hair 32 would recognize it as having come from a suspect the same way we recognize the unique features of the human face. This type of testimony has been labeled by the FBI as Type I errors, defined as a conclusion that occurs when the examiner “stated or implied that the evidentiary hair could be associated with a specific individual to the exclusion of all others.” Ex. V, FBI, Microscopic Hair Comparison Analysis Agreement, Nov. 9, 2012. In Perrot’s trial, Oakes twice testified that comparing hairs was like comparing two faces: [W]hen you look at a person’s face you look at their eyes, their ears and mouth. Most people have those characteristics, but when you walk into a room that’s crowded you can spot and recognize someone you know even though everyone has two eyes, two ears and a nose and mouth because you know that person. You can detect subtle differences. You can pick out—oh, yeah, that’s the person that I’m pointing out. Tr. II, 341–42; see also id. at 339 (“I’m looking at all of the microscopic characteristics that go to make that hair somewhat unique. It would be just like if you asked me to look at an individual’s face. What would I look at? I would look at his eyes, his ears, his nose, the mouth, the hair, the che[e]k structure, the bone structure, etc.”). Perrot’s objections on this point proved futile. Id. at 342–43. The prosecution then reinforced the grossly misleading face analogy a third time. Id. at 343 (“[Y]ou were indicating that an experienced hair examiner would look at each of these various parts of the hair [the way] that we might look at various parts of the face.”). Oakes’s comparison to recognizing a friend in a crowd suggests a virtually unique identification. Laypersons understandably believe that, when they see someone they know in a crowded room, they can identify that person to the exclusion of all others. This common understanding is born out in cognitive research, which demonstrates that humans have remarkable and unique abilities to distinguish between faces. See Ex. W, Elizabeth Norton, Identifying the Brain’s Own Facial Recognition System, Science, Oct. 23, 2012. Oakes exploited this common-sense understanding when explaining his methods. He putatively offered this 33 testimony to explain that, when examining a hair, he looks at the hair’s various features—a relatively straightforward concept that is not assisted by the analogy. In fact, the analogy likely undermines the very point Oakes was trying to make because when looking at faces, most people do not go through a checklist of individual features—Do the eyes match? Does the nose match? The ears? The facial structure?—to determine whether they recognize a face. Instead, facial recognition is an immediate and automatic reaction to the face as a whole. Nevertheless, Oakes, who referred to himself as a “scientist with a microscope,” offered “expert” testimony on the point to convey to the jury that he could provide a virtually unique association between a defendant and an evidentiary hair. Oakes levered this false testimony with other statements that implied that Perrot was the unique source of the hair. He testified that he would conclude that a hair matched when it was “microscopically indistinguishable” from the known samples. Tr. II, 348–49. In this case he explained that the evidentiary hair “exhibits all the same microscopic hair [characteristics] arranged in the same way as the characteristics present in the known hair from Perrot.” Id. at 353. In no less than 17 places in the record, Oakes testified that Perrot’s hair “matched” the evidentiary hair, see id. at 349, 350, 361, 364, 376, 378, 379, 381, 384, and he reinforced the inference through purposefully abstruse testimony detailing the many technical sounding and jargon-heavy variables that would have to be consistent for a hair to match precisely, see, e.g., id. at 347, 362. And Oakes insisted that this supposed match was evidence of whether Perrot was at the crime scene. Id. at 336 (“[I]f we have a particular crime scene it might be probative to determine whose hairs are at the crime scene to establish associations”). The NAS Report has criticized the use of this type of vague terminology because it could “be misunderstood to imply individualization.” See Ex. Q, NAS Report at 161. Likewise, Professor Karen Kafadar of 34 Indiana University and a member of the committee that produced the NAS Report has explained “that affirmative statements by an FBI hair examiner that imply the ‘uniqueness’ of the association . . . could well be misinterpreted to imply a very small probability that the two hairs did not come from the same individual—an interpretation that would be unjustified given the lack of relevant data needed to assure such probabilities.” See Ex. X, Kafadar Aff. Professor Kafadar explained that statements implying individualization include testimony that “that the questioned hair is microscopically indistinguishable from the known hair or from the known individual; or that the questioned hair is consistent with coming from the known individual.” Id. The prosecution committed even more egregious Type I errors than Oakes, telling the jury outright that the evidentiary hair had been deposited by Perrot at the scene of the crime. See Tr. I, 107 (in opening statement, prosecutor claiming that “when George Perrot was there . . . he left something behind. He left some of his hair.”). The prosecution echoed this false argument in his closing statement, falsely claiming that the only way that Perrot could be innocent is if the police framed him by planting his hair at the scene of the crime. Tr. IV, 627 (arguing that in order for Perrot to be innocent there would have to be a “conspiracy you can’t believe” where the police “Pull out one of George’s hair. . . . [a]nd the FBI finds that single hair. That’s preposterous. Its [sic] preposterous.” (emphasis added)). This line of argument could only be true if the prosecutor had a valid basis to conclude that the hair had come from Perrot to the exclusion of all others. None of this testimony and none of the prosecutor’s summation remarks would be possible today in light of the newly discovered evidence. 2. Type II Errors Oakes further testified that the “match” was “the basis of strong association” and categorically denied what the FBI now concedes is true: that, at most, his analysis showed that Perrot could not be excused as a possible source of the hair. Tr. II, 350. These statements are 35 quintessential Type II errors, which occur when the examiner “assigned to the positive association a statistical weight or probability or provided a likelihood that the questioned hair originated from a particular source, or an opinion as to the likelihood or rareness of the positive association that could lead the jury to believe that valid statistical weight can be assigned to a microscopic hair association.” Ex. V, FBI, Microscopic Hair Comparison Analysis Agreement, Nov. 9, 2012. There is no scientific foundation for Oakes’s testimony that his findings were “the basis of a strong association,” because applying “probabilities to a particular inclusion of someone as a source of a hair of unknown origin cannot be scientifically supported.” See id. The reason for the “strong association” was because he claimed that microscopic characteristics can make “hair somewhat unique” and that “one person’s head hairs tend to differ greatly from [an] other person’s.” Tr. II, 338–39, see also id. at 340 (“internal microscopic characteristics . . . can vary greatly from one person to another”); id. at 346 (“these characteristics can vary greatly from one person to another”); id. at 346–47 (“Pigment distribution and size granules can vary greatly” and cortical fusi “can vary greatly from one person to another”); id. at 347 (Ovoid body characteristics are consistent within an individual, but “can vary greatly in one person to another”). Oakes exacerbated this Type II error on cross-examination. At trial, defense counsel haplessly attempted to elicit accurate testimony by asking Oakes if it was true that all that could be determined from his analysis was that the evidentiary hair was “sufficiently similar to one or some of the hairs you received from Perrot so that you could not excuse Perrot as having been the source of that hair.” Tr. II, 371. Yet Oakes flatly rejected this correct categorization of the evidence and suggested that it was defense counsel who was misleading the jury: “That’s your characterization of my results, that’s not how I state my results.” Id. Defense counsel then 36 asked if his statement was accurate, and Oakes testified falsely that it was not. Id. The most that defense counsel could get Oakes to concede was that “a hair is not like a fingerprint,” a disclaimer he took pains throughout his testimony to thoroughly undermine. Id. But as the FBI has now acknowledged, the most that an examiner can testify to is that “a contributor of a known sample could be included in a pool of people of unknown size, as a possible source of the hair evidence (without in any way giving probabilities, an opinion as to the likelihood or rareness of the positive association, or the size of the class).” Ex. V, FBI, Microscopic Hair Comparison Analysis Agreement, Nov. 9, 2012; see also Ex. Q, NAS Report at 156 (a “match” between two hair samples “means only that the hair could have come from any person whose hair exhibited— within some levels of measurement uncertainties—the same microscopic characteristics” (emphasis added)). 3. Type III Errors Finally, Oakes compounded these errors by committing Type III errors, which occur when the examiner “cites the number of cases or hair analyses worked in the lab and the number of samples from different individuals that could not be distinguished from one another as a predictive value to bolster the conclusion that a hair belongs to a specific individual.” Ex. V, FBI, Microscopic Hair Comparison Analysis Agreement, Nov. 9, 2012. In discussing his background, which he capitalized on throughout his testimony to mislead the jury to believe that it was virtually impossible for a “scientist” like him to make a mistake, Oakes claimed that “on average I work anywhere from three to four hundred, possibly five hundred times a year,” and that he had thus conducted “[s]omewhere in the vicinity of three thousand to five thousands cases” in his ten years at the FBI. Tr. II, 335. Later in his testimony Oakes again tried to quantify his experience, stating at one point that he had “conducted thousands of examinations in thousands of cases” that “involved literally hundreds of thousands of hair comparisons or 37 examinations,” id. at 354, and stating at another that he had “done thousands and thousands of cases,” id. at 376. In all of his purported experience, Oakes claimed that “it’s extremely rare I will have known hair samples from two different people that I can’t tell apart,” and that it has been only “an extremely rare number of occasions” where he could not distinguish between known samples. Id. at 349. Because Oakes testified repeatedly that the evidentiary hair “matched” Perrot and that it was the basis “for strong association,” that it was like recognizing a human face, his implications that he was almost certain of his findings improperly linked Perrot to the crime scene. Further highlighting Oakes’s Type III errors were his assertions on cross examination that no other qualified hair examiners would reach a different conclusion. Specifically, when defense counsel asked whether another examiner might disagree with him, Oakes stated that “[i]f one [expert] had a lesser amount of training it would certainly be a possibility, sure, because it depends on how much experience you have and someone with less experience [than] myself might have a different opinion.” Id. at 374–75. Just minutes later, Oakes reiterated the point, stating that “if you have qualified experienced competent hair examiners that do this day in and day out and for years on end they are not going to differ, in my opinion, on the conclusion that I reached.” Id. at 376. Oakes’s testimony that he had conducted countless hair examinations, that he was virtually never wrong in his analysis, and that no other qualified expert could possibly disagree with him is exactly the type of error that the FBI has conceded is false as a matter of science. 4. Other Errors In Oakes’s Testimony Oakes repeatedly pushed back on any effort to elicit information about the basis of his conclusions. Instead, he offered a conclusory assertion that the crime scene hair had to have been Perrot’s because it “exhibit[ed] all the same microscopic hair [characteristics] arranged in 38 the same way the characteristics present in the known hair from Mr. Perrot” and concluded, once again, that the “question hair was consistent with coming from Mr. Perrot.” Tr. II, 353. Oakes could not recall the length or diameter of the crime-scene hair, but claimed—years later, with no notes—that whatever the length of the hair was, it was consistent with Perrot’s. Regarding length, Oakes could only say that the hair was not “like a quarter of an inch fragment” to argue that it was sufficient for comparison purposes, and therefore that his conclusion could not be challenged. Id. at 358. The failure to present evidence of the length of the hair is particularly significant because Prekop had testified that her attacker had short hair, whereas Perrot had long, unkempt hair. Thus, a short hair with a cut tip and root would provide further evidence that Perrot was not the attacker. But Oakes could not recall “whether it was one half inch, [or] three and a half inches.” Id. at 358–59. He also could not recall the length of hair samples from Perrot. Id. at 360. Yet Oakes insisted that the hair was “consistent in length with the known hairs from Perrot” even though he could not recall their length. Id. On recross, Oakes offered only a few specifics about the hair in question, but he could not provide thorough testimony about which features he believed were consistent. Id. at 385. Perrot’s attorney failed to object despite the fact that Oakes’s inability to discuss the underlying specifics of his hair examination meant there was no meaningful opportunity to challenge it. Cf. United States v. Willock, 696 F. Supp. 2d 536, 572 (D. Md. 2010) (noting that expert’s conclusion of a ballistics “match” was “only as good as the underlying photographs, sketches and notes that support it”). Oakes also repeatedly claimed to be able to determine race based on hair analysis, even using a low-powered microscope. Tr. II, 336, 345. As the FBI has made clear in the case of Manning v. Mississippi, 2013-DR-00491-SCT (Miss. S. Ct.), “an examiner cannot testify with 39 any statement of probability whether the hair is from a particular racial group.” See Ex. Y, FBI Letter of May 4, 2013, at 1–2. Oakes violated this restriction by testifying several times that the evidentiary hair was a “Caucasian head hair.” Tr. II, 352, 358, 385. Moreover, Oakes could not recall how many hairs he obtained from Perrot for his analysis, but even his highest estimate was insufficient for accurate results. Specifically, Oakes estimated that it may have been “12, 15, or 20” hairs. Tr. II, 362. NAS recommends that 50 comparison hairs be used for accurate results, id. at 354, and the FBI recommends at least 25, see also Forensic Science Communications, Part 1: Hair Evidence (July 2000), available at http://www.fbi.gov/about-us/lab/forensic-science-communications/fsc/july2000/deedric1.htm (FBI recommendation that hair examiners should “obtain at least 25 full-length hairs” for analysis). 40 5. Oakes Repeatedly And Deliberately Undermined His “Disclaimer” That Hair Is Not A Unique Identifier 41 Despite’s his inaccurate claims that hair microscopy was highly probative, that the evidentiary hair matched Perrot’s in “every respect,” and that anyone with his training and experience would conclude the same, Oakes carefully added a disclaimer to his testimony, but then did everything he could to again overinflate his identification of a “match.” Oakes conceded that hair microscopy is not the basis of a “positive” association “like a fingerprint,” and that he could not eliminate the “possibility” that the hair could “match someone else.” Tr. II, 349; see also id. at 371 (Oakes conceding that he “can’t say with one hundred percent certainty that [the] question hair positively came [from] Perrot to the exclusion of any other person in the world [because the possibility] exists that someone else may have the same characteristics.”); id. at 381 (Oakes testifying that he cannot “discount the possibility that there is someone else in the area that might match” the hair). But any fair review of Oakes’s testimony demonstrates conclusively that the so-called disclaimer was insufficient to counter the false certainty of the overwhelming majority of the testimony proffered to the jury, which falsely and repeatedly rejected the limits of hair microscopy. Moreover, Oakes asserted that no other examiner with his level of skill would possibly disagree with his conclusions. The fact that Oakes stated that he could not be “one hundred percent” certain that the hair matched Perrot was an obvious concession to reality—it nonetheless implied that Oakes’s strongly believed that when he declared a “match” it was virtually a one-to-one association between Perrot’s hair and the evidentiary hair. Oakes reiterated that false impression by simultaneously claiming a purported “match” was the basis of a strong association, that hair characteristics in individuals are “somewhat unique,” and that he had failed to identify two known hairs only a few times out of hundreds of thousands of hairs examined. Testimony that a hair match is not like a “fingerprint match” and cannot be used to exclude every person on the planet is only a necessary condition, 42 and not a sufficient condition, to accurately describe the science of hair microscopy. See Ex. X, Kafadar Aff. When the remainder of the testimony made abundantly clear that hair matches are virtually a unique identification, the examiner’s testimony could only have led the jury to believe what he was telegraphing to them: Perrot was guilty and the hair evidence proved it. Indeed, Oakes’s confidence in his own abilities and the strength of his conclusions indicate that he also believed that Perrot was guilty, and that he expected the hair evidence would prove it. As noted in the NAS Report, cognitive psychologists have found that forensic scientists are vulnerable to cognitive and contextual bias in using techniques such as microscopic hair microscopy. See Ex. Q, NAS Report, at 8 n.8.14 For example, numerous studies demonstrate that when a forensic examiner is required to make a determination that a suspect’s sample is similar to an evidentiary sample—as is true in hair microscopy, fingerprint, bullet, and bite mark analyses—examiners are significantly influenced by psychological factors such as knowing the details of the crime, pressure exerted by police with whom they are working, and appearing in an adversarial justice system. See Ex. Z, Saul M. Kassin et al., The Forensic Confirmation Bias: Problems, Perspectives and Proposed Solutions, 2 J. of Applied Research in Memory & Cognition 42, 44 (2013) (“A wealth of evidence indicates that an observer’s expectations can impact visual and auditory perception. [C]onfirmation biases are a natural and automatic feature of human cognition that can occur in the absence of self-interest and operate without conscious awareness.” (citations omitted)). Moreover, “expectancy effects” can cause humans to search out and selectively interpret evidence that confirms their hypothesis, thereby creating a “process that transforms expectations into reality.” Id. These cognitive biases would have affected Oakes, 14 See also, e.g., Ex. Q, NAS Report, at 123 (“Forensic scientists also can be affected by this cognitive bias if, for example, they are asked to compare two particular hairs, shoeprints, fingerprints—one from the crime scene and one from a suspect—rather than comparing the crime scene exemplar with a pool of counterparts.”). 43 who worked closely with the police and ADA Bloom, knew Perrot was a suspect as a result of other evidence in the case, and worked as part of the prosecution “team.” Given the subjective nature of hair microscopy, Oakes’s overconfidence in his abilities and the accuracy of his findings reflect the cognitive biases recognized by the NAS Report. Ex. Q, NAS Report, at 122 (“[C]ognitive biases are not the result of character flaws; instead, they are common features of decisionmaking, and they cannot be willed away.”). 44 V. PERROT IS ENTITLED TO A NEW TRIAL FREE FROM THE NEWLY DISCREDITED HAIR MICROSCOPY TESTIMONY THAT WAS NEVERTHELESS PRESENTED TO THE JURY AS IRREFUTABLE “SCIENTIFIC” EVIDENCE OF GUILT 45 Under Massachusetts Rule of Criminal Procedure 30(b), the Court “may grant a new trial at any time if it appears that justice may not have been done.” In ruling on a motion for a new trial, the Court must “make such findings of fact as are necessary to resolve the defendant’s allegations of error of law.” Mass. R. Crim. P. 30(b). To obtain a new trial on the basis of newly discovered evidence, the defendant must show that the evidence is both “newly discovered and that it casts real doubt on the justice of the conviction.” Commonwealth v. Grace, 397 Mass. 303, 305 (1986). Evidence is newly discovered if it was “unknown to the defendant or his counsel and not reasonably discoverable by them at the time of trial (or at the time of the presentation of an earlier motion for a new trial).” Id. at 306. Grounds for relief not raised in an original motion for a new trial are waived unless they could not have been raised in an earlier motion, or if the court exercises discretion to allow them to be raised in a subsequent motion.15 A new trial should also be granted where a defendant demonstrates that his due process rights were violated because prosecutors presented uncorrected false or misleading testimony that the Commonwealth knew was false or misleading. See Commonwealth v. Hurst, 364 Mass. 604, 608–09 (1974) (citing Napue v. Illinois, 360 U.S. 264 (1959)). Moreover, Perrot is entitled to a 15 As explained below, because Perrot could not reasonably have been aware of the FBI’s new admission, these claims could not have been raised in prior motions. Even if the Court were to find that Perrot’s claims were waived, it should exercise discretion under Rule 30(c)(2) to consider these arguments because “a miscarriage of justice” might otherwise result. Commonwealth v. Watson, 409 Mass. 110, 112 (1991). The court may grant a new trial based on claims that have been waived if there is a “serious doubt whether the result of the trial might have been different had the error not been made,” Commonwealth v. LeFave, 430 Mass. 169, 174–75 (1999), or where allowing the conviction to stand “will result in manifest injustice,” Commonwealth v. Amirault, 424 Mass 618, 666 (1997) (internal quotations omitted). Where, as here, there is a strong basis to think that the defendant is innocent and the forensic errors in question could very plausibly have resulted in a guilty verdict, there is a serious risk that a miscarriage of justice will occur unless a new trial is granted. Indeed, a trial judge in the Court previously found that the prosecution’s improper closing argument was alone sufficient to meet the standard for a “substantial risk of a miscarriage of justice.” When the Court of Appeals overturned that decision, see Perrot, 58 Mass. App. Ct. 1102, the appellate court went out of its way to note that the Commonwealth’s hair evidence “corroborated” the other “circumstantial” evidence against Perrot, and that the prosecutors admittedly unlawful statements were just another harmless error. Id. With the Commonwealth’s hair evidence discredited, the basis for the Court of Appeals reinstatement of the conviction has disappeared. 46 new trial because he received ineffective assistance of counsel in violation of his constitutional rights. See Commonwealth v. Saferian, 366 Mass. 89, 96 (1974). B The NAS Report And FBI Review Of Hair Testimony Are Newly Discovered Evidence Providing Objective Scientific Standards For Hair Microscopy That Was Previously Unavailable The FBI’s acknowledgment of its errors is “new evidence” that no amount of diligence could have uncovered, and that “casts real doubt on the justice” on Perrot’s conviction. Commonwealth v. Buck, 64 Mass. App. Ct. 760, 762, 765 (2005) (granting new trial where “defense counsel never had the opportunity to argue the various possibilities to the jury, nor did the jury have an opportunity to evaluate” new evidence). Evidence is newly discovered if it was “unknown to the defendant or his counsel and not reasonably discoverable by them at the time of trial (or at the time of the presentation of an earlier motion for a new trial).” Grace, 397 Mass. at 306; cf. Commonwealth v. Conaghan, 433 Mass. 105, 109 (2000) (“Evidence of battered woman syndrome may be considered newly discovered evidence warranting a new trial because usually there is delay in coming forward with information on the abuse, even if there were some knowledge of the abuse at trial.”). The FBI’s acknowledgment of error qualifies as “new evidence” for the simple fact that it did not exist at the time of Perrot’s trial. See Commonwealth v. Lykus, 451 Mass. 310, 330–31 (2008). In Lykus, the SJC held that a National Academy of Science Report finding voice spectography and bullet analysis techniques were unreliable and should not be admitted as evidence constituted “newly discovered” evidence to the extent that it was not available at trial. See id. at 330–31.16 Likewise, in Buck, the Massachusetts Court of Appeals held that evidence was newly discovered where the prosecution had failed to disclose the existence of a day-long 16 The Court ultimately denied the defendant’s motion for new trial, however, concluding that the “evidence against the defendant was voluminous and, ultimately, overwhelming.” Lykus, 451 Mass. at 329. No such “voluminous” and “overwhelming” evidence exists in this case. 47 surveillance tape. Buck, 64 Mass. App. Ct. at 764. Although defense counsel in Buck knew that fragments of the surveillance tape were available, the Court held that the prosecution’s failure to disclose the existence of the full-tape meant the exculpatory evidence was not available at trial. Id. Here, Perrot has never had the opportunity to present the scientific advances entirely discrediting hair microscopy science or the FBI’s acknowledgment of testimonial error at trial or in his prior motions for a new trial. Although hair microscopy has been the subject of some academic debate in recent years, the limitations of the science were only catalogued and given wide credence in the NAS Report that was published in August of 2009. The first indications of serious weaknesses in hair microscopy science appeared in the academic literature only beginning in 2002, but those scientific articles were only precursors to the NAS Report and Perrot did not have access or the specialized knowledge necessary to interpret them. Ex. AA, Perrot Aff., June 25, 2014. Moreover, the FBI’s independent review of the testimony given by FBI hair examiners and the cataloging of the three error types was not publicly acknowledged until July 2013. The extent of the false evidence in Perrot’s case did not become apparent until he retained counsel, who became aware of the NAS Report and the recently announced review conducted by the FBI. See In re Pers. Restraint Trapp, 165 Wash. App. 1003, at *5 (2011) (holding that, despite some prior indications, the extent of the FBI’s misleading testimony only became apparent “after a detailed review of the trial record by specialists at the FBI laboratory sometime in 2009”). In analogous situations, Massachusetts courts allow defendants to raise constitutional claims where they do not have a “genuine opportunity” to raise the claims before watershed Supreme Court decisions recognizing new types of constitutional error. See Commonwealth v. 48 Burkett, 396 Mass. 509, 511 (1986); see also Commonwealth v. Rembiszewski, 391 Mass. 123, 127–28 (1984). Perrot was unrepresented from the end of the appeal overturning the Superior Court’s order granting a new trial until undersigned counsel began to represent him in 2012. As a factual matter, Perrot was not aware of the scientific or forensic developments during his long incarceration, during which he had “limited access (if any) to technical, scientific research.” In re Pers. Restraint Trapp, 165 Wash. App. 1003, at *5. It would be unreasonable to expect Perrot to have discovered the flawed evidence that is the basis of this motion before the issuance of the NAS Report and the announcement of the FBI review. Buck, 64 Mass. App. Ct. at 764 (defendant not “obliged to search for something he did not know existed”).17 Notably, the DOJ has acknowledged that the FBI review of hair microscopy testimony is new evidence by waiving the statute of limitations and procedural-default defenses under 28 U.S.C. § 2255. See Ex. BB, FBI/DOJ Letter in Preston case (Aug. 20, 2013). Although DOJ cannot waive state procedural defenses, which makes it more urgent for Perrot to move now for a new trial, the DOJ’s discussion of how it is proceeding in federal cases should encourage state prosecutors to follow suit. Id. Considering the technical and highly specialized nature of the testimony at issue here, and the limited access Perrot had to scientific research or legal counsel, Perrot cannot be penalized for failing to predict that the FBI would dramatically reverse course on the scientific validity of its testimony. 17 See also Wilson v. Beard, 426 F.3d 653, 661 (3d Cir. 2005) (holding that due diligence did not require prisoner to monitor local news 12 years after conviction when there was no reasonable basis to conclude that local news would provide information on prisoner’s case); Poole v. Woods, No. 08-cv-12955, 2011 WL 4502372, at *17 (E.D. Mich. Aug. 9, 2011) (“[T]his Court does not equate reasonable diligence with requiring Petitioner to regularly scour the Detroit Free Press and Michigan Court Reporters more than a half-decade after his direct appeal was exhausted in the off-chance that something unforeseeable yet useful to his case would be found.”); United States ex rel. Fergerson v. Atchison, No. 09 C 2105, 2012 WL 581163, at *5 (N.D. Ill. Feb. 22, 2012) (“Due diligence” does not require prisoners “to hunt through haystacks trying to figure out whether one of them might contain a needle.”). 49 A. The Introduction Of The FBI’s False And Misleading Testimony Violated Perrot’s Due Process Rights The newly discovered evidence establishes that Perrot’s due process rights were violated by the introduction of false scientific evidence that was nonetheless presented to the jury as highly probative “scientific” evidence of guilt. Moreover, if this Court should find that the FBI’s admission is not “newly discovered” evidence, then it necessarily follows that the Commonwealth knew or should have known that the presentation of Oakes’s testimony regarding hair evidence it knew or should have known to be false and misleading violated Perrot’s due process rights and forms a separate basis for granting a new trial. Napue, 360 U.S. at 269 (“[A] conviction obtained through use of false evidence, known to be such by representatives of the State,” violates a defendant’s due process rights). “[T]he well-established rule” is that “a conviction obtained by the knowing use of perjured testimony is fundamentally unfair, and must be set aside if there is any reasonable likelihood that the false testimony could have affected the judgment of the jury.” United States v. Bagley, 473 U.S. 667, 678–79 (1985) (citations and quotations omitted). A defendant’s due process rights are equally as violated by a prosecutor who allows false or misleading testimony to go uncorrected. DeChristoforo v. Donnelly, 473 F.2d 1236, 1240 (1973) (“For a prosecutor to convey, or even to permit, a false impression, invades the area of due process.” (citing Miller v. Pate, 386 U.S. 1 (1967))), rev’d on other grounds, 416 U.S. 637 (1974); Commonwealth v. Gilday, 382 Mass. 166, 176 (1980) (holding that even testimony that is merely “substantially misleading” establishes a Napue violation where the prosecution allows it to go uncorrected); Hurst, 364 Mass. at 608 (When “the State, although not soliciting false evidence, allows it to go uncorrected when it appears,” the defendants due process rights are violated (quoting Napue, 360 U.S. at 269)); see also Alcorta v. Texas, 355 U.S. 28, 31–32 (1957) (misleading testimony that gives a “false impression” to the 50 jury is sufficient to create a constitutional violation). Where “false testimony could in any reasonable likelihood have affected the judgment of the jury,” a “new trial is required.” Giglio v. United States, 405 U.S. 150, 154 (1972) (quoting Napue, 360 U.S. at 271). In this case, the Commonwealth proffered “scientific” evidence that members of the prosecution team—which includes testifying FBI agents—knew or should have known was false and substantially misleading. See Kyles v. Whitley, 514 U.S. 419, 437 (1995) (prosecution’s knowledge encompasses what “others acting on the government’s behalf in the case, including the police” know).18 Knowledge of any member of the prosecution team, including FBI agents, is imputed to the Commonwealth under the SJC’s holding in Lykus. 451 Mass. at 328 (“[A]ctions of one sovereign may be imputed to another . . . [where the] heft of the Commonwealth’s case was provided by the FBI.”). The prosecutor’s office is the “spokesman for the Government,” and the “taint on the trial is no less if [the police], rather than the State’s Attorney, were guilty of the nondisclosure.” Commonwealth v. St. Germain, 381 Mass. 256, 261–62 n.8 (1980) (quoting Giglio, 405 U.S. 150 and Barbee v. Warden, Maryland Penitentiary, 331 F.3d 842, 846 (4th Cir. 1964)); Smith v. Commonwealth, 331 Mass. 585, 591–92 (1954) (“It is well understood that the duty of a district attorney is not merely to secure convictions” but “to secure them with due regard to the constitutional and other rights of the defendant.” (citing Berger v. United States, 295 U.S. 78, 88–89 (1935))). Here, even as doubts about the limits of hair microscopy were not widely known, the FBI itself was aware of such limitations and nevertheless encouraged examiners such as Oakes to 18 Indeed, because the prosecutor has a freestanding “obligation to refrain from presenting testimony at trial which he or she knows or should know is false,” the failure to correct false or misleading testimony is not excused by the defense’s knowledge of the testimony or its failure to object or cross-examine the witness. Commonwealth v. Sullivan, 410 Mass. 521, 532 (1991) (emphasis added); Sivak v. Hardison, 658 F.3d 898, 909 (9th Cir. 2011) (citing N. Mariana Islands v. Bowie, 243 F.3d 1109, 1122 (9th Cir. 2001). 51 offer “inappropriate” testimony that was not “scientifically supported.” In a deposition in the Donald Gates case, FBI Special Agent Michael Malone, a former supervisor in the FBI’s Hair and Fiber Unit, agreed that FBI agents had conversations “to the effect” of “[s]ince we didn’t have a database and we didn’t have, you know, real probabilities, scientifically valid probabilities, let’s try and use these numbers of the cases that we have looked at in lieu of real probabilities.” Ex. CC, Malone Dep. 330:15–22, Jan. 10, 2013. Put differently, Malone admitted that the FBI deliberately and inappropriately decided to proffer Type 3 errors to jurors because there was, in fact, no scientifically valid, empirical evidence to support the claims of the hair microscopists. Like Malone, Oakes specifically testified as to the number of cases he had looked at in the lab as lending credence to his ability to identify the evidentiary hair as consistent with Perrot’s hair. Tr. II, 349 (“[I]t’s my experience in 10 years [that] it’s extremely rare I will have known hair samples from two different people that I can’t tell apart,”); id. at 354 (noting he “conducted thousands of examinations in thousands of cases”). Malone’s testimony indicates that members of the FBI Hair and Fiber Unit were well aware that they lacked scientifically valid probabilities, and that testimony about the number of cases they have looked at was a misleading substitute.19 Such knowledge of FBI Agent Oakes is imputed to the Commonwealth’s prosecution team. Lykus, 451 Mass. at 328. Perrot’s due process rights under the Sixth and Fourteenth Amendments were also 19 Indeed, the FBI has known for years that Malone’s testimony had misidentified suspects. The Inspector General and DOJ reviewed Malone’s casework starting in the 1990s, and although they found widespread errors, they disclosed this exculpatory evidence only to prosecutors and not to defense counsel. See Ex. R, Spenser S. Hsu, Convicted Defendants left uninformed of forensic flaws found by Justice Department, Wash. Post, April 16, 2012. Although the Justice Department was aware that these problems could be broader, they limited the scope of their review at that time solely to Malone. Id. 52 violated by a fundamental breakdown in the adversarial process. The prosecutor and defense counsel all improperly assumed that Oakes’s forensic testimony was valid and reliable. Indeed, defense counsel never did so much as conduct a voir dire of Oakes, see Tr. II, 333, nor did he retain an expert to rebut Oakes’s false and misleading claims. Defense counsel was then powerless to refute Oakes’s specific rejection of the limitations of hair microscopy. As a result, there was no meaningful adversarial testing of critical evidence that was wrongly believed to be valid and reliable “scientific” evidence. See Estelle v. McGuire, 502 U.S. 62, 70 (1991) (“‘[T]he Due Process Clause guarantees the fundamental elements of fairness in a criminal trial’” (quoting Spencer v. Texas, 385 U.S. 554, 563–64 (1967))); cf. United States v. Cronic, 466 U.S. 648, 656 (1984) (“The right to the effective assistance of counsel is thus the right of the accused to require the prosecution’s case to survive the crucible of meaningful adversarial testing.”). The fundamental unfairness of Perrot’s trial was exacerbated by the prosecution’s closing argument—already acknowledged as improper by this Court and the Court of Appeals, the coercive interrogation of Perrot leading to a highly suspect partial “confession,” the prosecutor’s forged admission, and the long trail of police and prosecutorial misconduct regarding the collection and testing of evidence, discussed infra. B. Perrot’s Counsel Provided Ineffective Assistance By Failing To Adequately Investigate and Challenge The Commonwealth’s Hair Microscopy Testimony, And By Failing To Properly Argue The Voluntariness of Perrot’s Statement On The Night Of His Arrest Perrot also received ineffective assistance of counsel in violation of his constitutional rights. See Strickland v. Washington, 466 U.S. 668, 686 (1984). Defense counsel is ineffective under both the United States and Massachusetts Constitutions if his “incompetency, inefficiency, or inattention” fall “measurably below that which might be expected from an ordinary fallible lawyer,” and such 53 conduct “likely deprived the defendant of an otherwise available, substantial ground of defence.” Saferian, 366 Mass. at 96. Perrot’s defense counsel’s failure even to consult another expert witness to evaluate the hair evidence was unreasonable. See Strickland, 466 U.S. 691 (decision by counsel “not to investigate must be directly assessed for reasonableness in all [of] the circumstances”). Massachusetts courts have squarely held that failure to consult a forensic expert to investigate evidence that the Commonwealth will rely upon constitutes ineffective assistance of counsel. Commonwealth v. Baker, 440 Mass. 519, 529 (2006). In Baker, the Court held that defense counsel’s failure to investigate or consult a hair microscopy expert regarding the identity of an evidentiary hair constituted ineffective assistance of counsel. Id. at 529–30; see also Commonwealth v. Farley, 432 Mass. 153, 156–57 (2000) (“[N]o aspect of our adversary system could be more important than the opportunity finally to marshal the evidence for each side before submission of the case to judgment.” (citations and quotations omitted)). Numerous other Massachusetts cases similarly require defense counsel to obtain expert opinions if needed to substantiate or rebut the Commonwealth’s case. See Commonwealth v. Alvarez, 433 Mass. 93, 103–04 (2000) (defense counsel rendered ineffective assistance by failing to review or provide medical records to defense expert to substantiate theory that defendant lacked culpability because of mental defect); Commonwealth v. Haggerty, 400 Mass. 437, 441–42 (1987) (failure to seek expert opinion on whether assault was cause of victim’s death was ineffective assistance of counsel); see also, e.g., Pavel v. Hollins, 261 F.3d 210, 217–18 (2d Cir. 2001) (counsel’s failure to call important fact witnesses and medical expert at trial was ineffective assistance because testimony of those witnesses would have rebutted prosecution’s already weak case). Even though new evidence about the lack of any scientific basis for hair microscopy was not available 54 at the time of trial, by not at least consulting an opposing expert, Perrot lost the ability to challenge Oakes’s dubious and misleading hair microscopy testimony. Nor was defense counsel’s decision only to challenge Oakes on cross-examination without calling a rebuttal witness “arguably a reasoned tactical or strategic judgment”—it was simply a failure to pursue one of Perrot’s strongest defenses. See Commonwealth v. Rondeau, 378 Mass. 408, 413 (1979). Though defense counsel attempted to question Oakes about a study of scalp hairs done in 1990, he forgot the study’s authors. See Tr. II, 379–80. Following an objection, defense counsel dropped the subject entirely. Id. at 380. Likewise, defense counsel never directly attacked Oakes’s implication that he was able to distinguish between hairs and to associate them with an individual. Cf. Fisher v. Gibson, 282 F.3d 1283, 1309 (10th Cir. 2002) (defendant received ineffective assistance where, among other things, counsel’s failed attempts at cross-examination directly bolstered credibility of the prosecution’s witnesses). With such weak cross-examination and no rebuttal expert witness for the defense, the jury was left only with Oakes’s unsupported claim that he was almost certain the evidentiary hair belonged to Perrot. Moreover, defense counsel failed to appropriately assess the voluntariness of Perrot’s statement to the police and properly argue that the statement was, in fact, involuntary. The voluntariness of a criminal defendant’s statement to police is based on the totality of the circumstances. Commonwealth v. Hoose, 467 Mass. 395, 403 (2014). Factors relevant to this analysis include “promises or other inducements, conduct of the defendant, the defendant’s age, education, intelligence and emotional stability, experience with and in the criminal justice system, physical and mental condition, the initiator of the discussion of a deal or leniency (whether the defendant or the police), and the details of the interrogation, including the recitation of Miranda warnings.” Id. (quotation omitted). Here, in moving to suppress Perrot’s statement 55 as involuntary, defense counsel merely pointed to Perrot’s age, emotional state, drug consumption, and allegations that police had beaten him. Defense counsel should have stressed that the police were familiar with Perrot and that the officers responsible for investigating the elderly-women rapes arrived soon after Perrot’s interrogation began to question him about the rapes, even though Perrot was arrested for something entirely different. In fact, the very prosecutor, ADA Bloom, who was found to have later forged an affidavit attributed to Perrot also participated in the interrogation. The use of improper police tactics is critical to the voluntariness inquiry. See Commonwealth v. Ortiz, 84 Mass. App. Ct. 258, 266–70 (2013). In addition, defense counsel completely failed to explain that Perrot had no recollection of making this statement to the police. Defense counsel’s “manifestly unreasonable” decisions—in failing to investigate and challenge Oakes’s hair microscopy testimony, and in failing to properly argue that Perrot’s statement to police was involuntary—were severely prejudicial. As a result of these shortcomings in his representation, Perrot lost “otherwise available, substantial ground[s] of defense.” See Commonwealth v. Adams, 374 Mass. 722, 728 (1978). The constitutional deficiencies of Perrot’s trial counsel warrant a new trial. C. The Erroneous Hair Testimony Was Material To Perrot’s Conviction In order to assess the materiality and weight of the new evidence, the Court must compare the new evidence against the evidence of defendant’s guilt. A defendant does not need to show that “the verdict would have been different, but rather whether the new evidence would probably have been a real factor in the jury’s deliberations.” Grace, 397 Mass. at 306 (emphasis added). Instead, a new trial is appropriate where the new evidence is “material and credible” and “carr[ies] a measure of strength in support of the defendant’s position.” Id. at 56 305–06; see also Giglio, 405 U.S. at 154 (“A new trial is required if ‘the false testimony could . . . in any reasonable likelihood have affected the judgment of the jury.’” (quoting Napue, 360 U.S. at 271)). In cases such as this one, where a victim provides a description of the perpetrator that does not match the defendant, affirmatively states that the defendant is not the perpetrator when presented with him in a lineup, and where there is, at best, weak evidence of guilt and abundant police and prosecutorial misconduct, a new trial may be granted even for newly discovered evidence that bears on only minor or collateral issues. See Commonwealth v. Bennett, 43 Mass. App. Ct. 154, 162 (1997) (because case was a weak one for conviction, “it would be quite natural for the jury to pay attention to collateral factors and even to make them decisive”). A judge considering a new trial motion may consider other evidence of innocence in conjunction with the newly discovered evidence in determining whether a new trial is warranted. Grace, 397 Mass. at 312 (holding trial judge “has the flexibility . . . to consider in the interest of justice all evidence that might bear on the issues presented” (citations omitted)). 1. Scientific Evidence Is Particularly Difficult For A Jury To Evaluate Testimony offered by scientists or forensic experts is among the most persuasive evidence that can be presented to a jury. Courts have repeatedly recognized the effect of forensic witnesses, who are presumed neutral and objective thanks to their impressive credentials as scientists. “Expert evidence can be both powerful and quite misleading because of the difficulty in evaluating it.” Daubert v. Merrell Dow Pharm., 509 U.S. 579, 595 (1993) (citations and internal quotation marks omitted). Indeed, juries can be unduly influenced by the imprimatur imparted on a witness when the Court authorizes him or her to testify as an expert. United States v. Lester, 57 254 F. Supp. 2d 602, 608 (E.D. Va. 2003) (“Expert testimony has the potential to be substantially prejudicial because of the ‘aura effect’ associated with such testimony.”); State v. Young, 35 So. 3d 1042, 1050 (La. 2010) (explaining that “merely being labeled” an expert has the potential to unduly influence the jury). Further, when the topic of testimony is scientific or forensic in nature, the jury assumes that these experts are basing their testimony on objective fact and reason, and that their conclusions are not influenced by subjectivity or personal bias. Thus, with images of Sherlock Holmes lurking in the minds of jury members, scientific or forensic testimony may “assume a posture of mystic infallibility in the eyes of a jury of laymen.” United States v. Addison, 498 F.2d 741, 744 (D.C. Cir. 1974); see also United States v. Frazier, 387 F.3d 1244, 1263 (11th Cir. 2004) (“[E]xpert testimony may be assigned talismanic significance in the eyes of lay jurors”). In addition to courts, academics and the media continue to demonstrate that modern juries place undue significance on and have great difficulty in evaluating scientific evidence. As NAS explained, the various actors in the legal system “often have insufficient training and background in scientific methodology, and they often fail to fully comprehend the approaches employed by different forensic science disciplines and the reliability of forensic 58 science evidence that is offered in trial.” Ex. Q, NAS Report at 27.20 Oakes played up his role as the infallible expert to the jury. Oakes did not deem it necessary to explain his basis for concluding that the hairs were microscopically similar, and instead merely told the jury that he had concluded that they “matched.” He emphasized his extensive trial experience in over 150 cases, claiming to have examined “hundreds of thousands of hairs” over the course of “thousands and thousands of cases.” Tr. II, 354, 376. He considerably embellished his forensic powers, claiming that only an expert with “a lesser amount of training” could ever disagree with him. Id. at 374–75; see also id. at 376 (“[I]f you have qualified experienced competent hair examiners that do this day in and day out and for years on end they are not going to differ, in my opinion, on the conclusion that I reached.”). This testimony was so crucial that the prosecution reemphasized it during closing argument. Tr. IV, 619–20 (“I don’t know what we would all see if we look through a microscope. We don’t have the training that he has. ‘Would an expert disagree?’ What was his answer to that? ‘If somebody had the amount of training and experience not one of them would disagree with my opinion in this case.’”). And most egregiously, Oakes told the jury that he was almost never 20 The notion that modern juries have great difficulty in evaluating—and place undue significance on—scientific evidence is widely accepted. See, e.g., Keith A. Findley, Innocents at Risk: Adversary Imbalance, Forensic Science, and the Search for Truth, 38 Seton Hall L. Rev. 893, 948 (2008) (“[R]esearch indicates that jurors often do not understand the fundamentals of scientific evidence, and lack the ability to reason about statistical, probabilistic, and methodological issues effectively.”); Mark A. Godsey & Marie Alao, She Blinded Me with Science: Wrongful Convictions and the “Reverse CSI Effect,” 17 Tex. Wesleyan L. Rev. 481, 495 (2011) (“It is clear to me that jurors in this country often accept state forensic testimony as if each prosecution expert witness is the NASA scientist who first put man on the moon.”); Tom R. Tyler, Viewing CSI and the Threshold of Guilt: Managing Truth and Justice in Reality and Fiction, 115 Yale L.J. 1050, 1068 (2006) (“There is widespread evidence indicating that people already overestimate the probative value of scientific evidence.”); Richard H. Underwood, Evaluating Scientific and Forensic Evidence, 24 Am. J. Trial Advoc. 149, 166 (2000) (“Given their lack of scientific sophistication and innumeracy, jurors are likely to overestimate the significance of [expert testimony].”). 59 wrong about a hair match. Tr. II, at 349 (“[I]t’s my experience in 10 years [that] it’s extremely rare I will have known hair samples from two different people that I can’t tell apart.”). Thus, Oakes did everything to distance himself from laypersons, and create the image of an unrebuttable, unassailable expert dispensing objective truth to the jury. 2. The Prosecution’s Evidence Against Perrot Was Insufficient To Sustain A Conviction The Commonwealth’s introduction of Oakes’s false and misleading hair testimony was no mere “harmless error.” This Court previously held that the prosecution’s misstatements were sufficiently prejudicial on their own in an otherwise circumstantial case to warrant a new trial under Rule 30(b). See Ex. G, Perrot, No. 85-5415, slip op. Although the Massachusetts Court of Appeals found that the misstatements were “harmless error,” it did so after specifically noting that the Defendant’s guilt was “corroborated by blood and hair samples.” Perrot, 58 Mass. App. Ct. 1102, at *8. That conclusion only reinforces the gravity of error created by Oakes’s testimony, whose prejudicial effect is overwhelming when it is compared to the prosecution’s non-forensic evidence offered against Perrot. First, Prekop had at least three separate opportunities to identify Perrot as her attacker: at the lineup one week after the attack, when presented at trial with photos of Perrot one week after the attack, and when viewing Perrot in open court, at that time clean-shaven. Ex. C, Perrot Police Lineup Summ. 2; Tr. I, 141, 154–56, 158, 160. Each time, she affirmatively stated that he was not the perpetrator. Second, Perrot’s partial and circumstantial “admission” was not strong evidence of guilt. The “admission” was obtained after 12 hours of on-and-off interrogation, when Perrot had not slept more than an hour or so of the previous 36 hours and was suffering the after-effects of heavy drug and alcohol use. See Tr. II, 408; Tr. III, 436, 438, 458, 471–72, 474, 481, 484, 486, 509, 60 513–14, 536, 541, 544. There was extensive testimony that indicated Perrot was extremely emotional and suicidal when he finally signed the “confession” after the long interrogation. See, e.g., Perrot, 407 Mass. at 542 (one hour after “confession” Perrot “became emotional, began to cry, and asked for a police officer’s gun so he could shoot himself”). Indeed, this court recognized the effects of this drug abuse when granting Perrot a new trial in 1991: “There was evidence that the Defendant had ingested drugs and alcohol shortly before his arrest; that he was either not permitted to sleep or unable to sleep during the many, many hours of questioning, and that he was interrogated intermittently during a period of at least 12 hours commencing as early as 5:05 a.m.” Ex. G, Perrot, No. 85-5415, slip op. at 37. The veracity of the “confession” is further undermined by the fact that Perrot was 17 years old at the time, and the interrogation took place outside of the presence of counsel or Perrot’s parents or guardians. There is no audio or video recording of the interrogation. See Commonwealth v. DiGambattista, 813 N.E.2d 516, 533 (Mass. 2004) (expressing “skepticism of the officers’ version of events” in instances where interrogations have not been audio- or videorecorded). If the trial were to happen today, the hair testimony would be excluded and the trial court would have discretion to suppress the purported “admission” on this ground or alternatively instruct the jury that the lack of a recording requires them to “weigh evidence of the defendant’s alleged statement with great caution and care” and that the jury can conclude that the lack of the recording permits the jury to conclude that the Commonwealth has failed to prove it was voluntary beyond a reasonable doubt.” Id. at 533–34. Perrot has no recollection of signing the “admission,” and disputes whether he in fact did so. Given the fact that ADA Bloom—who was later reprimanded by the Massachusetts Bar for forging Perrot’s signature on a fabricated affidavit was present for Perrot’s interrogation—was present during Perrot’s long interrogation 61 casts serious doubts about the authenticity and voluntariness of the “confession.” Third, nor did the expert serology testimony of Eubanks provide strong evidence associating Perrot with the crime scene. With respect to the gloves, Eubanks found only one genetic marker, PGM 1+, which is present in 41 percent—or two out of every five—people. Tr. II, 245, 248. Because both Perrot and Prekop had this genetic marker in their blood, neither could be excluded as the source of the blood. Id. at 299–300. Eubanks testified that he found two bloodstains on the bedsheet, one of which did not yield serological results beyond an identification as human blood, and one of which contained four genetic markers: PGM 1+, EAP BA, HP 2-1, and Tf C. Id. at 298. Tests for ABO blood type and eight to ten other testable genetic markers produced inconclusive results on this second stain. Id. at 292. Eubanks testified that Perrot’s samples contained the same four genetic markers as the second stain on the sheet, but Prekop was excused as the source of the blood. Id. at 299–300. Eubanks further testified that only eight percent of the population had blood matching the second stain on the sheet, a figure he calculated by multiplying the population frequency of each of the four blood markers as reported in national studies. Id. at 301. Serology has only “modest powers of discrimination where the chance of two unrelated people matching is one in several hundred.” See John M. Butler, Fundamentals of Forensic DNA Typing 47 (2010). Here, however the profile was so incomplete that Eubanks could not produce results even that impressive, where even the eight percent figure he forwarded equated to more than 14,000 white males in Hampden County alone.21 Furthermore, the weak association suggested by Eubank’s serological testing is undermined by Prekop’s trial testimony calling into question the prosecution’s theory that the 21 See supra note 5. 62 blood evidence from the gloves and bedsheet belonged to the attacker at all. Prekop testified that the sexual assault took place on the floor, and that none of the attack took place on the bed. Tr. I, 163. She therefore could not see how her attacker’s blood would have ended up on her bedsheets. Id. Rather, Prekop believed that the blood on the sheets had come from her uncle, who used to live with her and had a respiratory illness that caused him to cough up blood on the sheets. Id. at 163–65. Prekop did not believe there was any fresh blood on her sheets as a result of the November 30 attack, nor did she see blood on any other objects or items of clothing, or anywhere else in the house. Id. at 169; Ex. B, Edwards Aff. Fourth, the highly conflicting testimony about the ownership of the gloves leaves them of no incriminating value. In the first place, the Commonwealth presented a conflicting and contradictory account of how the gloves were discovered. Prekop testified that she found the gloves on the floor at the foot of the bed when she returned home from the hospital. Tr. I, 144. Officer Kelly, on the other hand, first testified that he found both of the gloves on the floor. First Trial Tr. III, 195–97. At Perrot’s second trial, however, Officer Kelly claimed that he had found one glove on the floor and the other on the bed, conveniently providing a theory for how blood could have been deposited on the bedsheet. Tr. II, 215. Additionally, the ownership of the gloves was subject to conflicting testimony that undermined their evidentiary value. Perrot’s sister, Nancy Westcott, testified that she told Officer Kelly that the gloves were not hers but may have belonged to Perrot’s girlfriend. Tr. III, 546. Officer Kelly testified instead that Westcott told him that she “had a pair just like that” but could not find them at home. Tr. II, 406–07. However, in contrast to Officer Kelly’s version of events, Detective Jarvis later went to Perrot’s girlfriend to ask if the gloves were hers, MTS, 156–57, but Perrot’s girlfriend denied ownership and was able to produce her pair of gloves, Tr. III, 498–99. 63 Finally, Perrot was positively excluded from other attacks against elderly women, some involving the same victims, because of his prior stint in juvenile detention and a forensic test that excluded him as the donor of the Marchand semen stain. Without the hair microscopy evidence, the Commonwealth’s evidence was highly circumstantial and error-ridden, and the prosecution had no evidence actually linking Perrot to the crime scene. Against this backdrop, the false hair microscopy testimony most certainly “affected the judgment of the jury” and now “casts real doubt on the justice of the conviction,” therefore requiring a new trial be granted. Giglio, 405 U.S. at 154 (“A new trial is required if ‘the false testimony could . . . in any reasonable likelihood have affected the judgment of the jury.” (quoting Napue, 360 U.S. at 271)); Grace, 397 Mass. at 305 (requiring a new trial when newly discovered evidence “casts real doubt on the justice of the conviction”). 3. The Police And Prosecutorial Misconduct Evident In The Record Call Into Question The Chain Of Custody For All Of The Evidence In Perrot’s Case The remaining evidence further undermines the case against Perrot and indicates a pattern of police and prosecutorial misconduct that calls into question the chain of custody of all of the evidence at Perrot’s trial. First, Sgt. Kelly’s testimony regarding his own collection of the evidence plainly contradicts the Springfield Police Department’s evidence receipts. Kelly testified that he personally gathered the gloves and the bedsheet. Tr. I, 194–95 (“I folded the bedsheet from the corners in and recovered that. . . . I placed it into a paper bag, brought it to the police station and placed it into our evidence locker in the crime prevention bureau.”), at 196 (stating that for the “ladies knit gloves that I recovered from the bedroom of Miss Prekop … I placed them into a bag, tagged them and placed them in the evidence locker”). But the evidence tags for those items contradict Kelly’s testimony. Despite Kelly’s clear testimony that he logged the bedsheet and pillowcase, the evidence tag demonstrates that Officer Popko did so. Ex. DD, 64 S.P.D. Evid. Tag 49452. As for the gloves, inexplicably there are two evidence tags. The first was submitted by Kelly on November 30, 1985. Ex. EE, S.P.D. Evid. Tag 49449. The second was submitted by Kelly and Sgt. Louden under circumstances that appear to indicate it was submitted at a later date, but was made to appear as though it was submitted on November 30th with the rest of the evidence. More specifically, the second gloves evidence tag is dated November 30, 1985, but the date is inserted to the right of what appears to be a crossed out “12,” indicating the writer initially wrote “12” for December and then backdated the tag. This theory is bolstered by the fact that the evidence tag number is more than 40 digits later than the rest of the November 30th evidence recorded, indicating that may other pieces of evidence had been collected in the interim. Ex. FF, S.P.D. Evid. Tag 49492. Kelly was never asked and it was never explained why there were two evidence tags for the gloves. Second, Prekop testified that the perpetrator left a footprint on the door where he broke into her home. Tr. I, 151. Thus, the Springfield Police Department collected a pair of sneakers from Perrot that Detective Jarvis testified he had with the gloves as part of his investigation into the rapes. MTS, 170; see also Ex. GG, S.P.D. Evid. Tag 51436 (Dec. 7, 1985) (documenting the collection of Perrot’s sneakers). And in a report dated May 1, 1984, after the first assault of Prekop, Sgt. Louden recorded that the perpetrator left a sneaker print on her door measuring 10.5 inches.22 But Perrot wore only a size eight shoe at the time and the Commonwealth introduced no evidence of a footprint match at Perrot’s trial. Similarly, a palm print found on Lichwala’s window also did not match Perrot. Tr. III, 461–62 (ADA Vottero telling the Court at side bar that there is “a palm print that’s recovered that they’re unable to match to anybody”). This exculpatory evidence was never disclosed to the jury. 22 Perrot no longer has access to this record, but it should be available in the Springfield Police Department and/or District Attorney files. 65 Finally, it is also questionable as to why ADA Bloom—who personally drove much of the evidence to the FBI lab in Washington, D.C.—asked the FBI to test evidence that had already been tested by the Springfield Police Department’s regular contract serologist, John Abbott. By early 1987, Abbott had already tested at least the Marchand rape kit and semen stain, the Prekop rape kit, and Marchand, Prekop, and Perrot’s blood.23 Abbott concluded that Perrot was not the contributor of the semen from the Marchand rape. There was no semen available to test in the Prekop rape kit. Nonetheless, the FBI was asked to retest all of the evidence, including evidence that Bloom personally drove to the FBI after Abbott had reported his results. D. The Prosecution’s Error-Ridden Closing Argument, Repeatedly Referring To Oakes’s Hair Testimony, Sealed Perrot’s Fate As this court has already acknowledged, the prosecution’s opening and closing statements cast serious doubt on the justice of Perrot’s conviction, even without the exploitation of the false “scientific” testimony. One of the most foundational and fundamental protections of our criminal justice system is the right of every defendant to be presumed innocent until proven guilty beyond a reasonable doubt. In re Winship, 397 U.S. 358, 363 (1970) (“The [reasonable doubt] standard provides concrete substance for the presumption of innocence—that bedrock ‘axiomatic and elementary’ principle whose ‘enforcement lies at the foundation of the administration of our criminal law.’” (citation omitted)). The presumption of innocence is not a mere starting point; the presumption that a criminal defendant is innocent remains when trial begins and “continues throughout the case” until a verdict is rendered. Commonwealth v. Kane, 19 Mass. App. Ct. 129, 139 (1984). Yet the very opening lines of the prosecution’s closing argument in this case denied Perrot his constitutional birthright. The prosecutor told the jury that 23 John Abbott worked for the Serological Research Institute. Records of his testing should be available in the District Attorney and/or Police Department’s files. 66 Perrot “before you at the beginning of this trial [was] presumed innocent and no evidence had been introduced. He is not innocent anymore.” Tr. IV, 615. The prosecutor echoed this error again towards the end of closing: “He was presumed innocent. He is not any longer.” Id. at 629. Yet this constitutional deficiency was only one of many missteps. In discussing the testimony of Officer Kelly and Detective Jarvis, the prosecution asserted: What kind of world is it where George Perrot is innocent? It’s a world where two FBI agents get together and risk their reputations and careers to create evidence to reach conclusions not supported. . . . It’s a kind of world where Tom Kelly and Mr. Jarvis get together to create evidence to pin on this poor man. It’s the kind of world where for him to be innocent there has to be such a conspiracy you can’t believe. . . . That’s preposterous. Its [sic] preposterous. Id. at 626–27. This is a textbook example of improper vouching for the government’s witnesses. See Commonwealth v. Ciampa, 406 Mass. 257, 265 (1989); Commonwealth v. Griffith, 45 Mass. App. Ct. 784, 788 (1998). The prosecution further compounded that mistake by referring to highly damaging evidence—about the former prosecutor forging a confession purportedly by Perrot, to be used in interrogating other suspects—that had been specifically excluded as a result of the Commonwealth’s pretrial motion in limine. This, once again, is textbook error. See Commonwealth v. Mosby, 11 Mass. App. Ct. 1, 9 (1980) (“[A] party’s success in excluding evidence from the consideration of the jury does not later give that party license to invite inferences (whether true or, as in . . . false) regarding the excluded evidence.”). In September 2001, the Hampden County Superior Court granted Perrot a new trial because of the prosecution’s error-ridden closing argument. The court focused on the cumulative effect of these many errors, explaining that it was “unpersuaded that the prosecutor’s pervasive and persistent improper comments in closing did not materially influence the guilty verdict in this case.” Ex. G, Perrot, No. 85-5415, slip op. at 39. Though the Massachusetts Court of Appeals reversed this ruling, it largely based its conclusion on the fact that hair microscopy 67 evidence linked Perrot to the crime scene. 58 Mass. App. Ct. 1102, at *8 (“[T]he Commonwealth’s case was strong, based on evidence of a confession circumstantially identifying the defendant as the victim’s attacker, corroborated by blood and hair samples consistent with the defendant’s samples” (emphasis added) (quotation marks omitted)). The NAS Report and FBI review fatally undermine this conclusion. CONCLUSION For these reasons, Perrot respectfully requests that the Court grant his Motion for a New Trial. 68 Dated: July ___, 2014 Respectfully submitted, __________________ Susanna G. Dyer BBO# 671694 Christopher J. Walsh BBO# 685252 ROPES & GRAY LLP Prudential Tower 800 Boylston Street Boston, MA 02199-3600 (617) 235-4736 susanna.dyer@ropesgray.com christopher.walsh@ropesgray.com Nicholas C. Perros (Admitted pro hac vice) David A. Young (Pending admission pro hac vice) ROPES & GRAY LLP One Metro Center 700 12th Street, NW, Suite 900 Washington, DC 20005-3948 (202) 508-4663 nicholas.perros@ropesgray.com david.young@ropesgray.com 69 CERTIFICATE OF SERVICE I HEREBY CERTIFY that a true and accurate copy of the foregoing Memorandum in Support of Defendant’s Motion for a New Trial was sent by Fedex to the Office of the District Attorney, Hall of Justice, 50 State Street, Springfield, MA 01102, this ___ day of July, 2014. ___________________________ Susanna G. Dyer (BBO# 671694) 70