Key Findings

Key findings from DPIC’s analysis of more than 9,700 death sentences include:

Case Outcomes: The single most likely outcome of a capital case once a death sentence is imposed is that the conviction or death sentence will be overturned and the defendant will not be resentenced to death.

Fewer than 1 in 6 death sentences result in an execution.
A death sentence is 3 times more likely to be reversed as a result of a court decision than it is to result in an execution.
At least 189 people who were wrongfully convicted and sentenced to death have been exonerated since 1973. That is one person exonerated for every 8.2 people who are executed. (189 exonerations/1547 executions based on data updated through June 29, 2022.)
More than 550 capital convictions and/or death sentences have been overturned because of, or resulted in exonerations involving, prosecutorial misconduct. That is over 5.5% of all death sentences imposed in the U.S. in the past 50 years.

Geographic arbitrariness: The U.S. death penalty is geographically arbitrary and is disproportionately carried out in a small number of states and counties characterized by outlier practices and lack of meaningful judicial process.

Just 34 counties — fewer than 1.1% of all the counties in the U.S. — accounted for half of everyone on death row in U.S. states as of 1/1/2021. 2% of U.S. counties accounted for 60.8% of all state death-row prisoners. 82.8% of U.S. counties did not have anyone on death row.
Just 5 counties — Harris, TX; Dallas, TX; Bexar, TX; Tarrant, TX; and Oklahoma, OK — account for more than 1/5 of all executions in the U.S. in the past 50 years. 2% of U.S. counties accounted for half of all executions in the U.S., while 84% of U.S. counties have not had any executions in the past 50 years.
Fewer than 2.4% of all counties in the U.S. (just 75 counties) account for half of all death sentences imposed in state courts in the past 50 years.
Outlier practices disproportionately contribute to death sentences and executions. Counties in Alabama and Florida, which authorized non-unanimous death sentences, accounted for the most death sentences and had the highest per capita death-row populations in counties with smaller populations.
The quality of counsel had a direct impact on sentencing outcomes. More than 200 death sentences were imposed in Philadelphia, PA in cases in which defendants were represented by poorly trained and underfunded court-appointed counsel. No one represented by the Philadelphia public defender’s specialized homicide unit ever was sentenced to death.
Likewise, states with the highest execution rates also tended to have the worst access to meaningful judicial review. More than 100 people were executed in Texas after U.S. Supreme Court case precedent had already established the unconstitutionality of their death sentences. In 96% of capital post-conviction proceedings in Harris County, Texas, in which factual issues were contested, county judges adopted the prosecutors’ proposed findings of fact and legal conclusions word for word.

Race: While the vast majority of murders are committed against individuals of the same race as the perpetrator, the census data showed:

It is 5.9 times more likely that an execution carried out in the U.S. over the past 50 years involved one or more white victims (1207 cases) than that it involved only Black victims (206 cases). 78.1% of executions in the U.S. in the past 50 years were in cases involving white victims. 13.3% were in cases with only Black victims. 6.5% were in cases with only Latinx victims.
60% of Black defendants who have been executed were sentenced to death for killing white victims. Just 34.4% of executions of African Americans were for killing Black victims. 3.2% involved Latinx victims.
48.1% of executions of death-sentenced Latinx prisoners occurred in cases involving at least one white victim, identical to the percentage involving only Latinx victims. 2.3% involved Black victims and 1.6% involved Asian victims.

By contrast, 94.4% of executions of white prisoners were in cases involving white victims. Just 2.4% of executions of white prisoners were for killing Black victims and 2.3% were for killing Latinx victims.

The census data also showed that members of vulnerable groups who are sentenced to death (juveniles, intellectually disabled prisoners, the innocent) are disproportionately likely to be defendants of color. DPIC found that:

Defendants of color constitute 52.8% of those sentenced to death over the past 50 years. Yet they make up 64.2% of death-row exonerees and 83.1% of all death-row prisoners later found to be ineligible for the death penalty because of intellectual disability.
While Black defendants constitute 41.7% of those sentenced to death in the past 50 years, they are 54.0% of death-row prisoners who have been exonerated. 68.3% of death-row prisoners later found to be ineligible for the death penalty because of intellectual disability are Black.
Latinx defendants constitute 8.9% of those sentenced to death and are exonerated in roughly the same percentage (8.6%). However, they comprise 14.1% of death-sentenced prisoners subsequently found ineligible for the death penalty because of intellectual disability.
Defendants of color constitute 44.3% of those executed in the U.S. over the past 50 years but are 54.5% of those executed for offenses committed as juveniles, 69.8% of the intellectually disabled defendants to be executed before Atkins v. Virginia declared the practice unconstitutional, and 75.0% of the likely intellectually disabled prisoners who have been executed despite Atkins’ prohibition of the practice.
While Black defendants constitute 34.2% of those executed in the modern era, half of those executed for offenses committed as juveniles were Black. 63.0% of the intellectually disabled defendants executed before Atkins and 62% of the likely intellectually disabled defendants executed post-Atkins also have been Black.
Latinx defendants constitute 8.3% of those executed in the U.S. in the past 50 years but 14.3% of the likely intellectually disabled prisoners to have been executed post-Atkins.

The Decline of the Death Penalty: The dataset quantifies the decline of the death penalty across the United States by multiple measures.

Death sentences have declined in the United States by approximately 90% after peaking at more than 300 per year for three consecutive years in the mid 1990s, including a 70% decline over the decade preceding the pandemic. There have been fewer than 50 new death sentences imposed each year since 2015.
The decline has occurred at all levels, from the number of states and counties imposing death sentences to the number of death sentences imposed in those states and counties. More than 200 counties sentenced defendants to death in 1986, and an average of 180 counties per year imposed death sentences from 1993 – 1996. The number of counties imposing death sentences each year has declined more than 80% since then, to an average of 31.6 between 2015 and 2019.
The number of counties imposing more than one death sentences per year has declined even more. Since peaking at an average of 51.2 multiple death-sentencing counties in 1994 – 1999, it has fallen by nearly 94% to 3.2 such counties in the years 2015 – 2019. No more than 3 counties have imposed multiple death sentences since 2016.
Executions have declined in the United States by approximately 75% after peaking at 98 in 1999, to an average of 23.6 in the five years before the pandemic. There have been fewer than 30 executions in the U.S. every year since 2015.
No executions have been carried out in the Northeast in more than fifteen years and no northeastern state has carried out an involuntary execution in more than fifty years. Arizona is the only state west of Texas to have executed any person in more than a decade.
The DPIC census also confirms data from the Bureau of Justice Statistics and the Legal Defense Fund (formerly the NAACP Legal Defense and Educational Fund) that the number of people on death row or facing jeopardy of being resentenced to death on state death rows has declined to about 2,500 for the first time since 1991. BJS and LDF have documented that death row has decreased in size every year since peaking at more than 3,700 in 2001.

Explore by State