Utilitarianism and punishment



On utilitarianism, punishment is justified only insofar as it promotes the general happiness.


The following are potential utilitarian reasons to punish:


            · To prevent offenders from doing further harm

· Deterrence (to prevent recidivism and discourage other potential offenders)

            · To give satisfaction to victims/society

·  To reform/educate offenders and make them into more productive members of society.


For the utilitarian, it is not morally permissible to punish criminals in order to give them what they deserve by exacting revenge or retribution on them.  No one “deserves” anything…our only duty is to maximize utility.  The basic problem with retributivism, for the utilitarian, is that “it advocates the infliction of suffering without any compensating gain in happiness.” (Rachels, p.133)



Kantianism and punishment


For Kant, giving criminals what they deserve is the only legitimate reason to punish them.  If we punish them in order to promote happiness, then we are violating the categorical imperative by treating them as a mere means to an end.  So, all of the utilitarian justifications for punishment are bad ones, according to Kant.

            Kant advocated two principles regarding how punishment should be administered.


(1)  People should be punished simply because they have committed crimes, and for no other reason.


(2)  Punishment should be proportionate to the crime.


            Notice that utilitarianism does not endorse either of these principles.  Contrary to (1), it advocates punishing as a deterrent or in order to reform criminals.  Contrary to (2), the severity of punishment will depend on what maximizes utility, regardless of whether it is proportionate to the crime.  This might involve making an example of an offender by administering a severe punishment in order to discourage others from committing the same offense.  It might also involve letting some people (such as first time offenders) off light, on the grounds that a lesser punishment will be enough to discourage them.

            In keeping with (2), Kant supported capital punishment for murderers.  The question arises, how does Kant justify punishment in the first place?  Doesn’t punishment itself involve violating the categorical imperative?  If you punish someone, aren’t you violating their autonomy?  In the case of capital punishment, the punishment literally ends a person’s autonomy.  Wouldn’t the same kind of reasoning that prohibits suicide also prohibit killing a criminal?

Not according to Kant.  Criminal actions, like any other actions, are associated with maxims.  The maxims of criminal actions (theft, assault, etc.) are ones that endorse harming or otherwise violating the autonomy of others.  For example, I will rob you when I have the opportunity to do so in order to promote my own interests.  As rational agents, people who act on these maxims are endorsing them as universal law.  They are saying, in effect, this is how people ought to be treated. 


When we decide what to do, we in effect proclaim our wish that our conduct be made into a “universal law.”  Rachels on Kant, p. 139


Now, let’s focus on the second formulation of the CI, CI2.  It tells us that we ought to treat others always as ends, and never merely as means.  By punishing a criminal, we are respecting his ends, because we are treating him in the way that he thinks people ought to be treated.  Thus, we are not punishing him for our own benefit (nor for his benefit), but because it is in accordance with the principles that he has endorsed through his actions.  For Kant, this would be a way of “sharing in his ends”.  Now we can see how Kant could make a distinction between suicide and capital punishment.  If someone commits suicide in order to avoid the unpleasantness of a long death, Kant would say that person has been used as a mere means to an end:  that person’s rational agency—their freedom—has been destroyed in order to prevent unhappiness.  But if we kill a murderer because he has murdered, then we are not using him as a mere means to an end.  We are not killing him in order to increase happiness.  Rather, we are killing him in accordance with the principle of action that he chose to live by.  As Kant says, “His own evil deed draws the punishment upon himself.”

Of course, the Kantian justification for punishment hinges on the assumption that the punished actually chose their actions autonomously, using their own rationality.  It depends, in other words, on the assumption that they were moral agents when they acted.  If they were not moral agents; if instead, they were like wild animals, or they were insane, then Kant’s reasoning would not apply.


Capital Punishment


Some Recent Capital Punishment Statistics

From the Bureau of Justice Statistics


About 3% of those convicted of criminal homicide are eventually executed for it.

The median stay on California’s death row is about 13 years (it is somewhat shorter in some other states).

In 2003, 65 inmates were executed, 6 fewer than in 2002.

Prisoners under sentence of death

  At yearend 2002, 37 States and the Federal prison system held 3,557 prisoners under sentence of death, 20 fewer than at yearend 2001. All had committed murder.  

More statistics available at:  http://www.ojp.usdoj.gov/bjs/cp.htm

Notes on


“The Case Against the Death Penalty” by Adam Bedau




“A Defense of the Death Penalty” by Ernest van den Haag

We will consider Bedau’s points, one by one, followed by the rebuttals of Haag.

Bedau begins by noting the  1972 case of Furman v. Georgia, in which the Supreme Court effectively struck down the death penalty as it then existed in many states on the grounds that it had been applied in a “harsh, freakish, and arbitrary” way.  Many states then re-instated the death penalty with new standards that required juries to decide on penalties separately from guilt and innocence, and to use uniform criteria in applying the death penalty.  (The “special circumstances” deemed relevant to deciding whether capital punishment is appropriate include premeditation, the number of victims, and whether additional crimes were committed during the homicide.  For example, a rape + homicide is a better candidate for the death penalty than a mere homicide.)

Bedau claims that capital punishment is not a deterrent to capital crimes.

This is obviously relevant to whether the death penalty is justified on utilitarian grounds.  His argument is as follows:

1.  A punishment can be an effective deterrent only if it is consistently and promptly employed.

2.  The proportion of  first-degree murderers who are sentenced to death is small, and of this group, an even proportion of people are executed.  [Specifically, only about 1% of homicides (about 300 per year) result in the death penalty.  About 3% of those convicted of criminal homicide are eventually executed for it].

3.  Mandatory death row sentencing is unconstitutional.  (Woodson v. North Carolina 1976 ruling).

4.  A considerable time between the imposition of the death sentence and the actual execution is unavoidable, given the procedural safeguards required by the courts in capital cases.

5.  We can reduce delay and costs only by abandoning the procedural safeguards and constitutional rights of suspects, defendants, and convicts—with the attendant high risk of convicting the wrong person and executing the innocent.

6.  When crime is planned, the criminal ordinarily concentrates on escaping detection, arrest, and conviction.

7.  Most capital crimes are committed in the heat of the moment.

8.  For these reasons, capital punishment cannot be administered consistently or promptly employed.

9.  Therefore, capital punishment is not an effective deterrent.



The first premise is an over-simplification.  It should be changed to read, “Punishments tend to have more of a deterrent effect to the extent that they are promptly and consistently employed.”  After all, deterrence comes in degrees.  But now the objection arises:  the death penalty might have some deterrent effect when applied inconsistently after long delays, even though it would be a stronger deterrent if applied promptly and consistently.

Consider an analogy.  The RIAA now sues people who share copyrighted music on peer-to-peer computer networks.  But they only sue a relatively small number (probably less than 1% of the total).  The evidence suggests, however, that the threat of lawsuit has had a significant impact on file sharing – fewer people are doing it, and many are sharing fewer files.  (One informal study showed that file sharing was down about 30% after the first round of lawsuits).  In fact, inconsistent enforcement might be a more effective deterrent than a consistent standard by which only the “worst of the worst” are sued.  If enforcement is inconsistent, no one can be sure that they won’t be next.

The data regarding the death penalty is hotly debated.  Here are excerpts from a paper by Jay Gaskill (my father and retired Public Defender) “Death, Deterrence and Reform” where he argues for a deterrence effect:

“I find the evidence in favor of the death penalty’s deterrent effect on homicidal behavior to be highly persuasive, leaving aside the more difficult issue of measurement of the power of the effect on a given, demographically mixed population. The so called side-by-side studies that purport not to reveal any deterrent effect (for example comparing death penalty enforcing state A with non-death penalty state B over the same time frame) fail to normalize for demographic differences.  There are always higher “crime prone” sub-populations in any geographic area. At any given moment, virtually all states differ in political and social attitudes, police funding and activities, and in the detailed operation of their respective criminal justice processes.


The temporary imposition of a well publicized death penalty moratorium in a given jurisdiction provides a better quasi-controlled experiment, particularly when demographic factors remain relatively stable over the sample period.  And, in these samples, the larger the population that is included, the less that pockets of demographic variations will skew the outcome.  That said, there can be demographic and cultural changes with time.


With those qualifications, analysis of the available data has persuaded me that the death penalty for murder may have saved a significant number of lives over the last decade in those jurisdictions where it was used.”


Several recent studies have concluded that there is a significant deterrent effect:


“Journal of Applied Economics.”

4/01, Vol 33, N 5, p569 -- p576,  Execution Moratorium Is No Holiday For Homicides; Execution and deterrence: a quasi-controlled group experiment.” Professors Dale O. Cloninger and Roberto Marchesini University of Houston. [Professor Cloninger (cloninger@cl.uh.edu), Professor Marchesini (marchesini@cl.uh.edu)]


“The (Texas) execution hiatus (in 1996), therefore, appears to have spared few, if any, condemned prisoners while the citizens of Texas experienced a net 90 (to as many as 150) additional innocent lives lost to homicide. Politicians contemplating moratoriums may wish to consider the possibility that a seemingly innocuous moratorium on executions could very well come at a heavy cost.”


Professor Cloninger: “ . . .  (Our recent) study is but another on a growing list of empirical work that finds evidence consistent with the deterrence hypothesis.  These studies as a whole provide robust evidence -- evidence obtained from a variety of different models, data sets and methodologies that yield the same conclusion. It is the cumulative effect of these studies that causes any neutral observer to pause.”


Emory University Paper:

Does Capital Punishment Have a Deterrent Effect: New Evidence from Post-moratorium Panel Data”, Located at userwww.service.emory.edu/~cozden/dezhbakhsh_01_01_paper.pdf


Emory University Economics Department Chairman Hashem Dezhbakhsh and Emory Professors Paul Rubin and Joanna Shepherd: “Our results suggest that capital punishment has a strong deterrent effect.  An increase in any of the probabilities -- arrest, sentencing or execution -- tends to reduce the crime rate. In particular, each execution results, on average, in eighteen fewer murders -- with a margin of error of plus or minus 10.” Their data base used nationwide data from 3,054 US counties from 1977-1996.


University of Colorado (Denver) Paper:

Pardons, Executions and Homicide”, H. Naci Mocan (mmocan@carbon.cudenver.edu) and R. Kaj Gottings (rgitting@carbon.cudenver.edu), October 2001, located at http://econ.cudenver.edu/mocan/papers/deathpenalty1007.pdf


University of Colorado (Denver) Economics Department Chairman Naci Mocan and Graduate Assistant R. Kaj Gottings found “a statistically significant relationship between executions, pardons and homicide. Specifically each additional execution reduces homicides by 5 to 6, and three additional pardons (commutations) generate 1 to 1.5 additional murders.” Their “data set contains detailed information on the entire 6,143 death sentences between 1977 and 1997.



Between June 1967 and January 1976,  these was an effective national moratorium on executions (see the California experience below for more details).  From 1966-1980, basically for the entire moratorium, U.S. criminal homicides jumped from 11,040 to 23,040, and the murder rate increased from 5.6 to 10.2/100,000.  During that quasi moratorium period, the US averaged only 1 execution every 3 years, with a maximum of two executions per year.  After the moratorium: From 1995-2000 executions averaged 71 per year. The US murder rate dropped from a high of 10.2/100,000 in 1980 to 5.5/100,000 in 2000, a 46% decrease.  For a population of 200 million, the net 4.7% increase in annual murders represents an additional 9,400 killings.



Bedau’s second line of argument:  Capital Punishment is Unfair

            What he means is that capital punishment is unfairly applied.  He cites statistics that show that blacks are far more likely to get the death penalty than whites.  Even accounting for the fact that the murder rate is higher among blacks than whites, there is evidence of racial discrimination.  An “exhaustive statistical study” of capital cases in Georgia concluded “the average odds of receiving a death sentence among all indicted cases were 4.3 times higher in cases with a white victim.”  As Bedau points out, “Of the 313 persons executed between January 1977 and the end of 1995, 36 had been convicted of killing a black person while 249 (80%) had killed a white person…Our criminal justice system essentially reserves the death penalty for murderers (regardless of their race) who kill white victims.”

            Bedau also notes that defendants in capital cases are usually unable to afford their own lawyers, so are appointed lawyers by the state.   He insinuates that this often results in an inferior defense.

His argument seems to be:

1.  We are incapable of applying the death penalty fairly.

2.  If something cannot be done fairly, it should not be done at all.

3.  Therefore, the death penalty should not be used.


Haag, in “A Defense of the Death Penalty”, responds by denying the second premise.

  “Maldistribution of any punishment among those who deserve it is irrelevant to its justice or morality.  Even if poor or black convicts guilty of capital offenses suffer capital punishment, and other convicts equally guilty of the same crimes do not, a more equal distribution, however desirable, would merely be more equal.  It would not be more just to the convicts under sentence of death.”

            Essentially, Haag is saying that it is not unjust to give people a punishment that they deserve simply because there are others who deserve the punishment who are not getting it.  Faced with a choice between giving justice to some or justice to none, it is better to give justice to some.  Presumably, he would also favor making application of the death penalty more uniform.

            An analogy:  Suppose you are sitting on a jury.  You know that the defendant is being charged with a crime that he would not be convicted of in the next county (perhaps because prosecutors in the neighboring county would not prosecute for shoplifting).  On the assumption that the evidence indicates guilt, would what happens in the next county be a reason not to convict?  Arguably not, for what goes on in the next county has no bearing on the question of guilt and innocence in the case at hand.  Similarly, you might think that whether the defendant would be given the death penalty if his victims were black is irrelevant to whether he deserves the penalty in the case at hand.

Reply:  If there is a suitable substitute punishment (life without parole) that could be applied fairly, then we should use that instead.


Bedau’s third line of argument:  Capital Punishment is Irreversible

Occasionally, innocent people are executed.  In many more cases, there are “near misses” in which people are given last minute reprieves in light of new or unexamined evidence.  This would not be an issue if the death penalty were done away with.

In reply, it can be argued that the justifications for the death penalty outweigh the cost of occasional miscarriage of justice.  Whether you think that is true will depend on what you think of those justifications.  We have already discussed deterrence.  The other main justification is that the death penalty is what some criminals deserve, quite apart from whether it deters others.  This brings us to:


Bedau’s fourth line of argument:  Capital Punishment is Unjustified Retribution

Bedau accepts that “the severity of punishments must be proportional to the gravity of the crime—and since murder is the gravest crime, it deserves the severest punishment”

But he denies that the severest punishment must be the death penalty.  It could be life imprisonment without the possibility of parole, instead.  All that the premise requires is that punishments for murder be more severe than punishments for other crimes.  “But severity of punishment has its limits—imposed by both justice and our common human dignity.”

Replies:  Execution can bring “closure” to the loved ones of victims, which is not the case for life imprisonment.  Also, the prospect of life imprisonment just doesn’t have the same deterrent effect as the prospect of execution.