NEWS 2012

Death Penalty Studies: Three Decades Later, Still Fundamental Flaws

Most recent report points to path forward for researchers

Deterrence and the death penalty

Does the death penalty deter homicides?

Two separate reports, published 34 years apart by the National Research Council (NRC), both came to the same conclusion: Studies of the death penalty are so flawed that they cannot determine “whether the death penalty increases, decreases, or has no effect on homicide rates,” as cited in the 2012 report.

IPR economist Charles F. Manski was a member of the most recent NRC committee that took up this “incredibly hard and value-laden” question again.

Looking back, the first report was prompted by a “highly controversial” 1975 American Economic Review article, in which researcher Isaac Ehrlich concluded that each execution deters eight murders, Manski said. The article appeared at the same time that the U.S. Supreme Court was considering the constitutionality of the death penalty, ruling a year later in the 1976 Gregg v. Georgia decision that it was. This ended the four-year moratorium on executions.

Following the court ruling, the number of executions started to grow again—and following publication of the 1978 NRC report, empirical research on the subject dried up. By the mid-1990s, however, a new generation of researchers began to re-examine the topic. They asserted that their studies relied on better data and were methodologically superior to those of the past.

“Instead of just one paper, there have been dozens of studies that were getting results all over the place,” Manski said. Some of these contradictory studies found that “executions save large numbers of lives; others conclude that executions actually increase homicides; and still others conclude that executions have no effect on the homicide rate,” the 2012 report stated. They also led to a lot of “acrimonious” scientific debate about their validity.

Taking up the issue were the eight members of the NRC Committee on Deterrence and the Death Penalty. In addition to Manski, they included criminologist Daniel Nagin, the 2012 chair who had also worked on the 1978 report as a staff member; a sitting federal judge, Gerard Lynch; and the late, influential criminologist, James Q. Wilson, an originator of the “broken-windows theory.”  

The committee met intensively starting in October 2010. Avoiding questions of morality and equal justice, the members zeroed in on a simple assessment: Have these empirical studies provided “scientifically valid evidence” to determine if the death penalty affects murder rates?

While the two reports came to similar conclusions on the lack of evidence for or against a deterrent effect, Manski said, the primary difference between the two was that the 1978 report basically declared the issue “insolvable.” The final chapter of the 2012 report, however, concludes by offering a path forward for researchers.

Charles manski
Charles F. Manski

The first point is that the failings of studies on the topic stem in part from performing statistical analyses of the death penalty and murder rates without inquiring about how other possible noncapital deterrents, such as a sentence of life without parole, might affect murder rates, Manski said. To help remedy this, a national database is needed that would detail alternatives to capital punishment state by state.

The second, he continued, concerns our lack of knowledge of how a criminal might form expectations of committing a crime against the risk of getting caught, or, “What is the chance that I will be given the death penalty if I murder someone?”

The final recommendation deals with another research area “near and dear” to Manski—partial identification analysis, an area that he has done much to develop since the 1990s.

To date, most death penalty studies have combined available data with strong assumptions in order to reach strong conclusions, he said. Like the report, Manski urges researchers to “face up” to the likelihood that getting such definite results requires “assumptions so strong as not to be believable.” Although using weaker assumptions would lead to weaker results, it would also generate more believable ones, he said.

“Despite the lack of empirical evidence, society still has to make decisions about what is the right thing to do,” Manski said.  Therefore, judgment about whether there is a deterrent effect is still relevant to policymakers, as are other moral and legal arguments, the current report states, but this “judgment should not be justified based on evidence from existing research on capital punishment’s effect on homicide.”

Conclusion and Recommendation:

The committee concludes that research to date on the effect of capital punishment on homicide is not informative about whether capital punishment decreases, increases, or has no effect on homicide rates. Therefore, the committee recommends that these studies not be used to inform deliberations requiring judgments about the effect of the death penalty on homicide. Consequently, claims that research demonstrates that capital punishment decreases or increases the homicide rate by a specified amount or has no effect on the homicide rate should not influence policy judgments about capital punishment.

The summary and complete report are available here.

Read Manski’s related working paper with John Pepper on “Deterrence and the Death Penalty: Partial Identification Analysis Using Repeated Cross Sections.

Charles F. Manski is the Board of Trustees Professor in Economics at Northwestern and an IPR fellow.

National Research Council. Deterrence and the Death Penalty. Washington, DC: The National Academies Press, 2012.