Retributivism: Essays on Theory and Policy

Retributivism: Essays on Theory and Policy

Editor: Mark D. White
Oxford University Press, 2011
Reviewed by Adil Ahmad Haque  |  September 2011

The philosophy of punishment has fallen on hard times. In a recent volume of essays by leading criminal law theorists, boldly titled Philosophical Foundations of Criminal Law, only one of the twenty-two essays discusses the justification of punishment. Worse still, in that sole essay Mitchell Berman argues that theorists have converged on a justification of punishment that Berman calls ‘desert-constrained consequentialism’ (Berman 2011). The essay is of the highest quality, but makes one feel that the End of History has come to criminal law theory, that the grand narratives of retributivism and consequentialism have run their course, and that a master synthesis has been achieved. Yet it appears that a few Cold Warriors have not accepted the new reality and are still fighting the last war. Or perhaps the last war is not over after all. 

Retributivism: Essays on Theory and Policy brings together twelve leading and emerging thinkers for a series of reflections on the nature, foundation, and application of retributive justice. The first eight essays are conceptual and philosophical in nature. Antony Duff and Dan Markel cast the criminal law as a system for holding fellow citizens morally accountable for flouting the values of the polity, and defend criminal punishment as an appropriate condemnatory response. Michael Cahill and Mark White attempt to balance the pursuit of retributive justice against the varied and competing imperatives of any criminal justice system. Sarah Holtman and Jane Johnson examine the role of retribution in the thought of Kant and Hegel, respectively, while Gerald Gaus inquires into the evolutionary origins and survival value of our retributive impulses. Jeffrie Murphy reflects on his long struggle to ground retributive justice in other values, such as fairness and moral emotion, and shares his current doubts about this justificatory project. Each of these essays is worth reading, though Duff’s stands out for an important development of his widely-followed and universally-admired body of work. Duff argues that while conviction alone might communicate the moral condemnation of the polity, punishment aims to ensure that the message gets through: punishment focuses the minds of offenders on the reason for their punishment and thereby enables genuine remorse, apology, and reparation. Unfortunately, Duff introduces his new proposal in the last three pages of his essay and does not answer the many questions the proposal will raise in the minds of his readers. Nor shall I presume to answer these questions on his behalf. 

I will focus, instead, on the book’s final four essays, grouped together under the heading of “Retributivism and Policy,” which I found the most provocative. Mark Tunick’s essay, “Entrapment and Retributive Theory”, attempts to show how an entrapped defendant might lack moral blameworthiness and therefore not deserve punishment. Tunick first argues that an entrapped defendant will generally neither cause nor risk harm to others and therefore commits no punishable wrong. Of course, it not generally a defense to a criminal charge that the defendant’s action, though it satisfies the elements of an offense, neither caused nor risked harm to others. Indeed, there is at least one area in which such a defense is explicitly rejected:  it is not a defense to a charge of criminal attempt that it was factually impossible for the attempt to have succeeded. Moreover, it is simply untrue that an act must cause or risk harm to others in order to constitute a punishable wrong.  If an act-type is morally wrong and if the criminalization of that act-type will prevent harm to others then that act-type may be justly criminalized and punished even if not every act-token causes or risks harm to others (Gardner & Shute 2007).

Tunick next argues that “a person enticed by the police may not deserve punishment  . . . insofar as . . . it is not they but the police who cause the crime” (198). Now, it is certainly true that a defendant deserves no punishment if the police cause her to involuntarily commit a crime (for example by intoxicating the defendant without her knowledge or consent). But entrapment does not generally render a defendant’s actions involuntary. Tunick argues that an enticement causes a crime if either (a) the defendant’s disposition to commit the crime was weak and unlikely to be triggered absent police intervention or (b) the enticement takes inappropriate advantage of a defendant’s weaknesses.  These circumstances may establish the causal responsibility of the police, but they do not negate the causal responsibility of the defendant.

Finally, Tunick argues that police enticements are coercive because they conceal the true expected value of the options presented, in particular the probability that the crime will be detected and punished. Tunick’s argument seems to conflate coercion and deception, force and fraud. It also has the surprising implication that someone could be coerced without her knowledge. In any event, a mistake of fact generally exculpates a defendant only if the mistake refers to an element of the offense or a justifying or excusing circumstance. It is not clear how mistakes regarding the probability of detection are relevant to moral blameworthiness. 

Mark DeGirolami’s essay, “The Choice of Evils and the Collisions of Theory” defends the doctrine that a defendant who commits an offense in order to avert a greater evil may not benefit from the ‘lesser evils’ defense if she culpably created the choice between doing the lesser evil and allowing the greater evil. Specifically, DeGirolami argues that “an actor who has purposely and criminally caused the conditions leading to a choice of evils does not deserve the defense” (214). DeGirolami offers this as a retributivist account of the restriction, but this is somewhat misleading. Retributivism is primarily concerned with what punishment an individual deserves, not with what defense an individual deserves. Presumably, a defendant deserves a defense just in case the defense explains why she deserves no punishment or, in the case of partial defenses, less punishment than she would otherwise deserve. But DeGirolami does not tell us why a defendant who culpably creates a choice of evils and then chooses the lesser evil deserves the same punishment as a defendant who faces no choice of evils at all. For example, DeGirolami describes a man who, while drunk, strikes his girlfriend but, after realizing that her injuries are life-threatening, drives her to the hospital. DeGirolami says that retributivism explains why this man does not deserve to defend his drunk driving as the lesser of two evils. But DeGirolami does not explain why this man deserves the same punishment as a man who injures his girlfriend and then drunkenly drives away to escape detection. Instead, DeGirolami defends the restriction by claiming that an act is only justified if it is “right, laudable, or praiseworthy” (215) and that the choice of a lesser evil is not laudable or praiseworthy if the need to choose was of the chooser’s own making. I doubt very much that this is the correct account of justification (for example, there is seldom anything particularly laudable or praiseworthy about acting in self-defense). But there is certainly nothing particularly retributivist about this account of justification.

Richard L. Lippke’s essay, “Retributive Sentencing, Multiple Offenders, and Bulk Discounts,” is my favorite of the four and deals with one of the most important and difficult issues in sentencing theory. Lippke asks whether individuals convicted of multiple offenses should serve the normal sentence for each crime consecutively (“Strong Consecutivism”), concurrently  (“Strong Concurrentism”), or something in between (“Moderate Consecutivism” and “Moderate Concurrentism”). Lippke rejects Strong Concurrentism based on the principle that “Each Victim Counts,” according to which it would be unjust to give a defendant the same punishment irrespective of the number of his victims. Lippke rejects Strong Consecutivism based on the “Totality Principle,” according to which a defendant’s total punishment should not exceed her overall desert, as it might if, in Lippke’s example, a defendant convicted of defrauding 100 people of a given amount is sentenced to 100 times the sentence for defrauding a single person of that amount. Lippke ultimately defends a form of Moderate Concurrentivism according to which “multiple offenders should not be made to endure penal losses and deprivations the total impact of which exceeds, by too significant a margin, the harms which they culpably inflicted on any one of their victims” (220). Lippke’s position is intriguing, but he will need to say more in its defense. For one thing, his position seems to entail that as the number of victims goes up, a defendant must receive less and less punishment per victim in order to keep her total punishment from too significantly exceeding the harm suffered by any individual victim. His position therefore seems to entail either that the more victims you wrong the less punishable each past wrong becomes, or that the more victims you wrong the less punishable the next wrong becomes. In any event, Lippke’s discussion of the possible responses to multiple offenders is invaluable and I recommend it strongly. 

Finally, Thom Brooks’s essay, “Retribution and Capital Punishment”, argues that we can never know a defendant’s guilt with sufficient certainty to permissibly sentence her to death. Brooks acknowledges that certainty is never possible in criminal cases, but argues that since a defendant can challenge her conviction both before and during incarceration but only before execution capital punishment denies her additional opportunities to prove her innocence. This denial, Brooks argues, is impermissible. Brooks argues that execution punishes the innocent far more than they deserve, while life imprisonment punishes the guilty only slightly less than they deserve. This is true, but it is also misleading, since it suggests that if the guilty are not executed they will be imprisoned for life, while if the innocent are not executed they will go free. Sadly, this is not the case. If capital punishment is abolished then all of the guilty and most of the innocent will be imprisoned for life rather than executed; all of the guilty will get slightly less than they deserve, while most of the innocent will still get far more than the deserve (and only slightly less than they would if they were executed). If life imprisonment gives the guilty only slightly less deserved punishment than execution then execution gives the innocent only slightly more undeserved punishment than life imprisonment. Brooks does not explain why one deviation from desert is more significant than the other. 

Brooks is clearly right that some of the innocent who would be executed if sentenced to death will overturn their convictions if sentenced to life imprisonment. But this simply brings us back to the question with which we began:  are we morally required to abolish capital punishment in favor of life imprisonment in order to avoid these undeserved executions? If the innocent and the guilty were equal in number, then the slight reduction of deserved punishment of most of the guilty would be offset by the slight reduction of undeserved punishment of most of the innocent (both would serve life sentences); while the remaining reduction of deserved punishment of the guilty would be outweighed by the great reduction of undeserved punishment of the small number of the innocent who overturn their convictions. But of course the innocent and the guilty are not equal in number. The vast majority of those sentenced to death committed the crimes for which they were convicted. Brooks does not confront this fact, and this omission severely undermines his argument. Brooks believes we are morally required to give a large number of guilty defendants slightly less deserved punishment in order to give a small number of innocent defendants slightly less undeserved punishment and an even smaller number of innocent defendants far less undeserved punishment. He may be right. But his conclusion requires a sharper moral asymmetry between inflicting undeserved punishment and foregoing deserved punishment than his arguments establish. Perhaps such a moral asymmetry could be established by invoking the evaluative asymmetry between positive and negative desert as well as the deontic asymmetry between doing and allowing. Or perhaps Brooks will find another, more ingenious route from retributive theory to criminal justice policy. In this project he will enjoy the good company of his co-contributors as well as the continued admiration of this reviewer. 


Mitchell Berman, Two Kinds of Retributivism, in Philosophical Foundations of Criminal Law (R.A. Duff & Stuart P. Green eds., 2011). 

John Gardner & Stephen Shute, The Wrongness of Rape, in John Gardner, Offences and Defences (2007).


Adil Ahmad Haque is an Associate Professor of Law at Rutgers School of Law-Newark

Start typing and press Enter to search