The Problem of False Confessions in the Post-dna world

Steven A. Drizin^** & Richard A. Leo^***

Introduction 894

I. The Role of False Confession in the Study of Wrongful Conviction 901

II. The Social Psychology of Police Interrogation and False Confession 907

III. Methodology and Sources of Data 924

IV. False Confessions and Case Outcomes: Quantitative Trends 932

IV. False Confessions and Case Outcomes: Qualitative Trends 963

1. Ryan Harris 964

1. Allen Jacob Chesnet 969

1. Michael Gayles 971

1. Colleen Blue 974

1. Multiple False Confessions to the Same Crime 974

a. Frank Kuecken and Jonathan Kaled 977

2. Multiple Innocent Defendants Arrested, Charged, and Convicted Based on Co-Defendant’s False Confession 981

a. Calvin Ollins, Larry Ollins, Marcellus Bradford, and Omar Saunders 981

3. Multiple False Confessions from Same Defendants to Multiple Crimes: The So-called Serial Killer Cases 985

a. Innocents Who Falsely Confess to Crimes Committed by Serial Killers 986

i. Jerry Frank Townsend, Frank Lee Smith, and Eddie Lee Mosley 986

b. Closing Open Cases by Falsely Attributing Unsolved Murders to Guilty Defendants 992

i. Hubert Geralds, Derrick Flewellen, and Andre Crawford 992

1. David Saraceno 994

2. Teresa Sornberger 994

Conclusion 995

Introduction

In April 1989, a young woman was attacked while jogging in New York City’s Central Park. The jogger entered the park near 84th Street shortly after 9:00 p.m., traveled north along the East Drive, and then turned onto the 102nd Street Crossdrive heading west. At approximately 9:15 p.m.,² she was knocked down and dragged into a ravine where she was raped and sodomized. She was beaten so severely, particularly in the area of her left eye, that she lost nearly eighty percent of her blood.³

On the same evening, a large group of teenage boys, with estimates ranging as high as forty to fifty boys, entered Central Park near East 110th Street in Harlem and began walking south along the park’s East Drive. The boys subsequently encountered Antonio Diaz, who was eating dinner and drinking beer in the park that night. Some of the boys proceeded to beat up Diaz and dragged him into a nearby thicket of bushes.⁴ This group of boys then continued heading south on East Drive, harassing several cyclists along the way.⁵ Sometime thereafter, a police car passed the group, causing them to break up temporarily before regrouping near some ball fields in the park’s North Meadow. According to Jermain Robinson, who was one of the boys in the group that night, the group then left the ball fields, went south toward 97th Street, squirmed through a hole in the fence, and hid in some bushes in close proximity to the northern edge of the reservoir. While concealing themselves in the bushes, the boys waited for joggers to pass.⁶

Sometime before 9:30 p.m., five of the boys tried to assault David Lewis as he jogged near the reservoir. As Lewis sped past the boys’ hiding place, one of the boys struck him on the elbow with a blunt object. A second jogger, David Good, claimed that he was chased by a group of ten African-American youths who threw stones at him as he ran past them. Robert Garner, a third jogger, was chased and caught by fifteen to twenty youths on the reservoir’s northern edge. As they assaulted Garner, the mob made demands for his money. When Garner said that he had none, they released him. Finally, a fourth jogger, John Loughlin, a six-foot-four-inch ex-Marine, seeing that Garner was in trouble, approached the group. Loughlin was unable to render assistance, however, due to a blow to the head with a blunt object, later thought to be a solid metal pipe, which left him temporarily unconscious.⁷

Two plain-clothes officers, Eric Reynolds and Robert Powers, responding to complaints about the mayhem, spotted fifteen to twenty boys on Central Park West near 100th Street at around 10:15 p.m.⁸ As their squad car approached the group, all but two of the boys scattered. Steven Lopez, age fifteen, and Raymond Santana, age fourteen, remained at the scene and answered Officer Reynolds’s questions, insisting that they were not part of the larger group.⁹ Reynolds and a third officer, Ivelisse Flores, ran after the boys who had fled, eventually catching Kevin Richards and Clarence Thomas, both of whom were fourteen years old at the time.¹⁰ When Richardson and Thomas both identified Lopez and Santana as being part of their group, all four boys were arrested.¹¹ A fifth boy, Lamont McCall, age thirteen, was also among the first arrestees.¹²

Shortly before 1:00 a.m., two men, Benicio Moore and Carlos Colon discovered the body of the female jogger. The men were walking home when, after hearing moaning sounds in the darkness, they went to investigate the source of the noise.¹³ Prior to this discovery, the police officers had been focusing only on the Diaz assault, the attempted assaults on the cyclists, and the attacks on the joggers. Because the jogger’s body was discovered near the location where Diaz and the cyclists had been accosted, however, the police suspected that the boys involved in those crimes were also responsible for raping and beating the female jogger.¹⁴ Throughout the night and the next day, Manhattan North Homicide detectives interrogated the boys already in custody and apprehended others who had been named as accomplices. Antron McRay, age fifteen, was taken into custody at 11:00 a.m., and Yusef Salaam, age fifteen, and Kharey Wise, age sixteen, were both brought in at 10:00 p.m. on the next day.¹⁵ Two hours before the arrest of Salaam and Wise, prosecutors Elizabeth Lederer and Linda Fairstein arrived at the station to assist the detectives in the interrogations. Lederer and Fairstein worked in the Sex Crimes Unit of the District Attorney’s Office, and their involvement at this early stage of the investigation indicated that the D.A.’s office also believed the boys to be responsible for the sexual assault. Their arrival also coincided with the next phase of the investigation: videotaping the boys’ confessions to the rape of the Central Park Jogger.¹⁶

Ultimately, prosecutors were able to obtain five confessions to the rape of the Central Park Jogger; four of these confessions were captured on videotape and the fifth was an alleged “oral confession.” A sixth defendant, Steve Lopez, who had been identified as a ringleader by all of the other boys, refused to admit to any participation in the rape. Although the confessions were videotaped and most of the boys confessed in the presence of their parents,¹⁷ the earlier interrogation sessions had not been taped.

Precisely what happened during the hours of police interrogations was a matter of great dispute both in pre-trial motions and at trial. The boys and their parents claimed that the interrogations were highly coercive, alleging that officers slapped the boys,¹⁸ yelled and cursed at them, and called them liars. Several boys claimed that they were told that they were being questioned as mere “witnesses” and that they would be released from custody if they only confessed.¹⁹ The police officers denied that they used coercive tactics, although one detective did admit that he lied to Yusef Salaam when he told Salaam that his fingerprints would be found on the victim’s jogging shorts.²⁰ After hearing both accounts of what transpired during hearings for the defendants’ pre-trial motions to suppress their confessions, Judge Thomas Galligan found that the police detectives were more credible than the defense witnesses and ruled that the defendants’ statements were admissible in their trials.²¹

Five of the Central Park defendants took their cases to trial. The first trial’s defendants were Raymond Santana, Yusef Salaam, and Antron McCray; the second involved Kharey Wise and Kevin Richardson. All five defendants were convicted of participating in the rape of the jogger and the assaults on several of the cyclists and the other joggers. Jurors found Kharey Wise, who gave two conflicting taped confessions, not guilty of the rape but guilty of a lesser charge of sex abuse for “playing with the jogger’s legs.”²² Only Kevin Richardson, who had hair consistent with the jogger’s on his clothing and who was named as one of the boys who beat the jogger with a rock, was found guilty of attempted murder.²³ All the boys who went to trial were sentenced to between five and fifteen years in prison.²⁴ Steve Lopez was never brought to trial on the rape but pled guilty to one of the assaults near the reservoir.²⁵

In January 2002, a convict named Matias Reyes contacted authorities and informed them that he, acting alone, had raped the Central Park Jogger. Reyes was one of New York City’s most notorious serial rapists. Between June 1989 and his apprehension in August of that year, Reyes terrorized the Upper East Side, raping four women, one of whom, a pregnant woman, he killed after raping her in front of her children.²⁶ More significantly, when Reyes’s DNA was compared to that recovered from the Central Park Jogger crime scene, there was a match: semen stains on the jogger’s sock were proven to have come from Reyes. This newly discovered evidence prompted the Manhattan District Attorney’s Office to launch a reinvestigation of the case.

In the Fall of 2002, attorneys retained by three of the boys learned that DNA testing tended to exonerate their clients of the rape and filed a motion to vacate the convictions.²⁷ Meanwhile, Reyes gave a nationally televised interview in which he provided a detailed description of the assault and rape of the jogger, going so far as to draw a map of the area in which the attack took place.²⁸ Reyes also claimed in the interview that he did not know any of the boys who were convicted of the rape.²⁹ After an exhaustive eleven-month investigation, the Manhattan District Attorney’s Office was unable to establish any link between Reyes and any of the five defendants.³⁰

Additional new evidence emerged, much of which tended to undermine the validity of the boys’ convictions. At both trials, the prosecutors had stated that “hair consistent with the jogger’s” was found on Kevin Richardson’s clothing.³¹ Mitochondrial DNA testing—a technique not scientifically possible at the time of the trials—later demonstrated that the hairs were probably not the jogger’s.³² Similarly, hair and blood recovered from a rock found near the crime scene which prosecutors suggested was the murder weapon was found not to have been the jogger’s.³³

In light of this exculpatory evidence, the District Attorney’s Office ultimately decided to join the motion to vacate the boys’ convictions.³⁴ The District Attorney’s fifty-eight page memorandum in support of the defense motion outlines why prosecutors chose to believe Reyes’s confession that he acted alone and why they gave it greater weight than the boys’ videotaped confessions.³⁵ On December 19, 2002, Judge Charles Tejada of the New York Supreme Court (a trial court) granted the motion and vacated all of the convictions of the original Central Park Jogger defendants.³⁶

District Attorney Robert M. Morgenthau, and attorneys in his office, deserve credit for reinvestigating the jogger case when compelling evidence of the boys’ innocence first came to light. Rather than re-investigate such claims, police and prosecutors often vigorously defend the conviction, using the fact that a jury or judge must have found the confessions to be reliable in order to convict as justification for refusing to reopen the case.³⁷ Indeed, District Attorney Morgenthau’s decision to support the defense motion to vacate the convictions was sharply criticized by former prosecutors from his office, police officers involved in the original investigation, and others connected to the case.³⁸ In fact, the New York City Police Department conducted its own investigation and released its own report.³⁹ Although ultimately concurring with the D.A.’s decision to join in the motion to vacate, the NYPD report took issue with many of the D.A.’s factual findings, attacked the credibility of Matias Reyes’s statements, and advanced several theories in which both Reyes and the original defendants could have participated in the assaults.⁴⁰

The discovery that Matias Reyes’s DNA matched the DNA taken from a sock found at the Central Park crime scene triggered the re-investigation of the Central Park jogger case.⁴¹ This Article will examine how DNA testing has changed our understanding of wrongful prosecution and conviction in America, focusing specifically on the phenomena of interrogation-induced false confession. This Article will document and analyze more than 125 false confessions, the largest such cohort ever assembled. Unlike previous studies of false confessions, however, the confessions considered in this Article have been proven to be false. The Central Park Jogger case is representative of many of the trends we have observed in the cohort, including issues of youth, mental disability, multiple false confessions in a single case, and the utility of DNA testing in overcoming resistance to the notion of false confession.

In a development that can be traced directly to the increased use of DNA testing, most of the confessions in the cohort have been proven false in the past five years.⁴² Primarily as a result of DNA testing, approximately two-thirds of the exonerations have occurred pre-trial, rather than post-conviction.⁴³ In this way, DNA testing has helped to minimize the consequences of false confessions.

This Article proceeds as follows: Part I discusses, from a historical perspective, the study of wrongful convictions and the prominent role that false confessions have played in such studies. Part I also discusses the development of DNA testing and its role in renewing interest in the study of wrongful convictions. Part II highlights the connection between police interrogation methods and false confessions, focusing principally on the social psychology of false confessions and research on the causes and consequences of false confessions. Part III discusses the methodology used to compile and analyze the false confessions that make up this Article’s cohort, defines critical terms, and discusses the limitations of the data. Part IV sets forth the quantitative findings gleaned from the cohort. Part V takes a more qualitative approach to the data set, highlighting some of the common themes and trends that emerge from the cohort cases and describing illustrative cases in some detail. Finally, Part VI concludes this Article with several policy recommendations suggested by the aforementioned findings, and highlights some recent positive developments which suggest that reforms designed to reduce the frequency of false confessions may stand a better chance of being implemented now than ever before.

I. The Role of False Confession in the Study of Wrongful Conviction

The study of miscarriages of justice (i.e., wrongful conviction) in America begins with Edwin Borchard’s pioneering book, Convicting the Innocent.⁴⁴ Arguing against the conventional wisdom that innocent people are never convicted in the American criminal justice system, Borchard detailed sixty-five convictions in which innocent individuals were wrongfully prosecuted, convicted, and incarcerated.⁴⁵ The thrust of Borchard’s then-pioneering research was to shift the research question away from whether factually innocent individuals are wrongfully convicted in the American criminal justice system to the questions of why they are wrongfully convicted and what can be done to remedy the problem. Borchard identified a number of causes of wrongful conviction—e.g., eyewitness misidentification, perjured testimony, and police and prosecutorial misconduct—as well as policy solutions to reduce the frequency of wrongful conviction.⁴⁶

Subsequent empirical studies of error in the administration of criminal justice have elaborated on the multiple causes of wrongful conviction first identified by Borchard.⁴⁷ In all of these studies, the problem of false confession has been featured prominently as one among many of the leading causes of wrongful conviction. As discussed in more depth below, however, more recent studies have identified false confession as the leading or primary cause of wrongful conviction in anywhere from 14–25% of the sample cases studied. To a large extent, Borchard’s pioneering study laid down the template that subsequent empirical studies of miscarriages of justice would follow for many years to come. From the time Borchard published his book in the early 1930s until the advent of DNA testing in the late 1980s, there was typically one book or major article published every decade or so on the subject of miscarriages of justice, often following the same general format and repeating the same arguments but with newer (and sometimes even more compelling) cases. Although Borchard identified sixty-five cases of wrongful conviction, his book was primarily descriptive rather than analytical: Borchard briefly described how the error occurred, how it was later discovered, and how the original case against the innocent defendant subsequently unraveled. However, Borchard did not quantify, tabulate, or systematically analyze the causes of error in the cases he studied. Similarly, Erle Stanley Gardner’s The Court of Last Resort,⁴⁸ Jerome and Barbara Frank’s Not Guilty,⁴⁹ and S. Radhakrishnan’s The Innocents,⁵⁰ all follow the same format as Borchard’s classic, documenting and describing cases in which the State mistakenly prosecuted, convicted, and incarcerated the wrong man. In all three works, the specific causes of wrongful conviction—including the problem of false confession—are discussed in relation to the cases that are presented, but the authors make no attempt to quantify or systematically study the number and characteristics of false confessions (or other errors that they describe).

Until the late 1980s, there was no systematic, social scientific study of the causes, patterns, and consequences of miscarriages of justice in America. This changed with Hugo Bedau and Michael Radelet’s 1987 watershed study, “Miscarriages of Justice in Potentially Capital Cases,” published in the Stanford Law Review.⁵¹ Identifying 350 cases of wrongful conviction in potentially capital cases in America from 1900–1987, Bedau and Radelet systematically analyzed the causes of these errors, the sources of discovery of the error, and the number of innocents who had been executed.⁵² Significantly, Bedau and Radelet’s sample found that false confessions played a causal role in 49 of the 350 miscarriages of justice studied, approximately 14% of the cases in their sample. Bedau and Radelet’s article has been influential for a number of reasons. Most fundamentally, it introduced the largest and most compelling data set on wrongful convictions into the literature; that at least 350 individuals have been wrongfully convicted of capital crimes in the twentieth century is highly disturbing, if not downright horrifying.⁵³ Moreover, approximately 90% of the 350 wrongful convictions that Bedau and Radelet documented were based on official declarations of innocence.⁵⁴ Thus, even if one disputes Bedau and Radelet’s conclusion in any particular case, it would be difficult to meaningfully dispute the larger pattern of their findings. Bedau and Radelet have influenced numerous others to research and write about the causes and consequences of wrongful conviction;⁵⁵ they have inspired others to reanalyze and extend the insights offered by their data;⁵⁶ and they have continued to collect, analyze, and publish studies of wrongful convictions in capital cases.⁵⁷

Following Bedau and Radelet’s widely cited Stanford Law Review article, the 1990s were a period of renewed energy and activism in the study of miscarriages of justice. Unlike in the preceding six decades, journalists, lawyers, and scholars published a number of books in the 1990s on the problem of wrongful prosecution and conviction,⁵⁸ signaling a new and deepening interest in the study of miscarriages of justice. To be sure, most of the articles and books published in the 1990s were in the Borchard tradition of case description and policy prescription⁵⁹ or, alternatively, were individual case studies.⁶⁰ Nonetheless, these works created an emerging and expanding critical mass in the study of wrongful conviction, calling attention to old issues in new ways (or at least with newer cases), and laying the groundwork for the biggest, and potentially most significant, development yet in the academic study of miscarriages of justice.

The most significant development in wrongful conviction scholarship in the 1990s was the advent of increasingly sophisticated forms of DNA testing and the application of this new technology to criminal investigation, particularly in post-conviction cases in which a defendant had long claimed his conviction was erroneous and there remained biological evidence from the crime with which to conclusively test the convicted prisoner’s claim. DNA testing has established the fact of wrongful conviction in scores of cases, including capital cases.⁶¹ The earliest statement of DNA testing’s ability to conclusively establish the fact of wrongful conviction was contained in Edward Connors, Thomas Lundregan, Neil Miller and Tom McEwen’s study of twenty-eight wrongful convictions⁶² in which the testing of DNA evidence subsequently established the incarcerated prisoner’s innocence. In this study, approximately 18% (5/28) of the convictions were attributable to false confessions.⁶³ In the eight years since the publication of the Connors study, DNA testing has become increasingly sophisticated,⁶⁴ and numerous other wrongfully convicted individuals have been exonerated, declared innocent, and released from prison.⁶⁵ Barry Scheck and Peter Neufeld, co-founders of the Innocence Project at the Cardozo School of Law, and others, have continued to work on cases in which DNA testing has established factual innocence and led to the release of wrongfully convicted prisoners.⁶⁶ As of the year 2000, when Scheck and Neufeld (along with New York Times journalist Jim Dwyer) published Actual Innocence: Five Days to Execution and Other Dispatches from the Wrongly Convicted, sixty-two factually innocent individuals had been exonerated by DNA evidence.⁶⁷ Of those cases, approximately 24% (15/62) involved false confessions.⁶⁸ At the time of this writing 140 wrongly convicted prisoners have been exonerated and released as a result of DNA testing. Approximately 25% (35/140) of these wrongful convictions were caused by false confession.⁶⁹

The advent of DNA testing and the window it opened onto the errors of the legal system has permanently altered the nature and study of miscarriages of justice in America. Most importantly, DNA testing has established factual innocence with certainty in numerous post-conviction cases, so much so that it has now become widely accepted, in the space of just a few years, that wrongful convictions occur with regular and troubling frequency in the American criminal justice system, despite our high-minded ideals and the numerous constitutional rights that are meant to procedurally safeguard the innocent against wrongful conviction. It is one thing for Bedau and Radelet to argue, based on their own judgment of the totality of the facts and documentary record in individual cases, that hundreds of innocent individuals have been wrongfully convicted and incarcerated; it is quite another thing for DNA testing to establish prisoners’ factual innocence in case after case. Notwithstanding judgments of innocence from criminal justice and/or political officials, the former can always be disputed and impugned as the “subjective” interpretation of the scholar; the latter can be established conclusively and beyond dispute. More so than at any time since Borchard’s seminal book in 1932, the advent of DNA testing in the 1990s has established the problem as not whether, or how frequently, miscarriages of justice occur, but why they occur so frequently and what can be done to prevent and remedy them.⁷⁰

As we have seen from this cursory review of the miscarriage of justice literature, only a few studies have systematically aggregated, quantified, and analyzed the causal role of false confession in wrongful conviction cases. These studies report that the number of false confessions range from 8–25% of the total miscarriages of justices studied, thus establishing the problem of false confessions as a leading cause of the wrongful convictions of the innocent in America. Table 1 below lists these studies. If we remove the findings from one methodologically flawed (and otherwise questionable) study,⁷¹ the percentage of false confession in the miscarriages of justice studies ranges from 14–25%.

Table 1 The Percentage of False Confessions in Prior Studies of Wrongful Convictions