When the system gets it wrong: understanding the systemic causes of wrongful conviction

When an innocent person is convicted, the consequences extend far beyond a courtroom. A person loses years – sometimes decades – of freedom. Families grow up and grow older without them. Careers, relationships, birthdays, funerals and ordinary days disappear. The person who committed the crime may remain free. Survivors are denied the truth. Communities are left less safe, and confidence in the justice system is profoundly damaged.

Since 1989, the National Registry of Exonerations has documented thousands of exonerations across the United States. The Tennessee Innocence Project has helped secure ten exonerations since opening its doors in 2019 – approximately one-quarter of all state-court exonerations in Tennessee since 1989. Those ten Tennesseans collectively endured more than 300 years of wrongful imprisonment.

These figures are staggering, but they are not the full measure of the problem. An exoneration reveals the decisions, practices and pressures that allowed an innocent person to be convicted – and made the error extraordinarily difficult to correct.

Wrongful convictions are often described as the result of a single bad witness, flawed test or dishonest official. In reality, they are usually cumulative. A mistaken identification becomes more dangerous when investigators develop tunnel vision. An unreliable forensic opinion carries more weight when the defense lacks the resources to challenge it. A false accusation becomes a conviction when favorable evidence is not disclosed or when a jury is not given the information it needs to evaluate a witness’s credibility.

The lesson is clear: wrongful conviction is not merely a problem of individual mistakes. It is a systemic problem, and preventing it requires us to understand how the system’s failures interact.

Eyewitness misidentification: sincere memory can still be wrong

Eyewitness testimony can be deeply persuasive. A witness who points to a defendant and says, “That is the person I saw,” can give a jury a sense of certainty that other evidence cannot. But human memory does not operate like a video recording. It is reconstructed each time it is recalled and can be affected by stress, poor lighting, distance, the presence of a weapon, the passage of time, suggestive questioning and feedback from investigators.

Cross-racial identifications present an additional, well-documented risk. People generally have greater difficulty accurately identifying faces of a race different from their own. Yet jurors may not understand this phenomenon without proper instructions or expert testimony.

The danger is visible in the story of Eric Wright. In 1989, two men robbed a Memphis convenience store. Fingerprints recovered from the scene did not match Eric or his brother, and no physical evidence connected Eric to the crime. The case rested on a single cross-racial eyewitness identification made through procedures later recognized as deeply compromised. Eric was only 17 years old. He spent nearly 36 years imprisoned before his release in 2025.

Eric’s case shows why the method used to obtain an identification matters. Lineups should be administered by someone who does not know which person is the suspect, so that conscious or unconscious cues cannot influence the witness. Witnesses should be told that the perpetrator may not be present. Suspect photographs should not stand out. The witness’s confidence should be recorded immediately, in their own words, before later feedback can inflate it. Whenever practical, the entire procedure should be recorded.

These safeguards do not diminish eyewitnesses or discount the trauma they have endured. They protect witnesses from being unintentionally steered toward the wrong person – and protect everyone from the consequences of a mistaken identification.

False confessions: why innocent people sometimes admit to crimes

To many people, a false confession seems impossible. Why would anyone confess to something they did not do? Exoneration data and decades of research show that innocent people do confess, particularly when interrogations are prolonged, coercive or psychologically manipulative.

Young people and people with intellectual disabilities, mental illness, language barriers or limited understanding of the legal system face heightened risk. Exhaustion, hunger, isolation and fear can erode judgment. An innocent person may be falsely told that evidence proves their guilt, may come to believe that confession is the only way to end the interrogation, or may be promised – directly or implicitly – that going along with investigators will allow them to go home.

False confessions can sound convincing because details may be disclosed during questioning, intentionally or unintentionally, and then repeated by the suspect. In Innocence Project data on DNA exonerations, more than half of the false confessors examined had included nonpublic facts in their statements. The apparent “inside knowledge” was therefore not reliable proof that they committed the crime.

Recording custodial interrogations from beginning to end is one of the most important safeguards. A complete recording allows judges, jurors, attorneys and reviewing courts to see not only the final statement but also how it was obtained. Protections for children and vulnerable adults – including meaningful access to counsel – are also essential.

False accusations, incentivized testimony, and unreliable witnesses

Some wrongful convictions begin with an accusation that is mistaken. Others involve testimony shaped by pressure, self-interest or an offered benefit. Jailhouse informants, alleged accomplices and other incentivized witnesses may expect reduced charges, shorter sentences, money or favorable treatment in exchange for cooperation. That incentive does not automatically make testimony false, but it creates an obvious reason to scrutinize it closely.

The National Registry of Exonerations reported that perjury or false accusation appeared in 73% of the 97 exonerations that occurred in 2025. That category is broad and can include lies by civilian witnesses, informants, alleged victims, co-defendants or officials. The prevalence of the factor underscores a fundamental point: a factfinder cannot fairly assess testimony without knowing what a witness has been promised, what benefits the witness hopes to receive, whether the witness has offered similar information in other cases and whether the account is corroborated by independent evidence.

Reforms can include pretrial reliability hearings, full disclosure of benefits and expectations, careful jury instructions, corroboration requirements and centralized tracking of informants. Transparency is not an obstacle to prosecution; it is necessary to distinguish trustworthy evidence from testimony produced by pressure or reward.

Misapplied forensic science: when the authority of “science” exceeds the evidence

Scientific evidence can identify the guilty and clear the innocent. But not everything presented in a courtroom as science has been adequately validated, and even sound methods can be overstated or misapplied. Analysts may reach beyond what testing supports, use subjective methods without sufficient safeguards, rely on outdated assumptions or describe a possible association as a definitive “match.” Jurors may give those conclusions exceptional weight because they are delivered by an expert.

In 2025, false or misleading forensic evidence contributed to 40% of the year’s recorded exonerations, according to the National Registry of Exonerations. The figure is a reminder that accuracy requires more than a laboratory or an expert title. It requires validated methods, appropriate standards, quality assurance, independence, disclosure of limitations and language that accurately conveys uncertainty.

Tennessee has seen the human cost of flawed medical conclusions. Joyce Watkins and Charlie Dunn were convicted in the death of Joyce’s four-year-old great-niece after a medical examiner used scientifically baseless reasoning to conclude that the child’s injuries occurred during the nine hours she was in their care. That opinion became the cornerstone of the prosecution. Later medical review established that the injuries had occurred before Joyce and Charlie picked the child up.

Joyce and Charlie each spent 27 years imprisoned. Charlie died in 2015 before his name was cleared. On January 12, 2022 – exactly seven years after his death – the charges against both were dismissed. At age 73, Joyce became the first Black woman exonerated in Tennessee.

Flawed medical testimony also shaped Wayne Burgess’s case. Wayne was convicted in the death of his girlfriend’s infant daughter and spent nearly 25 years imprisoned. In 2023, a Tennessee judge vacated his conviction after concluding that the prosecution’s trial theory was medically impossible. The same medical examiner involved in Joyce and Charlie’s case had contributed to Wayne’s conviction; his medical license had been permanently revoked years earlier following findings that included falsified autopsy reports, false statements, negligence and incompetence.

Science evolves. Courts must have meaningful ways to reconsider convictions when the scientific foundation of a prosecution has collapsed or when an expert’s work has been discredited.

Official misconduct and tunnel vision

Police officers and prosecutors carry enormous responsibilities, and most work in good faith to protect their communities. But the system must be designed to detect and correct misconduct when it occurs. In 2025, official misconduct was identified in 72% of the year’s recorded exonerations. It was present in 78% of homicide exonerations.

Official misconduct can include concealing evidence favorable to the defense, pressuring witnesses, presenting testimony known to be false, using improper interrogation tactics, fabricating evidence or making misleading statements in court. Because officials control much of the evidence at the beginning of a case, failures at this stage can distort everything that follows.

Even without deliberate misconduct, tunnel vision can have a similar effect. Once investigators settle on a suspect, they may interpret ambiguous facts as proof of guilt and discount facts pointing elsewhere. Evidence is no longer assessed neutrally; it is filtered through the theory that the right person has already been found. Confirmation bias can affect investigators, forensic analysts, prosecutors, defense attorneys and judges alike.

Strong disclosure rules, open-file practices, independent conviction review, enforceable consequences for withholding evidence and a culture that rewards accuracy over conviction rates all help reduce these risks. Prosecutors should be supported in reassessing a case when new evidence raises serious doubt. Correcting an injustice is an act of public service, not an admission that the pursuit of justice has failed.

Inadequate defense and unequal resources

The constitutional right to counsel is a cornerstone of American justice, but a right on paper is not enough. Defense attorneys need time, manageable caseloads, training, investigators, experts and funding. Without them, counsel may be unable to locate witnesses, test physical evidence, consult a forensic specialist, challenge an identification procedure or uncover information that undermines the prosecution’s theory.

Inadequate legal defense contributed to one-third of the exonerations recorded in 2025. This does not mean every overburdened lawyer is ineffective. It means that a profoundly unequal contest – one side backed by the investigative power of the state and the other without basic resources – creates an unacceptable risk of error.

Poverty should never determine whether a person can prove their innocence. Adequately funding indigent defense, setting realistic workload standards and providing access to independent experts are public-safety investments. A reliable adversarial system depends on both sides being able to test the evidence.

Racial bias and the unequal distribution of error

The causes of wrongful conviction do not operate in a vacuum. Racial bias can influence who is viewed as suspicious, whose account is believed, how cross-racial identifications are conducted, which communities are policed most aggressively, whose defense is adequately resourced and how ambiguous evidence is interpreted.

Black people are disproportionately represented among exonerees, particularly in cases involving murder, sexual assault and drug offenses. Bias may be explicit, but it is often embedded in routine decisions and institutional practices. Because no single actor must intend a discriminatory outcome for unequal outcomes to occur, race-conscious safeguards and careful data collection are essential.

Eric Wright’s cross-racial identification and Joyce Watkins’s place in history as Tennessee’s first exonerated Black woman remind us that the story of innocence work is inseparable from the pursuit of racial justice. A system committed to accuracy must be equally committed to examining where error falls most heavily.

Why these causes compound one another

The percentages above overlap because wrongful convictions rarely have a single cause. Imagine a frightened eyewitness exposed to a suggestive lineup. Investigators become confident in the identification and stop pursuing another lead. A forensic analyst is told which suspect police favor and interprets an ambiguous result accordingly. A jailhouse informant offers a story after learning case details. Favorable evidence is not disclosed. An under-resourced defense cannot hire an expert or fully investigate. Each piece appears to corroborate the others, but all may flow from the same flawed starting point.

This is why prevention cannot rely on one reform. Recording interrogations will not cure unreliable forensics. Better lineups will not remedy hidden evidence. Improved disclosure will not help if a defense lawyer lacks time to review what is produced. Every safeguard is part of an interconnected system.

Exoneration is correction – and prevention

The National Registry of Exonerations recorded 97 exonerations in 2025. Those individuals lost more than 1,095 years in total – an average of more than 11 years each. Yet known exonerations represent only the errors that have been discovered and successfully corrected. Most crimes do not yield testable DNA evidence, and post-conviction claims face daunting legal and practical barriers. Evidence may be lost or destroyed. Witnesses disappear. Filing deadlines expire. Once a conviction is final, the burden shifts heavily to the person asserting innocence.

That is why the work of the Tennessee Innocence Project matters. Each year, TIP receives approximately 150 to 200 applications for representation. Every application must be carefully reviewed, records located, evidence examined and claims investigated. Since 2019, TIP’s work has helped ten innocent Tennesseans reclaim their names and freedom. Their cases have also exposed recurring weaknesses – from unreliable identification practices and flawed medical testimony to the enormous difficulty of correcting a conviction after the fact.

Exonerations do not undermine the justice system. They strengthen it by revealing where safeguards failed and showing how future harm can be prevented. The reforms illuminated by these cases are practical: evidence-based identification procedures; complete recording of interrogations; reliable, transparent forensic practices; full disclosure of favorable evidence; scrutiny of incentivized witnesses; adequately funded defense; conviction review; preservation of evidence; and accessible pathways for courts to consider credible claims of innocence.

When an innocent person is convicted, there are at least two injustices: the wrong person is punished, and the truth remains hidden. The response cannot end when prison doors open. We must learn from what happened, repair what we can and change the conditions that allowed it to happen.

The people exonerated in Tennessee lost more than 300 years behind bars. We cannot give those years back. But we can honor them by refusing to treat wrongful convictions as isolated accidents – and by building a system in which truth is pursued at every stage, accountability is welcomed and justice always retains the courage to correct itself.

Sourcees

• Tennessee Innocence Project, “Our Story,” “Our Impact,” and exoneree case profiles: https://www.tninnocence.org/
• National Registry of Exonerations, 2025 Annual Report: https://exonerationregistry.org/sites/exonerationregistry.org/files/documents/2025Exonerations.pdf
• Innocence Project, “Explore the Numbers”: https://innocenceproject.org/exonerations-data/
• Innocence Project, “Eyewitness Misidentification”: https://innocenceproject.org/eyewitness-misidentification/
• Innocence Project, “Five Facts About DNA Testing and Wrongful Convictions in 2026”: https://innocenceproject.org/news/five-facts-about-dna-testing-and-wrongful-convictions-in-2026/