:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/32312985/103898508.0.jpg)
At least 4 percent of people who receive death sentences in the United States are likely innocent, a 2014 study finds.
The paper, in the Proceedings of the National Academy of Sciences, borrows a technique from biomedical research to estimate the number of prisoners sentenced to death who are falsely convicted. The study, by Samuel R. Gross of the University of Michigan and Barbara O'Brien of Michigan State University, finds that at least 4 percent of people who get sentenced to death when they're convicted would ultimately be exonerated if their cases were closely examined for the next 21 years.
That doesn't just include current death row inmates: many people who initially get death sentences end up getting their sentences reduced to life in prison. And no prisoner serving a life sentence gets the same level of scrutiny as someone on death row. For this reason, the authors conclude that the rate of false convictions in life-imprisonment cases is probably much higher.
What the study says
The reason the authors were able to estimate the number of innocent people with death sentences is that, for as long as someone's on death row, a lot of resources go into making absolutely sure he belongs there. So an innocent person on death row is much more likely to be exonerated than another innocent prisoner, even if it takes an average of ten years to do it. In fact, the authors say, it's possible that a majority of all innocent death-row prisoners are exonerated.
The problem is that most prisoners originally sentenced to death end up getting their sentences reduced to life in prison, and once that happens their cases don't get the same level of scrutiny anymore. So to estimate how many death-sentenced prisoners would be exonerated if they continued to get the higher level of scrutiny that they get while they're on death row, the authors borrowed a technique from biomedical research called "survival analysis."
In biology, the analysis gets used to estimate how many, say, Lyme disease sufferers would survive if a particular treatment were extended for a long time. For this study, instead of considering the possibility that someone would leave the Lyme-disease population by dying, the researchers looked at the possibility that someone would leave the death-row population by getting exonerated. So instead of a "survival rate" and "death rate," as medical studies show, they found a "stayed-on-death-row rate" and an "exoneration rate," respectively.
The "exoneration rate" they found was 4.1 percent. That means that if everyone who got a death sentence when they were initially convicted got 21.4 years of high scrutiny for their cases, 4.1 percent of them would be exonerated by the end of that period.
What the study means
The authors say it's fair to consider 4.1 percent a very low estimate for the number of people sentenced to death who are wrongly convicted, since they assume it's more likely that there are other innocent defendants whose innocence was never discovered than guilty defendants who were wrongly exonerated.
The authors estimate that if every false death sentence resulted in an execution, 50 innocent people would have been executed in the last 35 years. But the actual number of innocents put to death is almost certainly a lot lower, because judges and prosecutors tend to reduce an inmate's sentence to life in prison if there's any doubt about his guilt. The flip side of that, however, is that the false conviction rate of people sentenced to life in prison to begin with is probably higher than the false conviction rate in death-penalty cases — especially because jurors tell researchers that the biggest factor in deciding to give someone a life sentence rather than a death sentence is lingering doubt that the defendant is guilty at all.
The authors conclude: "The great majority of innocent defendants who are convicted of capital murder in the United States are neither executed nor exonerated. They are sentenced, or resentenced to prison for life, and then forgotten."
