Reading Time: 16 minutes

When Larry Johnson walked out of a Missouri prison during the summer of 2002, exonerated by DNA testing from a wrongful rape conviction after avowing his innocence for 18 years, St. Louis legal community insiders nodded knowingly as word trickled out who had led the prosecution back in 1984—Nels C. Moss Jr.

Moss, assistant circuit attorney for the city of St. Louis and later a trial prosecutor in neighboring St. Charles County, earned a well-deserved reputation as an aggressive, effective trial prosecutor. During his 33 years of trying cases for the people, however, he simultaneously was a recidivist breaker of the rules by which prosecutors are supposed to operate.

After joining the St. Louis city prosecutor’s office in 1968, Moss found his conduct formally challenged in at least 24 cases. In seven of those, judges reversed the conviction, declared a mistrial or issued some other ruling adverse to the prosecution.

Over the course of his career as a prosecutor, Moss reneged during trial on a pre-trial stipulation with the defense; called the jury’s attention to the defendant’s failure to testify, thereby compromising the Fifth Amendment rights of the accused; alluded to the defendant’s uncharged criminal conduct, a violation of the rules of evidence; attacked the character of the defendant with information not in the court record; used inadmissible material from a separate trial of an accomplice; promised during jury selection or opening argument to present testimony never offered; attacked the truthfulness of defense counsel; cast aspersions on the integrity of an insanity defense; and inflamed jurors’ passions during closing argument.

When one appellate panel reversed a conviction in a case won by Moss, a judge writing a concurring opinion emphasized that the blame lay with the prosecutor and not with the courts:

“Most regrettable … is the fact that we are required to remand this case for retrial, with all of the expense, delay and inconvenience attendant thereto, because of a trial incident that need not and should not have occurred … It was a deliberate effort by one of the most experienced assistant circuit attorneys in the City of St. Louis to interject even more poison than his extensive review of defendant’s prior convictions had already accomplished … The sole purpose … was to poison the minds of the jurors regarding the defendant’s character … Where, as in this case, the record discloses a patent effort to deprive a defendant of a fair trial, the onus for the delay and added expense should be directed toward the prosecutor who caused it. This is especially true when, rather than resulting from youthful zeal, the error is but one example of a consistent pattern of improper tactics reflected by other transcripts in cases tried by the same experienced prosecutor.”

In another 17 cases prosecuted by Moss, appellate judges affirmed the conviction or trial judges allowed the proceeding to continue, despite finding Moss committed prosecutorial error.

Moss declined an interview request from the Center for Public Integrity, and wrote that he sees no point in subjecting himself to “second guessing by those that have not walked in my shoes.” In his response to the interview request, Moss characterized himself as “a hard-hitting but honest prosecutor.” He estimated that he tried more than 400 cases before juries, including “high-profile, racially and politically explosive cases. Obviously the friends and representatives of those convicted are dissatisfied with the outcomes and are prone to see fault and perceived injustice.”

Of judges who criticized his tactics, Moss wrote that some “have never tried cases as prosecutors or defense attorneys and have never experienced the heat of the courtroom.” Moss wondered if his detractors realize that, “I have refused to proceed on numerous cases where confessions did not match the evidence, where identifications did not measure up to appropriate standards, where alibis while not conclusive have left too much reasonable doubt. … I have never approved or sponsored testimony I suspected to be false. I have never prosecuted anyone on the basis of race; indeed most of the victims of the crimes I prosecuted were minorities.”

Though Moss’ record of 7 reversals due to misconduct and 17 other findings that he committed prosecutorial error is extreme, he is hardly an anomaly.

Recidivist prosecutors

Local prosecutors in many of the 2,341 jurisdictions across the nation have stretched, bent or broken rules while convicting defendants, the Center has found. Since 1970, individual judges and appellate court panels cited prosecutorial misconduct as a factor when dismissing charges at trial, reversing convictions or reducing sentences in at least 2,012 cases. The nature of the questionable conduct covers every type attributed to Moss, and more.

In 513 additional cases, appellate judges offered opinions—either dissents or concurrences—in which they found the prosecutorial misconduct serious enough to merit additional discussion; some of the dissenting judges wrote that they found the misconduct warranted a reversal. In thousands more cases, judges labeled prosecutorial behavior inappropriate, but allowed the trial to continue or upheld convictions using a doctrine called “harmless error.”

The Center analyzed 11,452 cases in which charges of prosecutorial misconduct were reviewed by appellate court judges. In the majority of cases, the allegation of misconduct was ruled harmless error or was not addressed by the appellate judges, and the conviction stood. The relative rarity of reversals makes these opinions useful from an empirical standpoint: Any prosecutor who has more than one reversal to her credit belongs to a select club.

Prosecutorial misconduct falls into several categories, including:

  • Courtroom misconduct (making inappropriate or inflammatory comments in the presence of the jury; introducing or attempting to introduce inadmissible, inappropriate or inflammatory evidence; mischaracterizing the evidence or the facts of the case to the court or jury; committing violations pertaining to the selection of the jury; or making improper closing arguments);
  • Mishandling of physical evidence (hiding, destroying or tampering with evidence, case files or court records);
  • Failing to disclose exculpatory evidence;
  • Threatening, badgering or tampering with witnesses;
  • Using false or misleading evidence;
  • Harassing, displaying bias toward, or having a vendetta against the defendant or defendant’s counsel (including selective or vindictive prosecution, which includes instances of denial of a speedy trial);
  • Improper behavior during grand jury proceedings.

Some of the most common allegations of prosecutorial misconduct involved improper closing arguments and excluding jurors on the basis of race, ethnicity, gender or some other discriminatory grounds.

In 28 cases, involving 32 separate defendants, misconduct by prosecutors led to the conviction of innocent individuals who were later exonerated, the Center found. Innocent men and women were convicted of serious charges, including murder, rape, and kidnapping and assault.

Guilty defendants have also had their convictions overturned. Sometimes those defendants cannot be retried because of double jeopardy rules, and are placed back on the streets of the community. In other words, prosecutorial misconduct sometimes has severe consequences for the entire citizenry, not just a lone defendant.

In addition, the Center found some prosecutors who had convicted innocent defendants in more than one case over the course of their careers; some of these prosecutors were cited multiple times for misconduct in other cases as well.

Most of the nation’s approximately 30,000 local trial prosecutors strive to balance their understandable desire to win—a desire supported by the vast majority of the citizenry—with their duty to ensure justice. There are some prosecutors, however, who have exalted winning and ignored the other half of the equation. Those prosecutors who repeatedly break the rules give recidivism—a word usually used to describe those they work to put behind bars—a disturbing new meaning.

It is impossible to know for sure how often a specific prosecutor (or a specific defense attorney, judge, police officer, etc.) bends or breaks the rules. In most jurisdictions, at least 95 percent of the cases that pour in from the police never reach a jury, which means any misconduct occurs away from public view. The only trial those defendants receive takes place in the prosecutor’s office; the prosecutor becomes the judge and the jury. The prosecutor is the de facto law after an arrest, deciding whether to charge the suspect with committing a crime, what charge to file from a range of possibilities, whether to offer a pre-trial deal, and, if so, the terms of the deal.

Katherine Goldwasser, a law professor at Washington University in St. Louis who served as a prosecutor in Chicago before joining academia, suggested that misconduct often occurs out of sight, especially in cases that never go to trial. Those cases by definition do not generate appellate opinions (and thus are for the most part beyond the scope of the Center study). Goldwasser told the Center. “It is not a safe assumption that cases ending with guilty pleas are absent prosecutorial misconduct.”

Perhaps the most difficult type of misconduct to unearth, Goldwasser said, is the failure of the prosecutor to turn over possibly exculpatory information to the defense. Such lack of disclosure is commonly known as a “Brady violation,” after the 1963 U.S. Supreme Court case Brady v. Maryland and its progeny. After all, if only police and prosecutors know about evidence that suggests innocence, how is defense counsel to know for certain such evidence even exists?

To complicate quantification, any listing of mistrials and appellate reversals involving a specific prosecutor might be incomplete. While legal databases like Lexis and Westlaw (both of which were used in this study) contain appellate rulings, some remain unpublished, and those that are published rarely identify the trial prosecutor. And, short of visiting every courthouse in the country, there is no way to determine how many cases are dismissed or ruled mistrials by trial judges (and thus never reaching the appellate courts) because of a prosecutor’s misconduct.

Despite those limitations in the data, the study determined that, like Moss in St. Louis, other prosecutors around the country have been found by appellate court or trial court judges to have bent or broken the rules multiple times.

In Cuyahoga County, Ohio, Carmen Marino, who served for 30 years as a prosecutor before retiring in 2002, won five convictions that were overturned by the Ohio appellate courts. Appellate judges have ruled that Montgomery County, Ala., District Attorney Ellen Brooks’ discriminatory tactics deprived defendants of fair trials four times since she began prosecuting in 1977. Former Hinds County, Miss., District Attorney Edward Peters was involved in six cases in which judges ruled that his conduct prejudiced a defendant.

The pattern of behavior of John Zimmermann, a trial prosecutor in Davidson County (Nashville), Tenn., so alarmed six former Tennessee prosecutors that, during July 2002, they filed an amici curiae brief to the U.S. Supreme Court on behalf of Death Row defendant Abu-Ali Abdur’Rahman. The six, who constitute a who’s who of the Tennessee legal profession, cited Zimmermann’s misconduct in the case – confirmed by the state Supreme Court but nevertheless ruled as harmless error – and his behavior in previous, unrelated cases.

The brief argues that Zimmermann withheld evidence from the defense and misrepresented a prior conviction of the defendant. Even more troubling, it cites Zimmermann’s conduct during three other, unrelated cases, including a murder case in which the verdict was overturned because of the Davidson County prosecutor’s behavior. (In a response to the Center for Public Integrity, Zimmerman vociferously contests the charges.)

Fighting Misconduct

Perhaps nothing better demonstrates the serious problem prosecutorial error and misconduct pose to the justice system than the efforts of a handful of jurisdictions to combat it. In Boston, Ralph Costas Martin II, who became district attorney of Suffolk County in 1992, worked hard to change a system that had been characterized by one prominent defense attorney as “the best place to have a guilty client, and the worst one to have an innocent client,” according to journalist Sean Flynn’s book, Boston D.A.: The Battle to Transform the American Justice System.

Among numerous examples, Flynn explains how Martin began hiring young trial prosecutors who understood the concept of elevating justice over winning at all costs. Job candidates would be asked how they would handle the following scenario: On a busy morning, with three dozen cases stacked up, you, the prosecutor, are approached by defense counsel in one of those cases. “My client will plead out if you recommend probation,” the defense lawyer says. You skim the case file, which contains nothing but a police report. It says the officer stopped the car because the driver allegedly made a “furtive gesture.” While searching the stopped car, the officer says he found a small amount of cocaine. What, the job candidate is asked, should a good prosecutor do?

The obvious answer seems to be accept the plea. But that is not the answer Martin wanted. Instead, he hoped job candidates would say they would ask the judge for a delay, in order to question the arresting officer privately. The first question of the officer should be to describe the “furtive gesture.” Much of the time, Martin believed, police officers listed that rationale to cover up a bogus stop of a racial minority or some other targeted group. Unless the police officer offered a plausible explanation, Martin hoped the job candidate would say the case should be dismissed. Why? Because an alert defense lawyer or judge will understand the traffic stop was a pretext, thus leading to suppression of the evidence that resulted—the cocaine. The prosecutor moving ahead with the case will lose the trust of the judge and the defense lawyer, and will have done nothing to halt the police officer’s improper behavior.

In San Diego County, elected District Attorney Paul J. Pfingst and his staff introduced a number of innovations, including distribution of a comprehensive training manual to all attorneys in the office, initiation of post-conviction DNA review of 766 pre-1992 cases, and even allowing television cameras to follow prosecutors around as they did their work. The videotape became the basis of a national television show that aired weekly on NBC during the summer of 2002. But not even all the legitimate reforms could save Pfingst from defeat at the polls in November 2002. Too many voters perceived his office as housing prosecutors whose misbehavior resulted in high-profile mistrials and appellate reversals.

Within a jurisdiction, elected prosecutors can preside over a culture of misconduct in their offices. Unelected, nearly anonymous prosecutors, like Nels Moss in St. Louis, do not operate in a vacuum. Individual prosecutors accused of misconduct must be understood within the context of the culture of the office to which that prosecutor belongs.

The St. Louis Circuit Attorney’s office has served as a home for aggressive trial prosecutors, including Moss. Sometimes that aggressiveness leads to brilliant lawyering within bounds. Other times it leads to rule breaking.

Since 1970, the Center study found, there were 129 rulings by trial judges and appellate judges, including the cases tried by Moss, that addressed alleged prosecutorial error by the circuit attorney’s office.

Those cases involve at least 40 St. Louis city prosecutors other than Moss. Of the 129 rulings, 45 resulted in reversals or acquittals. Another 13 upheld convictions, but at least one appellate judge issued a dissent in favor of reversing the conviction. The remaining 71 rulings found prosecutorial error, but the judges allowed the convictions to stand, without any dissents.

A culture of misconduct

Bennett L. Gershman, a former New York County (Manhattan) prosecutor who now teaches law at Pace University, is an authority on professional conduct. Gershman, who has written textbooks and law review articles about prosecutorial misconduct, testified at Congressional hearings, conducted seminars and provided perspective to countless journalists, told the Center about St. Louis, “Compared to other offices that I have looked at, the number of reversed cases for prosecutorial misconduct from a single office is rather large, particularly since reversals are not commonplace given the various techniques used by appellate courts to try to uphold convictions.”

Moss is not the only prosecutor in the St. Louis office to be cited multiple times for error and misconduct. Three assistant circuit attorneys other than Moss were cited or reversed multiple times for misconduct. Appellate judges cited misconduct when reversing three convictions won by Joseph W. Warzycki, who joined the circuit attorney’s office in 1977. In five other cases, defendants alleged misconduct, which was ruled as harmless error. Gordon L. Ankney, now in private practice, was reversed twice for misconduct. John D. Chancellor was reversed once for misconduct; in 13 other cases, defendants alleged misconduct but appellate courts ruled it harmless error. Chancellor left the prosecutor’s office in 1987 to become a trial judge, a position he held until his death in 1991.

Though St. Louis-area defense lawyers mention Warzycki, Ankney, Chancellor and other prosecutors from time to time, Nels Moss almost always is mentioned first, and discussed with the greatest amount of passion. Like so many other prosecutors who bend or break the rules in jurisdictions across the nation, Moss has never been publicly sanctioned by his office supervisors or by the state bar disciplinary counsel.

Gershman, the Pace University professor, said of Moss’ record that the number of “reversed cases from one prosecutor’s misconduct is fairly astounding.”

Ronald Weich, a former assistant district attorney in New York County (Manhattan), has studied the atmosphere of prosecutors’ offices and the conduct of individual prosecutors in those offices from his perch at the Washington, D.C., law firm of Zuckerman Spaeder. The co-author of a study for the Leadership Conference on Civil Rights that addresses what he terms “the unequal treatment of minorities in the exercise of prosecutorial discretion,” he told the Center that, “Prosecutorial abuses often arise from structural problems and an utter lack of accountability within the office.” When the elected prosecutor cares little about serving justice, that attitude trickles down to the trial prosecutors, Weich says. Even when the elected prosecutor sets the right tone, Weich believes many offices house what he calls “cowboys” who ignore reasonable doubt. Moss’s career in St. Louis suggests that both office culture and his own aggressiveness played a part in his “astounding” record of reversals due to misconduct.

Moss served 16 of his 33 years as a prosecutor under the supervision of former Circuit Attorney George A. Peach, the elected district attorney. Peach, who first won election in 1976, resigned in 1992 because of personal financial and sexual improprieties. A staff prosecutor before his election to the top spot, Peach is one of the more than 40 prosecutors whose conduct led to reversals and other findings of error.

Dee Joyce-Hayes and Moss worked as trial prosecutors from 1981 until 1992 under Circuit Attorney Peach. Both Joyce-Hayes and Moss wanted to succeed Peach after he stepped aside in 1992. Joyce-Hayes won the job.

Joyce-Hayes told the Center she thought of firing Moss, partly because she believed he behaved unfairly toward her as a political opponent, partly because of his over-aggressiveness as a trial lawyer. She retained Moss, she said, to keep from angering some of their office colleagues, to refrain from alienating certain St. Louis power brokers who admired Moss—and because he could win convictions that might have eluded other prosecutors. But, Joyce-Hayes added, “I isolated him in the homicide unit; I did not make him a team leader. I worried about him having too much contact with impressionable young assistant circuit attorneys.”

The current elected prosecutor in St. Louis is Jennifer M. Joyce, who joined the staff before Moss left but never served as his supervisor. Joyce, who has served about half of her first four-year term, has instituted a number of reforms aimed at cleaning up the office culture of St. Louis.

Though Moss’ elected supervisors failed to discipline or rein him in, he was ultimately responsible for his own conduct in the courtroom. And that conduct earned him the mistrust of lawyers who opposed him at trial. Some defense attorneys began to assume that Moss would bend or break rules.

During 1999, shortly after he left his position in St. Louis to prosecute cases in neighboring St. Charles County, Mo., Moss entered a high-profile murder case. Appellate courts had overturned the defendant’s death sentence two separate times before his involvement.

Upon Moss’ entry, public defenders collaborated on a petition they said was unprecedented in their experience: They asked the judge to bar Moss’ participation. Failing that, they asked the judge, before the proceeding even began, to admonish Moss “against engaging in any effort to circumvent the constitution … the rules of evidence, and from any effort to make prejudicial arguments, speaking objections or other improper remarks within the hearing of the jury.” The public defenders told the judge they had “no desire to engage in vexatious bickering, yet Mr. Moss comes to this case with a record of deliberate misconduct that sharply jeopardizes defendant’s right to and hope for a fair, constitutional sentencing hearing.”

The judge allowed Moss to enter the case. The trial got off to a bad start, as defense counsel had feared. The public defender complained to the judge that “Yesterday morning before we began opening statements, we … received from the prosecuting attorney, Nels Moss, an endorsement of four new witnesses and then a number of items, a photo album. … The matters that have now been disclosed are brand new witnesses, documents, items that have not previously been disclosed. … We are at a significant disadvantage.” The judge did not grant a mistrial. Moss eventually argued in favor of the death penalty, which the jury granted him.

Defense counsel appealed, alleging improper conduct by Moss. The Missouri Supreme Court agreed that, once again, Moss had conducted himself improperly. The court, however, upheld the death sentence, ruling Moss’ overall performance, despite some specific “improper” conduct, could not be proved to have prejudiced the jurors:

“The three trials of this case unfortunately exhibit a consistent attempt by the prosecutor to push the envelope of proper advocacy. We condone the prosecutor’s strategy no more in this trial than in the previous two that were reversed.”

Judges scolded Moss in others cases, too, where they nevertheless upheld the conviction under a doctrine called “harmless error.” Prosecutors and defense counsel disagree on the term’s meaning. The former tend to emphasize the word “harmless,” while the latter tend to stress the word “error.”

Moss’ partisans note that appellate judges—those with no trial experience, those with criminal defense backgrounds and even those who once served as prosecutors—sometimes misinterpret the antiseptic written trial record from which they work. As a result, appellate judges might find prosecutorial misconduct where none was intended. Those same partisans emphasize that Moss helped convict hundreds of guilty defendants without allegations of improper conduct arising on appeal.

Convicting the innocent

There is no doubt, however, that in cases like Larry Johnson’s, unambiguously innocent defendants suffer for a long time while the perpetrators remain at liberty.

Determining unintentional error or intentional misconduct by the prosecution in an actual innocence case can be difficult; sometimes the prosecutor has sound reason to believe in the suspect’s guilt until new evidence, or new ways to evaluate old evidence, emerge. The justice system is acknowledged by all its participants to be imperfect, and even when there is no misconduct, when there are no lapses—either intentional or unintentional—on the part of the police, the prosecutor, the judge, or the defense counsel, an innocent defendant can go to prison. At virtually any step in a trial, from the initial questioning of a suspect through the marshalling of forensic evidence and experts to closing arguments and appellate maneuvering, errors by the state—prosecutors and police—can convict the innocent.

In the Johnson case, Moss had plenty of reason at first to think the defendant guilty, given the accused rapist’s criminal past and the victim’s eyewitness identification. It is also worth noting that DNA testing, which eventually cleared Johnson, was not easily available nor particularly credible when the case first came to trial.

The rape that led to Johnson’s arrest occurred Jan. 31, 1984. The next day, the victim helped a police artist produce a composite drawing of her assailant. He appeared clean-shaven, based on the victim’s recollection. With the composite completed, the police artist showed the victim about 140 photographs of possible suspects. The victim set one of those photographs aside. The man pictured wore a mustache, contradicting the victim’s original description. Moss asked the police artist to add a mustache to the composite. The artist complied.

Later that day, the victim picked Johnson from a police line-up. He had facial hair, contradicting the original physical description and the original composite.

When Johnson’s public defender asked the victim questions at a deposition in late June, she said she was “fairly certain” the photograph she chose from the 140 looked like her assailant. During the August trial, a police crime analyst said he found sperm, but said nothing about using it to type the perpetrator’s blood. Defense counsel asked about blood typing. Moss objected, and the judge sustained the objection. Why? Because the judge opposed allowing forensic tests into evidence.

As a result of Moss’ objection and the judge’s ruling in Moss’ favor, the jurors convicted Johnson based solely on eyewitness testimony. He lost his only appeal, limited to jury selection issues, in 1986. So Johnson remained in prison, with no realistic hope of a hearing to reconsider his conviction, until 1995, when a letter he wrote to the Innocence Project staff in New York City meshed with their interest in post-conviction DNA testing.

Lawyer Barry Scheck, director of the Innocence Project, sued the St. Louis prosecutor’s office toward the end of Dee Joyce-Hayes’ term. Current prosecutor Jennifer Joyce inherited the litigation.

The Innocence Project staff claimed the prosecutor’s office was obstructing post-conviction DNA testing in at least six rape cases, including Johnson’s. Joyce responded there had never been obstruction. She said testing is being done, as time and budget permit, in accordance with a new Missouri law and U.S. Justice Department guidelines. Scheck and his staff are failing to consider the anguish of rape victims when cases are re-opened, Joyce said, as well as ignoring the common-sense theorem that testing should occur only if the results can definitively establish innocence or guilt. Since the litigation began, Joyce took the initiative to start examining about 1400 pre-1994 convictions, to determine if DNA evidence exists and, if so, whether it makes sense to test it. Joyce estimated each case will require about 10 hours of initial review after the paperwork is gathered.

The first scientific testing related to the Innocence Project litigation, on behalf of convicted rapist Fred Hamilton, confirmed his guilt. The finding cast a shadow over Scheck’s effort and led St. Louis prosecutors to complain about wasting time and money, as well as forcing Joyce’s staff to, as she put it, “re-victimize the victim” by asking her about sexual relations with men other than the rapist. “It was a gut-wrenching interview,” Joyce said. “It reminded me of my time as a sex-crimes prosecutor.” She wondered whether Hamilton believed he might benefit from a testing mistake, or whether he was acting sadistically.

Then the tables turned. The second DNA test, negotiated in court on behalf of Johnson, demonstrated a wrongful conviction. After spending 18 years behind bars for a crime he did not commit, Larry Johnson finally attained his freedom. And a longstanding miscarriage of justice was finally undone.

Help support this work

Public Integrity doesn’t have paywalls and doesn’t accept advertising so that our investigative reporting can have the widest possible impact on addressing inequality in the U.S. Our work is possible thanks to support from people like you.