homicide rates and the legal status or the actual implementation of the death penalty.
The studies have reached widely varying, even contradictory, conclusions, and commentary on the findings has sometimes been acrimonious. Some researchers have concluded that deterrent effects are large and robust across datasets and model specifications. For example, Dezhbakhsh, Rubin, and Shepherd (2003, p. 344) concluded that:
Our results suggest that capital punishment has a strong deterrent effect; each execution results, on average, in eighteen fewer murders with a margin of error of plus or minus ten. Tests show that results are not driven by tougher sentencing laws and are robust to many alternative specifications.
Similarly, Mocan and Gittings (2003, p. 453) stated the following:
The results show that each additional execution decreases homicides by about five, and each additional commutation increases homicides by the same amount, while an additional removal from death row generates one additional murder.
In 2004 testimony before Congress, Shepherd (2004, p. 1) summarized this line of evidence on the deterrent effect of capital punishment as follows:
Recent research on the relationship between capital punishment and crime has created a strong consensus among economists that capital punishment deters crime.
However, the claims that the evidence shows a substantial deterrent effect have been vigorously challenged. Kovandzic, Vieraitis, and Boots (2009, p. 803) concluded that:
Employing well-known econometric procedures for panel data analysis, our results provide no empirical support for the argument that the existence or application of the death penalty deters prospective offenders from committing homicide … policymakers should refrain from justifying its use by claiming that it is a deterrent to homicide and should consider less costly, more effective ways of addressing crime.
Others do not go so far as to claim that there is no deterrent effect, but instead argue that the findings supporting a deterrent effect are fragile, not robust. Donohue and Wolfers (2005, p. 794) reanalyzed several of the data sets used by the authors who claimed to have found robust deterrent effects and concluded that:
We find that the existing evidence for deterrence is surprisingly fragile, and even small changes in specifications yield dramatically different results.