Why not all executions deter murder

At approximately 12:00 a.m. Pacific Time on Dec. 13, Stanley "Tookie" Williams was executed in California. His execution was California's twelfth since 1977. Although California has the largest death row in the United States, with approximately 650 inmates, it has one of the lowest ratios of inmates executed to inmates on death row. California sentences many to death, but it executes few.

There are many moral and legal issues in the debate over capital punishment. One central issue is whether executions deter murder. Deterrence is the basis that many policymakers and courts cite for capital punishment. For example, in one of his presidential debates with Al Gore in 2000, President Bush stated that capital punishment deters crime and that deterrence is the only valid reason for capital punishment. Likewise, the Supreme Court, when it held in its landmark 1976 decision Gregg v. Georgia that capital punishment was constitutional, cited deterrence as one of its main reasons.

Theories are conflicting as to whether executions will reduce or increase murders. Executions may deter some criminals from committing murder because they raise the potential costs of committing a murder. Or, executions may lead to a brutalization effect, creating a climate of brutal violence and setting an example of killing to avenge grievances.

Recent empirical evidence initially seemed to confirm the deterrence theory. In the past decade, 12 empirical studies by economists, published in peer-reviewed journals, have found evidence consistent with a strong deterrent effect. Most of these studies, including three by me, use large data sets that combine information from all 50 states or all US counties over many years to show that, on average, an additional execution deters many murders.

Although the previous studies estimated a national average deterrent effect, it is reasonable to assume that the deterrent effect will vary based on the great differences in states' application of the death penalty. For example, states vary widely in: their definitions of capital crimes, their frequency of imposing capital sentences, their frequency of executions, their methods of execution, and the publicity those executions receive.

In a study recently published by the Michigan Law Review, I use three well-known data sets common to empirical studies of crime and well-tested empirical methods. I find that the impact of executions differs among states with the death penalty. Although executions appear to deter crime in approximately one-fifth of these states, in the remaining 80 percent, executions show no deterrent effect. Indeed, in some of these states, executions produce the opposite effect: Murders increase after executions.

Why does this happen? One important factor is that, on average and with exceptions, the states where capital punishment deters murder tend to execute many more people than do the states where capital punishment incites crime or has no effect.

An intuitive explanation for this is that each execution creates two opposing reactions: a brutalization effect and a deterrent effect. For a state's first few executions, the deterrent effect is small. Only if a state executes many people does deterrence grow; only then do potential criminals become convinced that the state is serious about the punishment, so that they start to reduce their criminal activity. For most states, when the number of executions exceeds some threshold level, the deterrent effect begins to outweigh the brutalization effect. In the four-fifths of states where executions either increased murders or had no effect, the brutalization effect either counterbalances or outweighs the deterrent effect.

What about the earlier studies showing deterrence as a national average? These results appear to be caused by the few states with many executions and deterrence outweighing the many states with no deterrence or increased murders.

Will Stanley Williams's execution increase or decrease murders? If Williams had been executed in Texas, his execution would be expected to deter future murders. My study shows a strong deterrent effect there, seemingly motivated by the hundreds of people that Texas has executed in recent decades.

California is different. The three data sets that my recent paper analyzes vary by time period and other characteristics. Therefore, there are some differences among the data sets in the states that fall into each group: deterrent, no effect, and brutalization. Nevertheless, none of the data sets in California show that executions deter murder.

Perhaps this is explained by Williams being only the twelfth execution in California in the past quarter century, compared with the 650 on death row there. Perhaps this low ratio of executions to death-row inmates is not enough to convince potential criminals that the possibility of execution is a real threat that could be imposed on them.

However, it is possible that future executions in California may have a deterrent effect. Ten of California's 12 executions have taken place within the past decade. Perhaps the increased frequency will shift California to being a deterrence state. If so, the execution of Williams may save the lives of other potential murder victims. Only time, and future empirical analysis, will tell.

• Joanna Shepherd is an assistant professor of law at Emory Law School.