A Reply to The Atlantic on the Death Penalty

by John R. Lott Jr.

Death-penalty opponents are engaging in a lot of hyperbole.

As an example, in the Atlantic this past Thursday, Andrew Cohen went after what I have written on the death penalty at National Review Online and in my book Freedomnomics, pushing the claim that “no reliable study by credible researchers has ever found any deterrent effect” from the death penalty. He also gets into the issue of race and quotes John H. Blume, a professor at Cornell Law School and death-penalty opponent, as asserting: “Every credible study has found a statistically significant race of victim effect” on who gets sentenced with the death penalty. 

As for the first claim, Cohen relies on a deceased economist who died before he could evaluate my claim and the April 2012 report from the National Research Council (NRC). But even that report, which was edited by two strong death-penalty opponents, contradicts the assertion that “no reliable study by credible researchers has ever found any deterrent effect.” Instead, the report concludes that there are approximately equal numbers of papers showing deterrence as showing no clear effect.

Unfortunately, the NRC report itself is rather biased as it excludes more than half the academic research done. It counts only ten studies (nine were peer-reviewed) that look at all 50 states and what happens when states adopt the death penalty. By my count, 20 peer-reviewed studies and four non-peer-reviewed ones were of the type that the NRC considered — following the 50 states over time to see how murder rates change when states use the death penalty. Without offering any explanation, the NRC just ignored the bulk of the research that showed the death penalty deterred murders. 

Politicians, such as those in the Obama administration, simply can’t keep politics out of the National Research Council studies, and they bias reports through whom the government puts on these panels. The Obama administration knows the views of the people they put on the panels.

If indeed they had surveyed all these peer-reviewed studies, they would have found that by more than 2-to-1, these studies found that the death penalty deters murders.

Cohen phrases the debate this way: “On deterrence, he told me, we have Lott, who argues that ‘innocent people’s lives are saved thanks to the death penalty,’ and then we have an April 2012 report from the National Academy of Science’s National Research Council.” But that is a strange way of casting the debate. In truth, it isn’t me versus the NRC report. It is all the various academic papers and researchers the NRC report ignored.

Of course, just because a study is peer-reviewed doesn’t mean that there aren’t problems with it. Not surprisingly, the NRC report couldn’t seem to find any problems with the studies that fail to find a deterrence effect from the death penalty. 

So why do some studies find an effect while others do not? As I explain over on my blog at the Crime Prevention Research Center website:

The few studies that fail to find any deterrence from the death penalty have done some odd things. For instance, they measure the execution rate in strange ways. Take an approach first used by Lawrence Katz, Steven Levitt, and Ellen Shustorovich and later by John Donohue and Justin Wolfers. They do not look at the percent of murders that result in execution, but instead at the number of executions per prisoner

That simply makes no sense. If more criminals are put in prison for crimes such as petty larceny or auto theft, would that lower the risk that murderers face from execution? Of course not. Comparing two unrelated statistics, it is hardly surprising that this research cannot identify any benefit from the death penalty.

Cohen would have known all this if he had simply looked at Chapter 4 of my book, Freedomnomics, before attacking me.

So what about Cohen’s assertion that “every credible study has found a statistically significant race of victim effect” on who gets the death penalty? Notice that Cohen and the people he quotes never directly address my point that “the race results disappear as soon as you account for even simple variables such as income or occupation.” 

The final issue is whether the innocent are accidentally convicted. There is no DNA evidence proving that the wrong person has ever been executed. The rate that innocent people are even convicted of murder, let alone sentenced to death, is just a tiny fraction of one percent. Cohen’s hyperbole in defending law professor Sam Gross’s claim that 4 percent of death-row inmates are wrongly convicted confuses convictions that are overturned with convictions that were mistakes on the merits. The paper effortlessly slides from using terms like “false convictions” to “exoneration” (e.g., top left column of page 2), but while the rate of overturned cases for any reason is indeed higher in death-penalty cases simply because so much effort is put into appeals, neither of these terms implies the defendant was innocent.

Assertions about “credibility” and appeals to authority won’t magically make the bulk of the academic research go away.

— John R. Lott Jr. is the president of the Crime Prevention Research Center and the author of More Guns, Less Crime.

The Corner

The one and only.