Virginia may be about to eliminate the death penalty, which would remove the state with the greatest historical use of the punishment from the execution business. Simultaneously, President Biden is the first sitting president to openly oppose capital punishment. Dozens of lawmakers have urged him to commute the death sentences of all those on federal death row.

Our recent research offers some useful insights for those considering whether Virginia or the federal government will retain their death penalty systems. Public opinion has shifted dramatically since the 1990s. Since 1973, 174 innocent men have been released from death row, which has shaken public confidence in the system. Today, few jurisdictions use the penalty. The counties most likely to condemn someone to death also had higher per capita numbers of lynchings in the Jim Crow era and have a higher Black population today than those counties that have made little or no use of the death penalty.

How we did our research

We spent years compiling a database of every death sentence in the nation since 1972, totaling more than 8,500 cases. Most were concentrated in just a few high-use counties including those surrounding the cities of Los Angeles, Houston, Philadelphia, Phoenix, Chicago and Miami. We developed a statistical model to predict how many death sentences each county would impose, for each year from 1976 through 2019. Naturally, we eliminated counties in states that had no death penalty. We looked at population size, homicides per capita and several other factors to see what might influence death penalty use.

The most influential factor, we found, is population size. Large population counties have more death sentences than small ones, a simple result of the number of homicides.

Some counties “learn to kill” while others never do

That’s far from the full story, however. Here’s a second important factor in boosting death penalty imposition: the cumulative number of death sentences a county had previously imposed. A district attorney’s office may or may not accumulate the skills, knowledge and practice needed to successfully carry out a capital trial leading to a death sentence. Most do not; a few become specialists.

We and others have previously explained how a small number of counties generate the majority of the nation’s executions. The reason is that they “learn to kill.” Some counties go down the path of death sentencing and executing, and others never do. This self-reinforcing process helps explain why two counties in the same state, each looking at its own history, might reach different decisions about whether a given crime is among the “worst of the worst,” deserving death. Such a system is akin to random, since small initial differences can amplify into large disparities. Inertia is a powerful force as prosecutors consider their own offices’ histories, not those of surrounding areas.

Public opinion and history of lynching are strong predictors of death penalty use

Public opinion is another important influence in deciding whether to seek execution. Over the past 25 years, popular attitudes have moved sharply against capital punishment, largely because of what we’ve called the “discovery of innocence.” As of last month, we have counted 174 death row exonerations since 1973. Innocence has transformed the debate to the point that some conservatives have moved against the penalty, partly because the system is costly and partly because they don’t trust the government to be perfect.

The death penalty is also highly racialized. Our study confirms that whether a county has a significant Black population is strongly related to whether it uses the death penalty: In counties with higher proportions of Black residents, courts are more likely to impose the death penalty. Southern counties use the death penalty more frequently. In addition to having larger Black populations, Southern counties had more lynchings during the Jim Crow period of 1883 to 1930; counties that lynched more at that time condemn more to death today.

Legal historians would not be surprised that historical lynchings have had a lasting effect. Capital punishment has long been linked to lynchings. Promising to bring the perpetrator to “justice” within the legal system helped suppress mob violence outside the legal system in the early 20th century. The figure below shows the link between legal executions, which rose sharply from 1919 to 1935, and lynchings, which declined in the same period.

Executions substituted for lynchings in the 1930s. Black people are arrested, prosecuted, convicted and imprisoned at higher rates than White people throughout the criminal justice system — and are sentenced to death at higher rates as well; 43 percent of those sentenced to death since 1972 are Black.

What did not influence death penalty rates?

We also checked to see whether a county’s death penalty rate went up or down based on whether that state had partisan judicial elections (as opposed to judges appointed by the governor or elected in a nonpartisan manner); whether the governor was a Republican or Democrat in that year; and the number of homicides in the county in the previous year. None mattered. Counties with higher homicide rates do not sentence more people to death, other things being equal. Once population size is accounted for, homicides don’t matter.

If homicide rates don’t affect death penalty rates, what matters is whether prosecutors are in the habit of seeking death sentences, a path that was much more common in counties with ugly histories of racial conflict.

The death penalty is on the decline

In 2006, 38 states allowed imposition of the death penalty and 23 imposed at least one death sentence; by 2020, just 28 states allowed the penalty, and only seven imposed it. Even in the dwindling number of states that have a legally valid death penalty statute, few death sentences are being imposed and even fewer carried out. The Trump administration’s pursuit of executions, with 13 in the six months before President Donald Trump left office, contrasts starkly with these trends, of course. But given Biden’s opposition to the practice, and Virginia’s move toward abolishing the death penalty, Trump’s actions may become a historic blip rather than the reversal of a trend.

Frank R. Baumgartner is professor of political science at UNC-Chapel Hill and author of “Deadly Justice: A Statistical Portrait of the Death Penalty” (Oxford University Press, 2018) and “The Decline of the Death Penalty and the Discovery of Innocence” (Cambridge University Press, 2008).

Christian Caron is a PhD candidate in political science at UNC-Chapel Hill. His research on the death penalty has appeared in American Politics Research and PLOS One.