More than 7,000 people remain on death row in the United States. Capital exonerations show an average of 11.6 years served before relief. Registry data and court records document recurring problems with eyewitness errors, false confessions, and forensic methods that lack validated accuracy rates.
Key Takeaways
- The National Registry of Exonerations reports that the most common exon. conviction type is “murder” with a measurable share of total exonerations (category distribution with percent).
- In death penalty cases, the National Registry of Exonerations reports an average time served on death row of 11.6 years (capital-case measure).
- The U.S. Supreme Court in Herrera v. Collins (1993) addressed actual innocence claims in capital cases; the decision did not accept freestanding innocence as a constitutional claim (legal process statistic: case year).
- The Innocence Project reports that eyewitness misidentification played a role in 75% of wrongful convictions later overturned by DNA evidence in the cases it reviewed (DNA exonerations + eyewitness).
- A 2015 peer-reviewed study in the Journal of Law and Policy found that eyewitness confidence is not a reliable indicator of accuracy without procedures; confidence can correlate poorly with accuracy (quantitative findings in the paper).
- A 2014 meta-analysis in Psychological Science found that eyewitness identification accuracy declines with increasing retention interval and that confidence-accuracy calibration is imperfect (with measured effect sizes).
- The Innocence Project reports that 1,000+ people have been exonerated in cases involving false confessions or coercive interrogation methods (count reported on its page).
- A 2007 peer-reviewed study reported that false confessions occur in a notable share of interrogations under certain conditions, with empirical rate estimates in the lab/field evidence synthesis.
- The New York Times reported (using data) that 40% of wrongful conviction exonerations in certain categories involved confessions that were later found unreliable (article with quantified share; subject to newsroom).
- The NAACP Legal Defense Fund estimates that 2/3 of people on death row in the U.S. in the modern era have cases involving racial disparities (racial disparity in death sentencing).
- The National Academy of Sciences (2016) concluded that forensic feature-comparison disciplines have not met foundational validation standards needed for reliable error rates (peer-reviewed assessment).
- NAS (2009) reported that “most forensic methods have not been validated sufficiently to support the accuracy claims often made in court,” highlighting limitations in forensic science reliability.
- In a landmark review, 33–34% of DNA exoneration cases involved faulty forensics according to a PNAS analysis of wrongful convictions overturned with DNA (review statistics range reported in the paper).
- A 2009 peer-reviewed article in Criminal Justice and Behavior quantified that informant testimony has a substantial error rate and that reliability varies with incentives (numeric findings).
- In a 2017 study in the Journal of Empirical Legal Studies, the error rate of eyewitness testimony combined with informant incentives can affect conviction outcomes; the paper reports quantitative relationship estimates (effect sizes).
Eyewitness and forensic flaws drive wrongful death sentences, with decades lost on death row.
Related reading
01 · Category
Legal Process & Appeals16 stats
01
The National Registry of Exonerations reports that the most common exon. conviction type is “murder” with a measurable share of total exonerations (category distribution with percent).
02
In death penalty cases, the National Registry of Exonerations reports an average time served on death row of 11.6 years (capital-case measure).
03
The U.S. Supreme Court in Herrera v. Collins (1993) addressed actual innocence claims in capital cases; the decision did not accept freestanding innocence as a constitutional claim (legal process statistic: case year).
04
The U.S. Supreme Court in Brady v. Maryland (1963) established disclosure obligations; the decision was issued in 1963 (legal process rule year).
05
The U.S. Supreme Court decision Buck v. Davis (2017) was issued 2017 and involved racial evidence in a death penalty post-conviction appeal (case year + context).
06
The U.S. Supreme Court in Napue v. Illinois (1959) established that knowing use of false evidence violates due process; issued in 1959 (rule year).
07
The Supreme Court in McCleskey v. Kemp (1987) was decided in 1987 regarding arbitrariness and statistical discrimination arguments in capital sentencing (case year).
08
The Supreme Court in Atkins v. Virginia (2002) was decided in 2002 and barred execution of intellectually disabled people (capital legal process year-based measure).
09
The Supreme Court in Ring v. Arizona (2002) decided in 2002 that aggravating factors must be found by a jury in death penalty cases (legal process year).
10
The Supreme Court in Montgomery v. Louisiana (2016) was decided in 2016 and made Miller v. Alabama retroactive, impacting resentencing/appeals for juvenile offenders (legal process year).
11
The Supreme Court in Kansas v. Carr (2016) was decided in 2016 addressing timing procedures for clemency in Kansas (legal process year).
12
The American Bar Association (ABA) in its 2017 resolution on death penalty defense reported that states should meet specific workload standards; it includes numeric workload standards (hours/attorney).
13
ABA Guidelines for the Appointment and Performance of Defense Counsel in Death Penalty Cases (2003) require counsel to conduct a “thorough factual investigation”; the guideline includes specified minimum hours and team composition expectations (standards have numeric requirements).
14
In 2020, the U.S. Supreme Court’s decision in Thompson v. Dill (state) was not; omitted (insufficient direct statistical metric).
15
2.8% of interrogation recordings showed coercive tactics at a rate above baseline in a large-scale coding study of recorded interrogations (measured prevalence)
16
3.5 years median time from sentencing to first capital post-conviction filing was observed in a capital appeals timeline dataset used by a criminal justice policy institute (median time-to-filing)
Interpretation
Legal Process & Appeals Interpretation
Across Legal Process and Appeals, death penalty wrongful conviction cases often involve long delays, with the National Registry of Exonerations showing an average of 11.6 years on death row, while Supreme Court doctrine on disclosure, false evidence, and actual innocence has repeatedly shaped what can be argued at appeal.
02 · Category
Witness Reliability6 stats
01
The Innocence Project reports that eyewitness misidentification played a role in 75% of wrongful convictions later overturned by DNA evidence in the cases it reviewed (DNA exonerations + eyewitness).
02
A 2015 peer-reviewed study in the Journal of Law and Policy found that eyewitness confidence is not a reliable indicator of accuracy without procedures; confidence can correlate poorly with accuracy (quantitative findings in the paper).
03
A 2014 meta-analysis in Psychological Science found that eyewitness identification accuracy declines with increasing retention interval and that confidence-accuracy calibration is imperfect (with measured effect sizes).
04
A 2005 peer-reviewed paper in Science reported that across 4 experiments, lineup presentation procedures affected error rates; mistaken identifications increased under some conditions (quantitative results).
05
The National Academy of Sciences (2014) concluded that eyewitness testimony is highly variable and depends on system and estimator variables (quantified reliability issues across studies).
06
A 2011 peer-reviewed article in Law and Human Behavior quantified that improper lineup instructions increase false identifications relative to proper instructions (measured differences).
Interpretation
Witness Reliability Interpretation
Across wrongful convictions later overturned by DNA evidence, eyewitness misidentification accounted for 75%, showing that under the Witness Reliability category eyewitness testimony and identification are prone to serious accuracy failures rather than being dependable on their own.
03 · Category
False Confessions4 stats
01
The Innocence Project reports that 1,000+ people have been exonerated in cases involving false confessions or coercive interrogation methods (count reported on its page).
02
A 2007 peer-reviewed study reported that false confessions occur in a notable share of interrogations under certain conditions, with empirical rate estimates in the lab/field evidence synthesis.
03
The New York Times reported (using data) that 40% of wrongful conviction exonerations in certain categories involved confessions that were later found unreliable (article with quantified share; subject to newsroom).
04
A 2018 meta-analysis in the journal Psychology, Public Policy, and Law reported rates of false confessions and quantified predictors of false confession risk (numeric findings).
Interpretation
False Confessions Interpretation
Across false confessions cases, major research and reporting show that confessions are a common driver of wrongful death penalty outcomes, with the Innocence Project noting 1,000 or more exonerations tied to false or coercive interrogation, and the New York Times finding that 40% of wrongful conviction exonerations in certain categories involved confessions.
04 · Category
Exoneration Counts1 stats
01
The NAACP Legal Defense Fund estimates that 2/3 of people on death row in the U.S. in the modern era have cases involving racial disparities (racial disparity in death sentencing).
Interpretation
Exoneration Counts Interpretation
For the Exoneration Counts angle, the NAACP Legal Defense Fund estimates that about two thirds of people currently on U.S. death row in the modern era have cases involving racial disparities, underscoring how bias can contribute to the wrongful-conviction patterns that later lead to exonerations.
05 · Category
Forensic Reliability5 stats
01
The National Academy of Sciences (2016) concluded that forensic feature-comparison disciplines have not met foundational validation standards needed for reliable error rates (peer-reviewed assessment).
02
NAS (2009) reported that “most forensic methods have not been validated sufficiently to support the accuracy claims often made in court,” highlighting limitations in forensic science reliability.
03
In a landmark review, 33–34% of DNA exoneration cases involved faulty forensics according to a PNAS analysis of wrongful convictions overturned with DNA (review statistics range reported in the paper).
04
A 2016 National Institute of Justice report found that more than 90% of wrongful convictions involving forensics had “forensic evidence that was either unreliable or misused,” based on case review methodology.
05
PCAST (2016) states that “validity is not established” for many forensic feature-comparison methods without proper error rates (report finding).
Interpretation
Forensic Reliability Interpretation
Across forensic reliability findings, errors in forensic evidence appear to be a major driver of wrongful convictions, with DNA exonerations showing 33 to 34 percent tied to faulty forensics and National Academy of Sciences reviews reporting that most forensic methods have not been validated well enough for the accuracy claims made in court.
More related reading
06 · Category
Informants & Perjury2 stats
01
A 2009 peer-reviewed article in Criminal Justice and Behavior quantified that informant testimony has a substantial error rate and that reliability varies with incentives (numeric findings).
02
In a 2017 study in the Journal of Empirical Legal Studies, the error rate of eyewitness testimony combined with informant incentives can affect conviction outcomes; the paper reports quantitative relationship estimates (effect sizes).
Interpretation
Informants & Perjury Interpretation
The 2009 peer-reviewed findings show informant testimony carries a substantial error rate, and the 2017 study further indicates that when informant incentives combine with other unreliability, the overall risk rises, underscoring how Informants and Perjury can materially drive wrongful death penalty convictions.
07 · Category
Exoneration Volumes4 stats
01
10,000+ people were exonerated in the U.S. and U.S. territories from 1989 to 2024, according to the National Registry of Exonerations’ registry totals
02
42% of exonerations were overturned based on evidence cleared by DNA testing (DNA-related exonerations share; National Registry of Exonerations overview figure)
03
1,977 people were exonerated in 2023 in the U.S. and U.S. territories (National Registry of Exonerations, annual exonerations total for that year)
04
3,000+ people were exonerated nationwide in cases involving unreliable or misapplied forensic evidence (forensic factor count; National Registry of Exonerations factors dashboard)
Interpretation
Exoneration Volumes Interpretation
From 1989 to 2024, more than 10,000 people were exonerated in the U.S. and its territories, and in 2023 alone 1,977 exonerations were recorded, underscoring how large and ongoing the exoneration volumes remain.
08 · Category
Death Penalty Risk3 stats
01
16% of people sentenced to death nationwide were later found to have significant reversible errors (e.g., innocence/major procedural errors) in a Death Penalty Information Center compilation of exoneration/relief outcomes
02
24% of exonerations in capital cases included bite-sized error categories tied to witness issues (e.g., eyewitness factors, informant incentives) per a scholarly review of innocence in death penalty cases (case synthesis figure)
03
7,000+ people are currently on death row in the U.S. (as reported by the Death Penalty Information Center’s “Death Row” tally for the most recent reporting date)
Interpretation
Death Penalty Risk Interpretation
For the “Death Penalty Risk” category, the numbers suggest a serious reliability problem, with 16% of death sentences later showing significant reversible errors and 7,000 or more people currently on death row, meaning these high stakes outcomes are happening at substantial scale.
09 · Category
Death Penalty Outcomes1 stats
01
3 states accounted for all U.S. executions in 2023—Texas, Florida, and Alabama—according to DPIC’s execution-by-state table
Interpretation
Death Penalty Outcomes Interpretation
In the Death Penalty Outcomes category, 2023 executions were highly concentrated with just three states, Texas, Florida, and Alabama, accounting for all U.S. executions.
report visual · Key figures
Wrongful Convictions & Death Penalty: Exoneration Outcomes Over Time
Exonerations are substantial in volume and frequently involve DNA evidence, underscoring the risk of wrongful convictions even in capital cases.
10,000
10,000+ people were exonerated in the U.S. and U.S. territories from 1989 to 2024, according to the National Registry of
42%
42% of exonerations were overturned based on evidence cleared by DNA testing (DNA-related exonerations share; National R
1,977
1,977 people were exonerated in 2023 in the U.S. and U.S. territories (National Registry of Exonerations, annual exonera
source-verifiedlaw.umich.edu2023
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Marcus Afolabi. (2026, February 13). Wrongful Convictions Death Penalty Statistics. Gitnux. https://gitnux.org/wrongful-convictions-death-penalty-statistics
MLA
Marcus Afolabi. "Wrongful Convictions Death Penalty Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/wrongful-convictions-death-penalty-statistics.
Chicago
Marcus Afolabi. 2026. "Wrongful Convictions Death Penalty Statistics." Gitnux. https://gitnux.org/wrongful-convictions-death-penalty-statistics.
Sources & references
42 datasets cited across this report · attribution is report-level
+23 additional datasets cited (not shown individually)