The leading causes of wrongful convictions are a short, well-documented list: mistaken eyewitness identifications, forensic methods that were never scientifically validated, false confessions pulled from long interrogations, misconduct by police or prosecutors, jailhouse informants trading testimony for leniency, defense lawyers who lack the time or resources to mount a real case, and plea bargains that make trial too risky even for innocent people. The National Registry of Exonerations has documented more than 3,700 cases since 1989, and eyewitness error alone contributed to roughly 69 percent of convictions later overturned by DNA testing.1National Registry of Exonerations. National Registry of Exonerations Home
These causes rarely appear alone. A shaky identification gets reinforced by a confident expert on the stand, a tired suspect confesses after being told fabricated evidence exists, and a defense attorney without the budget to hire a specialist can’t push back on any of it. Each failure amplifies the others.
Eyewitness Misidentification
Human memory does not work like a camera. Under stress, witnesses tend to fixate on a weapon rather than the person holding it, which sharply reduces how accurately they can later describe a face. Each retelling subtly reshapes the original memory, and by trial a witness may be recounting a reconstruction rather than a genuine recollection.
Lineup procedures often make the problem worse. When the officer running a photo array knows which person is the suspect, subtle body language or a shift in tone can steer the witness toward a particular choice. If the suspect is also the only person in the array who matches the general description, selection is close to guaranteed. These errors appear in roughly 69 percent of DNA-based exonerations.2Innocence Project. DNA Exonerations in the United States
Confidence is another trap. Witnesses who felt uncertain during the initial identification frequently become far more definitive by the time of trial. Juries treat a confident witness as a reliable witness, but research shows no meaningful link between how sure a person feels and how accurate their identification actually is.
The most effective safeguard is a double-blind, sequential lineup, in which photos are shown one at a time by an officer who does not know which person is the suspect. False identifications drop significantly because the administrator can’t give cues they don’t have and the witness evaluates each photo against memory rather than comparing faces side by side. Best practice also calls for a standardized script and a recorded confidence statement taken immediately after the identification, before outside influences inflate certainty. Adoption of these reforms across jurisdictions remains uneven.
Flawed Forensic Science
Forensic evidence carries enormous weight with juries, but several widely used techniques were never validated with the rigor expected of actual science. In 2016, the President’s Council of Advisors on Science and Technology (PCAST) evaluated seven common feature-comparison methods and found that bite mark analysis, footwear analysis, firearms examination, and microscopic hair comparison all lacked “foundational validity,” meaning no adequate studies had demonstrated these methods to be repeatable, reproducible, and accurate.3Executive Office of the President, President’s Council of Advisors on Science and Technology (PCAST). Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature-Comparison Methods PCAST specifically noted that bite mark analysis was “far from meeting” scientific standards and that footwear analysis had zero appropriately designed studies supporting its ability to link a shoe print to a specific shoe.
The FBI’s own review of microscopic hair comparison testimony illustrates how deep the problem runs. After auditing trial transcripts, the Bureau acknowledged that examiners gave erroneous testimony in at least 90 percent of the cases reviewed.4FBI. FBI Testimony on Microscopic Hair Analysis Contained Errors in at Least 90 Percent of Cases in Ongoing Review Post-PCAST court decisions have gone different directions: some judges exclude or limit this evidence while others continue to admit it, so admissibility depends heavily on jurisdiction.5National Institute of Justice. Post-PCAST Court Decisions Assessing the Admissibility of Forensic Science Evidence
The deeper issue is how expert testimony is framed. An examiner who tells a jury that a fiber is “consistent with” the suspect’s clothing without explaining how common that fiber is in the general population leaves jurors to assume the match is meaningful. A claim of a “100 percent match” using a subjective comparison method overstates what the science supports.
False Confessions
People confess to crimes they did not commit more often than most jurors would believe. Research on known false confession cases found that the average interrogation producing a false admission lasted about 16 hours, and 84 percent occurred after six or more hours of questioning.6Innocence Project. False Confessions After that kind of sustained pressure, exhaustion and desperation override rational self-interest. The person confesses so the interrogation will stop.
The Reid Technique, the most widely taught interrogation approach in American policing, relies on isolating the suspect and applying escalating psychological pressure while presenting the confession as the most reasonable path forward. Critics identify specific features that contribute to false confessions, including misreading truthful suspects as deceptive, coercive psychological manipulation, and feeding non-public case details that the suspect then incorporates into their story.7Connecticut General Assembly. OLR Research Report – Interrogation Techniques
Police in every state are legally permitted to lie to adult suspects during interrogations. Officers routinely claim to have DNA matches, fingerprints, or eyewitness identifications that do not exist. A handful of states have begun banning this practice for juvenile suspects, but for adults, fabricated evidence remains a lawful interrogation tactic. Young people and individuals with intellectual disabilities are especially vulnerable because they may not grasp the long-term consequences of what they are signing. Once a confession exists, juries struggle to accept that anyone would admit to a crime they did not commit, and that intuition is difficult for defense attorneys to overcome even when physical evidence points elsewhere.
Mandatory recording of custodial interrogations is the most direct safeguard. The Department of Justice adopted a policy in 2014 requiring its primary law enforcement agencies to electronically record custodial statements beginning as soon as the subject enters the interview area.8Department of Justice Office of the Inspector General. Audit of the Department of Justice’s Implementation of its Policy to Electronically Record Statements of Arrestees in Custody Roughly 30 states and the District of Columbia now require recorded interrogations in some form, though scope and exceptions vary.
Official Misconduct
When police or prosecutors cut corners, hide evidence, or steer a case toward a predetermined conclusion, the system’s built-in safeguards fail. The most consequential form of prosecutorial misconduct is suppressing evidence favorable to the defense. Under the Supreme Court’s decision in Brady v. Maryland, prosecutors must turn over any evidence that could help the accused, including anything that undermines the credibility of a government witness. Withholding this material strips the defense of its ability to mount a meaningful challenge at trial.
Investigative misconduct takes many forms: coaching witnesses, ignoring leads that point away from the primary suspect, and in extreme cases, fabricating or planting physical evidence. High-profile cases create intense public pressure to deliver a quick arrest and conviction, and that pressure can warp the priorities of an entire investigation. Investigators develop tunnel vision, lock onto a suspect early, interpret ambiguous evidence as confirmation, and dismiss contradictory information as noise. This kind of confirmation bias is not always deliberate dishonesty, which makes it harder to detect and prevent.
Consequences for prosecutors who violate their disclosure obligations are remarkably light. Under the Supreme Court’s 1976 decision in Imbler v. Pachtman, prosecutors enjoy absolute immunity from civil lawsuits for actions taken while initiating, pursuing, and presenting a criminal case.9Justia. Imbler v Pachtman A wrongfully convicted defendant cannot sue the prosecutor for damages even if the prosecutor deliberately hid exculpatory evidence. The Court acknowledged this leaves “genuinely wronged” defendants without a civil remedy but concluded that personal liability would chill the independent judgment the role requires. Professional discipline remains theoretically available and is rarely imposed in practice.
Jailhouse and Incentivized Informants
Informants have a straightforward incentive to fabricate testimony: they get something valuable in return. Reduced sentences, dropped charges, transfer to a better facility, or cash payments all give an informant a reason to tell prosecutors what they want to hear. These witnesses tend to have extensive criminal histories and enough experience with the legal system to construct a plausible story from publicly available case details or information overheard in a jail setting.
Prosecutors lean on informant testimony most heavily when physical evidence is thin. The informant fills the gap in the narrative, and juries rarely see the full picture of what the witness received in exchange. Federal data on sentence-reduction motions shows cooperating witnesses receive an average sentence cut of about 37 percent, with some reductions reaching well above 50 percent depending on the district and offense type.10United States Sentencing Commission. Rule 35(b) Sentence Reductions: An Empirical Analysis When years of prison time hang in the balance, the temptation to deliver useful testimony is enormous regardless of whether it is true.
Inadequate Defense Representation
The constitutional right to an attorney means very little when that attorney lacks the time, resources, or competence to mount a real defense. Under Strickland v. Washington, a defendant challenging their conviction on these grounds must show two things: that their lawyer’s performance fell below a reasonable professional standard, and that the deficiency actually changed the outcome of the case.11Justia. Strickland v Washington, 466 US 668 (1984) The second requirement is brutal to prove. Courts regularly acknowledge that a defense attorney performed poorly but still uphold the conviction because the evidence of guilt was strong enough that better lawyering probably wouldn’t have changed the verdict.
The most common failures are basic. Not interviewing alibi witnesses. Not hiring experts to challenge questionable forensic evidence. Not investigating the backgrounds of prosecution witnesses. Not filing motions to suppress illegally obtained evidence. A thorough defense requires money, and private criminal defense attorneys typically charge several hundred dollars per hour, putting competent representation out of reach for most defendants. The alternative is a public defender, but public defense offices are chronically underfunded. Older caseload standards once recommended a maximum of 150 felony cases per attorney per year, and the national public defense community now considers even those numbers far too high given expanded digital evidence, forensic challenges, and longer sentencing frameworks. In practice, many public defenders carry caseloads that make adequate preparation physically impossible. Some clients receive only a handful of hours of their lawyer’s attention before trial.
Plea Bargaining and the Trial Penalty
Nearly 98 percent of criminal convictions nationwide come from guilty pleas rather than trials. For an innocent defendant, that statistic is terrifying. The “trial penalty” describes the sentencing gap between what a defendant is offered in a plea deal and what they face if they reject it and lose at trial. One study of murder cases found that defendants convicted by jury received sentences averaging 22.3 years, compared to 11.1 years for those who accepted plea bargains.12Florida Atlantic University. An Economic Analysis of the Trial Penalty: A Comparative Analysis of Three Alternative Trial Settings
That math pushes innocent people to plead guilty. When a prosecutor offers two years on a plea but trial carries a potential sentence of 20, even someone who committed no crime can rationally conclude that the risk of trial is too high. Exonerees have described pleading guilty to murder to avoid the death penalty, or to sexual assault to avoid a 41-year sentence. These are survival calculations made inside a system that punishes defendants for exercising the right to trial.
The Alford plea reflects this dynamic directly. It allows a defendant to formally plead guilty while maintaining innocence, accepting the punishment because the evidence is strong enough for a conviction rather than because the defendant admits to the act. An Alford plea is not available as a right; both the prosecutor and the judge must agree to it. The practical effect is that people who may well be innocent end up with guilty pleas on their records, foreclosing many later avenues for exoneration.
After a Wrongful Conviction
Reversing a wrongful conviction is harder than it sounds. Federal habeas corpus, the primary mechanism for challenging a state conviction in federal court, carries a one-year filing deadline under the Antiterrorism and Effective Death Penalty Act (AEDPA), which generally starts running when the conviction becomes final after direct appeal.13Office of the Law Revision Counsel. 28 US Code 2244 – Finality of Determination Newly discovered evidence must meet strict criteria, and changes in legal interpretation do not qualify. Federal law and the laws of at least 46 states now provide some mechanism for prisoners to request post-conviction DNA testing, though access varies.14National Institute of Justice. Defendants’ Access to Postconviction DNA Testing
Exoneration does not automatically come with financial recovery. As of early 2025, 39 states and the District of Columbia have enacted compensation statutes for the wrongfully convicted, with amounts ranging from caps as low as $25,000 total to $80,000 per year plus an annuity.15National Registry of Exonerations. Compensation for Exonerees The remaining states have no compensation statute, leaving exonerees to pursue civil litigation or private legislation with no guaranteed outcome. Even in states with relatively generous programs, the money rarely reflects what was actually lost: destroyed careers, fractured relationships, and years spent in prison for something the person did not do.