The Theoretical Case Against Deterrence

The concept of deterrence is built upon the assumption that criminals are rational actors; if we regarded criminals as irrational, we might immediately give up on the possibility of influencing their decisions. But how rational are murderers, the particular subset of criminals that society is most interested in affecting? The answer is not so clear.

For almost the entirety of the modern death penalty era, for example, states have executed the mentally retarded; the Supreme Court ruled unconstitutional the execution of persons with an IQ of 70 or below only in 2005. The mentally ill, however, are still not exempt. This is not to say that either the mentally retarded or mentally ill do not know the difference between right and wrong, but it should be clear that the prospects for deterring these persons are not particularly good. Nor should this be read as suggesting that the majority of the death row population is mentally retarded or mentally ill, though many people would likely suggest that there is something quite clearly different about those who commit murder and those who do not.

While there is undoubtedly a difference of mindset between the inmate who murders his girlfriend’s husband and most other people, it seems obvious that even he is procedurally rational. It is not simply the case that most people are more risk-averse than the murderer, it is that murder is part of a menu of choices for one and not for most others. There is considerable debate, to be sure, about how much weight we should ascribe to biology and how much to circumstance,[1] but the fact remains that I am not in the least deterred from committing murder by the possible consequences of doing so; I do not even begin to consider the crime, which means that the penalty is of no personal concern. We need to focus, then, on the question of deterring those for whom murder is already on the menu of available choices.

The inmates on death rows across the country are a fairly uniform bunch. They are poor, they are male, and roughly half of them reside in only four states, California, Florida, Texas, and Pennsylvania.[2] They are often addicted to drugs or alcohol; they are the losers in our society, on the bottom rung of the ladder. Many of them –- and, indeed, convicted felons in general -– view the world and their place in it quite differently from the way those who hold to the idea of deterrence would expect. As Howard Zehr (54) argues, “If success comes, it is associated more with luck than hard work. If they are arrested for an offense, it has more to do with luck than something they did. Whether or not they do have the power to make real choices, many do not believe that they do.”

The implications of this sort of thinking are serious, for deterrence necessitates a belief that we choose what we will do in any given situation. Zehr (54) continues: “Every day these young people saw innocent people being arrested. Every day they saw people who were guilty going free. In their mind, there was little relationship between offense and punishment. Rather, they saw punishment like rain. Some days it rains, some days it doesn’t.” If offenders believe they lack agency, then the threat of punishment will not influence their actions because the link between crime and punishment is broken, or at least attenuated, in their minds.

It might be the case, then, that the particular group of people the death penalty is meant to deter simply will not be deterred. But perhaps there is another group, this time of potential offenders, who feel that they have choices to make and that there are consequences for actions stemming from these choices. Perhaps there are a lot of people in this group who are deterred every day by the specter of the death penalty because it is clear to them that the choice leading them to murder will lead also to the execution chamber. That is a fairly straightforward story, but one that is most likely incorrect.

The typical rational agent is not deterred by the death penalty because he has every reason to believe that he will not be executed, even if he commits murder. The death penalty is not, of course, the sole – or even the most frequent – punishment for homicide in the United States. Richard Dieter estimates that there are approximately 15,000 homicides annually and approximately 150 annual capital sentences. One percent of murderers, then, are sentenced to death. These numbers fluctuate a bit, certainly, but the greatest number of capital sentences in any year since 1991 was 317 (in both 1995 and 1996). If we assume that there are fewer homicides and more capital sentences -– say 10,000 and 500, respectively -– then we arrive at the conclusion that five percent of all homicides end in a death sentence. A rational agent who is considering homicide would conclude that he is very unlikely to be convicted and sentenced to death in this country.

As Isaac Ehrlich (415) admits, “The analysis rests on the presumption that offenders respond to incentives.” In weighing the costs and benefits of homicide, then, potential murderers must incorporate the death penalty into their calculation and this higher punishment thus deters at least some of them.[3] This model depends on several conditions. Criminals or potential criminals must have some understanding of the probability they will be apprehended and convicted, along with an idea of the likely punishment (Anderson, 6). The rarity of executions and long lags between crime, sentencing, and executions make these probabilities fairly small. Steven Levitt (175) concludes that “a rational criminal should not be deterred by the threat of execution” precisely because executions are so rare and time horizons are short. Exacerbating this concern is the tendency of people to not identify with people on death row or with people in bad situations generally (Bowers and Pierce, 455). An attitude of “It couldn’t happen to me” would also limit the effectiveness of a deterrent signal.

Since it seems unlikely that we’ll resolve the empirical debate about the ability of the death penalty to deter murderers any time soon, perhaps we can at least start to consider whether or not deterrence makes sense theoretically. While these are only a few of the possible theoretical arguments regarding the deterrent effect of the death penalty, I think they will likely provide us with much to discuss.