No, capital punishment is not morally required: deterrence, deontology, and the death penalty.

• Author: Steiker, Carol S.
• Position: Response to article by Cass R. Sunstein and Adrian Vermeule in this issue, p. 703

INTRODUCTION I. PRIVATE MURDERS ARE NOT EQUIVALENT TO EXECUTIONS AS A MORAL MATTER: NONPURPOSEFUL VERSUS PURPOSEFUL KILLINGS II. PRIVATE MURDERS ARE NOT EQUIVALENT TO EXECUTIONS AS A MATTER OF JUSTICE: REGULATION VERSUS PUNISHMENT A. Capital Punishment as a Failure of Proportionality B. Capital Punishment as a Failure of Equality C. Capital Punishment as a Failure of Dignity III. CONCEPTUAL SLIPPERY SLOPES A. Execution of the Innocent B. Execution of Innocent Members of an Offender's Family C. Imposition of Other Extreme Punishments D. Execution of Less Culpable Offenders IV. THE FAILURE OF THE "THRESHOLD" V. WHY EVEN CONSEQUENTIALISTS SHOULD WORRY CONCLUSION INTRODUCTION

As an opponent of capital punishment, I have participated in many (and witnessed many more) debates about the morality and wisdom of the death penalty. The debate usually begins with one of two dramatic gambits by the proponent of capital punishment, both of which derive their power from the grievous harms suffered by murder victims and their loved ones. The first gambit is to consider in detail the facts of one or more capital murders and to propose that only the punishment of death is an adequate and proportional response to the terrible suffering of the victim intentionally inflicted by the perpetrator--a predominantly retributive argument. The second gambit--a modified version of which Cass Sunstein and Adrian Vermeule use to begin their provocative article (1)--is predominantly consequentialist. This gambit is to suggest that if the death penalty can prevent--through incapacitation of the offender or general deterrence--the loss to murder of even one innocent life, then it is a morally justified or perhaps even morally required penal response. A common response to both of these gambits is to ask why it is we do not rape rapists, torture torturers, or rape and then murder those who rape and murder in order to provide a proportional response to the suffering they have inflicted or to adequately deter future rapists, torturers, and rapist/murderers. This response suggests that our rejection of such extreme punishments points the way to a categorical, deontological limitation on the kinds of punishments we are justified in imposing, on either retributive or consequentialist grounds. The usual counter to this response is to acknowledge that we do not and should not impose such extreme punishments--that there is some moral limit to what we can justify as punishment--but to deny that the use of the death penalty crosses that line.

The debate--like a stylized form of dance--then tends to move from consideration of capital punishment in the abstract to its application in contemporary society. Here the opponent of the death penalty goes on the offensive, arguing that regardless of whether capital punishment is justified in the abstract, the fact that it is too often imposed arbitrarily, invidiously, or in error in our imperfect legal system renders it a morally unacceptable practice in contemporary society. The usual counter here is some combination of denying that the problems are as big as the opponent claims (citing the opponent's abolitionist bias), denying that problems of arbitrariness and discrimination affect the justice of imposing the death penalty if the defendant is guilty, and acknowledging that the erroneous conviction and execution of innocents is unjust but maintaining that the problem is either small enough to be acceptable (in light of the greater number of innocent lives saved) or fixable.

Sunstein and Vermeule want to dance to a very different tune. They start with some recent statistical studies of the impact of capital punishment on homicide rates--studies that claim to find strong deterrent effects after controlling for potentially confounding variables with multiple regression analysis. (2) Sunstein and Vermeule do not purport to vouch for the validity of this recent spate of studies, acknowledging that "it remains possible that the recent findings will be exposed as statistical artifacts or found to rest on flawed econometric methods." (3) This is a prudent concession, given the powerful reasons that are offered by John Donohue and Justin Wolfers in this Issue, (4) along with many other experts, (5) to reject this body of work as the basis for any public policy initiative. Rather, Sunstein and Vermeule argue that if such deterrent effects could ever be reliably proven or even if the evidence demonstrated a "significant possibility" that the use of capital punishment saves a substantial number of lives by preventing future murders, then consequentialists and deontologists alike should join in supporting the retention and vigorous use of the death penalty. (6) Indeed, they contend that under such conditions, capital punishment should be considered not merely morally permissible (as any consequentialist would hold) but actually "morally obligatory." (7) What Sunstein and Vermeule add to prior debates between consequentialists and deontologists regarding the death penalty is their insistence that recognition of the inapplicability of the act/omission distinction to the government as a distinctive kind of moral agent should strengthen the consequentialist argument in favor of capital punishment and undermine deontological objections to capital punishment, under the stipulated conditions of deterrence from which the argument proceeds. (8)

This argument neatly sidesteps some of the central wrangles in the typical death penalty debate described above. First, under the terms of Sunstein and Vermeule's argument, there is no need to "draw the line" excluding some extreme punishments (like torture), because the argument denies the existence of any such categorical line prohibiting extreme punishments as a moral matter; (9) the only question is whether the government can prevent more suffering inflicted by future offenders than it metes out as punishment on current offenders. Second, there is no need to address the vexing issue of how to weigh innocent lives of murder victims against (usually, but not always) guilty lives of convicted capital defendants because the argument holds that the government is equally responsible for the harms that flow from its failure to impose the death penalty and for those that flow from its imposition. Thus, all lives (innocent or guilty) are counted equally, and all that remains to do is count: if more private murders would be prevented than executions imposed, the balance favors executions. Third, the argument insists that the distributional problems of arbitrary or invidious infliction of the death penalty disappear as moral problems, at least when there is reason to believe that private murders are at least equally arbitrary or invidious in their distribution. Sunstein and Vermeule contend that the belief that there is a categorical prohibition of extreme punishments or the belief that arbitrariness, discrimination, or error in the distribution of capital punishment count as distinctive moral failures are examples of the operation of a "moral heuristic"--by which they mean a form of moral shorthand that leads to error. (10) Specifically, they refer to error arising from the failure to fully appreciate the distinctiveness of the government as a moral agent that must treat the death penalty as an example of a "life-life tradeoff." (11)

The problem with Sunstein and Vermeule's argument is not their general premise regarding the government's distinctive moral agency, which, as they acknowledge, is likely to be far more congenial to the political opponents of capital punishment than to its supporters. (12) Rather, Sunstein and Vermeule's argument runs into serious problems when they attempt to transplant their insight about government agency from the arena of civil regulation to the arena of criminal justice. Sunstein and Vermeule's assertion that the state's execution of murderers is equivalent to its failure to adequately deter murders by private actors ignores the ways in which the construction of a governmental choice as a "life-life tradeoff" in the regulatory context does not map congruently onto the criminal justice context, either as a matter of morality or as a matter of justice.

As a matter of morality, Sunstein and Vermeule fail to grapple adequately with the fact that for their argument to succeed in the criminal context, they must jettison not only the act/omission distinction in the context of government action but also--and less convincingly--the distinction between purposeful wrongdoing on the one hand and merely reckless or even knowing wrongdoing on the other. Even more problematic is Sunstein and Vermeule's failure to acknowledge the social and political fact that executions are not mere fungible "killings" but rather are part of a practice of state punishment that can be unjust in ways quite distinct from the general wrongness of killing. Sunstein and Vermeule's reduction of the deontological objections to capital punishment to some version of the moral intuition that "killing is wrong" (13) thus evades and fails even to acknowledge long-standing and widely discussed deontological objections to capital punishment qua punishment.

Moreover, despite their protestations to the contrary, (14) Sunstein and Vermeule's argument in favor of capital punishment presents some conceptual slippery slopes upon which only the deontological arguments that they evade can offer some purchase. Their argument is unable to explain why we might not, under conceivable circumstances, be morally obligated to adopt punishments far more brutal and extreme even than execution, or to inflict similarly brutal and extreme harms on innocent members of an offender's family (as punishment of the offender, not of the innocent), or to extend the use of capital punishment to contexts in which many deaths result from behavior far less culpable than murder, such as highway fatalities due to drunkenness or...