« Further Comments on Capital Punishment--Posner | Main | »



Feed You can follow this conversation by subscribing to the comment feed for this post.


Becker: "I believe it has 'sizeable' deterrent effects."

A panel of the National Academy of Sciences reviewed the available research on the general deterrent effects of criminal sanctions. The panel reported that the empirical studies suffer from methodological weaknesses so severe that no conclusions could be drawn. Among the studies reviewed was Ehrlich's, "The Deterrent Effect of Capital Punishment: A Question of Life and Death," American Economic Review, June 1975.


I wonder what level of evidence it would take to convince you that "[the death penalty] is administered in such a racially biased manner as to wrongly convict many black persons, and to be little used against white murderers." I cannot cite you studies, but perhaps you could cite some studies you know of that dispute the often-cited claim that executions are racially-biased -- some modern ones.

In dealing with such powerful negative externalities (one of the strongest, really, though we can imagine stronger externalities yet), what level of wrongful execution are we willing to tolerate? Wrongful executions are perhaps one of the most fundamental violations of a human's right to life -- though admittedly most economists and consequentialists do not believe in "rights." Still, it seems odd to advocate strong property rights because of its consequence in raising welfare levels over time, but advocate weak life rights because of its possible consequence in reducing the deterrence effect of the criminal justice system.


"For those who find such a position "barbaric", I would ask how many innocent victims are they willing to tolerate before they might take a more positive position on these additional punishments?"

as a more-or-less determinist, I view extremely violent crime as indicative of a "diseased" mind (in a social, but not necessarily medical, sense ). from that perspective, the primary objective of "punishment" for such crime is to purge society (including that portion in prisons) of those who pose extreme danger to others and to do so in a humane way, including possibly capital punishment. to intentionally inflict suffering suggests holding criminals not only accountable for their antisocial acts but also responsible for being willfully "evil".
by itself, of course, this doesn't answer prof becker's cost-benefit trade question. to make it less abstract, suppose science and technology one day provide the ability to keep a person alive and conscious indefinitely and in continuous agony (sound familiar?). would you do that to one guilty of even an extremely violent crime if it saved scores of innocent lives? only one innocent life? just for vengeance? for lesser offenses if it had a large deterrant effect?

my concerns are that viewing criminal sentences as "punishment" rather than mere accountability is based on philosophically (and biologically?) questionable assumptions and diverts attention from the fundamental objective of protecting society, and that adding torment is way too close to religious inquisition, even if it can be rationalized due to its deterrence value.


Add to the hypothetical equations a very real number. For each innocent person convicted for a murder, a real murderer or murderers are free to kill or harm others again. Once the penalty has been carried out the search for the perpetrators will likely never begin again. This is rarely addressed by police, prosecutors or judges that are proponents of the death penalty.

However, an innocent man sentenced to life in prison may shout loud enough and long enough that someone will take notice and the investigation in to what really happened may take place. If this occurs, there is a chance that the real criminal may then be brought to justice at some point in the future. This chance seems to be virtually non-existent in the current system.


Granted what we are arguing here is much like debating politics and the existence of God, but I would still like to throw a few ideas out there that have been somewhat sidestepped. How does one begin to measure a murder that has been deterred? If we have an instance of no-murder, this does not mean a murder has been deterred. I seriously doubt the validity of the deterrent effect of the death penalty given that it is impossible to measure its effects in a sound, valid, and reasonable manner. Do murderers travel over state lines from pro-death penalty states that impose such a punishment to states where there are no such penalties to avoid being sentenced to death? Would that be considered the deterrent effect?

Allow me to play devil?s advocate and ask you run a thought experiment where the federal government forced each state to carry the death penalty and fill a quota (by quota I mean if there is a murder we issue the death penalty to the convicted murderer) of death sentences each year (or whatever time period you like) based on the amount of severe negative externalities we wish to reduce and based on the number of murders committed. The quota would have to be based on the evidence to date of the deterrent effect of the death penalty. The quota would also have to be based on the murder rates idiosyncratic to each state and would of course differ across states based on such information. Then, say we had no murders in a given state in a given year; we would obviously have no death penalty sentence handed out and no quota to be filled. Would you be comfortable saying that the death penalty has in effect done its job and deterred a given number of unknowable murders? What if we run the same experiment in the next time period and in the state where there were no murders, we have 3 (or any number) and in a state where we had X murders we now have none. Where is the deterrent effect in that? I could go on forever with such an experiment and ask you over and over again how you feel about the way I am measuring the deterrent effect of the death penalty and whether or not it has worked, what kind of conclusions do you think we would come to?

Just because have lower murder rates does not in anyway mean that this is because of the death penalty. Just because a town is having a lot of babies and that town also has a high population of storks, does not mean the storks are flying the babies to town in little white handkerchiefs and dropping them off in bedrooms across the town. We have to be careful deeming spurious relationships as causal relationships. The death penalty is in fact a useful tool when it is used as Hammurabi intended but to use it as a deterrent is somewhat questionable as we can neither measure uncommitted murders, murders deterred by such a penalty, etc.


I have a hard time believing 100% in deterrence as the sole reason for capital punishment. We have had capital punishment for a long time, and yet people keep murdering.

It becomes necessary to get into more details on academic studies.

Some people appear to have a death wish and appear to not care if they are executed after murdering other people. This situation seems to be happening a lot recently. It would be interesting to know how common this is and how it has trended over time. Perhaps my perspective is biased by frequent consumption of mass media. It may be possible that deterrence attributable to capital punishment has changed significantly over long periods of time due to changes in other factors (eg, media and information proliferation).

At least two movies may be relevant:

"Minority Report": Maybe science and technology will give us new methods to manage murders in the future (and new government abuses of power).

"Natural Born Killers": Some people may kill or murder for media attention and posterity. This motive may be getting worse in modern time - not sure. Murderers know that murderous events will be reported over cable news and the internet instantly.

P.S. What about "crimes of passion" where people murder spontaneously? Do such people grasp the consequences at the time? And how common are unpremeditated murders and is this even relevant to this discussion?


The singlest largest problem with the concept of deterrence I see, from studies of criminal behavior, is the questionable assumption that engagers of crime take the risk of punishment into consideration.
Deterrence would only work if there was a reasonable surety of consequences, and a rational understanding of those consequences.
However, as forensic psychologists have long known, few criminals believe that they will get caught, or care if they do (and that includes murderers). Not to mention the fact that a large percentage of the criminal population has rather low intelligences, the mean is considerable below that of the average -- their ability to make the kind of decisions that the judicial idea of deterrence implies is inherently handicapped.
The only viable use of punishment, whether execution or imprisonment, is not deterrence or revenge, but simply to withdraw those individuals who have commited crimes from the general population; preventing them from further actions.

Arun Khanna

Professor Becker said: "Some readers asked whether I also favor public executions of convicted murderers, mangling of their bodies, and other methods used in some countries still, and in most countries in the past? I do not because they seem unnecessarily abusive of convicted murderers without any compensating gains. However, I admit I would reconsider this position if it were demonstrated that such added punishments have a large effect in reducing the number of murders."
If it were demonstrated that such added punishments have a large effect in reducing the number of murders, it might be better to outsource such executions to certain Middle East countries. The information asymmetry introduced by such an outsourcing or rendition policy in the minds of criminals will act as an additional deterrent.


Nate said.."P.S. What about "crimes of passion" where people murder spontaneously? Do such people grasp the consequences at the time? And how common are unpremeditated murders and is this even relevant to this discussion?"

-----I believe premeditation is required for the death penalty. I'm no expert though, and this is an interesting question.


I am not an expert either.

HamletsMill makes interesting points. Murderers may not weigh the consequences prior to the crime - premeditated or unpremeditated.

And even though murders continue despite capital punishment and even though not all murderers weigh consequences, there may be to this day a small subset of potential murderers out there who will be deterred by the threat of capital punishment. Maybe this deterrence alone justifies capital punishment. Maybe the idea of removal from society and justice (or "revenge" or "retribution") is superfluos or dangerously subject to abuse.


"superfluous" not "superfluos"

Cogliostro Demon

Here in Florida we kill with great abandon, officially and privately. We are one of the top states when it comes to executions, yet we have a high rate of murders. If the death penalty deters murder why do states like Florida and Texas which have high execution rates have high murder rates?

Frankly, this is a topic which needs to be illuminated by rigorous mathematical analysis. I think that when the government kills it sends a message that killing is good. The death penalty is the opposite of a deterrent, it encourages people to kill.

Texas leads the country in executions, but, like Florida it has a high murder rate. Why?



When the state arrests people and locks them up for murder, does that send the message that kidnapping is good? (And let's not get into tax collection....)

Similarly, Chicago has a lot more snowplows and road salting trucks than Miami, yet also has more snow and more icy roads in the winter. Why?

Bob K

While economists have no authority nor competence to discuss the morality of public policy, and it is best that they abstain to enter into such discussions, it is a whole different issue to argue that morality should not matter at all.

Economists are needed to caution against possible unintended consequences of measures taken with lofty moral objectives.

But to say that morality does not matter at all? Does the end justify the means? Why do humans, generally, get repulsed and rebel against fully efficient governments with no morals? Maybe it is because morals enter our utility functions? Is it part of our biological survival programming? I don't know, but I would never say that morals are not important to evaluate public policy.

Bob K

Cogliostro has a point. If I murder someone else, then I know that if caught I will get the electric chair. After that first murder, the marginal cost of killing a cop, or another person is zero. I can't get the death penalty twice.
So, for criminals and serial killers, the death penalty makes murder cheaper, not more expensive as Becker says. That's why Florida and Texas might have a higher murder rate.

Bob K

In other words, the death penalty has a deterrent effect on the number of murderers per capita, but, in principle, a positive effect on the number of murders per murderer. The net effect is uncertain.


Professor Becker writes: "I believe that deterrence can be the only reasonable basis for capital punishment. Revenge, retribution, and other arguments sometimes made to justify capital punishment are too subject to government abuse, and have been abused."

What makes the case for capital punishment as a deterrent any less subject to abuses from government? In fact, give the age-old axiom of "lies, damn lies and statistics", I would think that "deterrence" as the sole policy motivator would be even more subject to abuse, especially given the decidedly mixed perponderance of the evidence. For example, unquestionably capital punishment if performed very quickly after the crime would have a great deterrent effect; maybe 10 lives saved per execution if all executions were performed within six months after the murder.

Economically speaking, such a policy would make tremendous sence, even if a few "innocents" were executed the greater good would have been served.

I would suspect that injecting some moral or social standard of abstract justice, such as revenge or retribution would in fact have the opposite effect. Revenge is an emotional good only obtained when the right person is put to death, the same with retribution. Sanitizing the policy considerations for capital punishment to merely "deterrence", in my opinion, increases the likelihood of government abuse, by reducing the considerations to a variable completely unconnected with the person to be killed.


Both deterrence and the satisfaction of revenge are based on the people experiencing these feelings/effects belief that the right person gets punished most of the time. If I'm a prospective murderer, and I think the police will, upon finding the body, round up the usual suspect and beat a confession out of one of them selected at random, then there's not much deterrent effect on me. If I'm an angry family member of a murder victim, and I think the police have beaten a confession out of some random street criminal and executed him, I'm not going to feel very satisfied.

These effects are independent of the truth. I'll be deterred by the threat that the police will catch and punish me if I believe it, even if the police are really too incompetent to catch any real criminals. I'll fail to be deterred if I think the police are incompetent, even if they're really very likely to catch me.


Bob K

My only discomfort with the claim that the marginal cost of a second murder is zero under the death penalty is that the chance of capture or of a death penalty verict can still increase by murdering more. The second murder is therefore still has costs for the criminal.

Btw, the second-murder argument doesn't require the death penalty to work: if a state's maximum sentence - whatever it is - is imposed for the first murder then a second murder will be similarly undeterred, notwithstanding my previous point.

Cogliostro Demon


A kidnapping is not an arrest. An execution, be it gangland or Bush style, is still an execution.

Killing a woman strapped to a hospital gurney and killing a man strapped to an electric chair are both killing. Why is murder OK if the government does it?

Northern Observer

Couple points to add:

The stubborn persistance of wrongful convictions in criminal cases, usually based on mistaken eyewitness accounts.

It does not deter. Too many studies say so and the ones that say it does upon closer scrutiny turn out to be gamed. To say that it does because men fear death is to move into the realm of faith and belief. Not that these fields have nothing to add to the debate but to frame a rational defence of capital punishment around them is darkly comic.

The company we keep. When you look at the list of countries that use capital punishment as a form of justice and those that don't one can only ask the United States. Why? Why must you do this?

The historical experience. The institution of African chattle slavery in the old confederacy and capital punishment are closely linked. It was a tool of social control and dominance in the Old South. To use it in the modern era without hesitation is to cheerlfully ignore America's darker impulses, as if the past had no possible hold on the present.

Which brings me to a final comment. To believe in Capital Punishment you must believe in a from of human perfection and human judgement that in my opinion surpasses the ability of real men to achieve in the persuit of justice. All the countries that execute have this in common, the arrogance to think that they can be perfect. It is a conceit that has led America to do evil in the world of men in many fields and I look forward to the day when those who shape public opinion in the Republic approach the world and its problems with a little more humility.



A dozen guys with guns show up at your house, and by threat of violence carry you off to some heavly secured house, where they hold you for several years against your will. How is this not kidnapping? If it were done by the Mafia or Al Qaida, we'd call it that without hesitation.

The point is, we normally don't accept private citizens applying punishments for crimes. Instead, we want the police and courts doing that, since it seems more likely that they'll be somewhat neutral about it, will make some effort to imprison or hang the right person, etc. That means we have the police/courts doing things we don't accept that private citizens can do to punish a crime, whether that's imprisoning people, flogging them, hanging them, torturing them, whatever. Which of those the state ought to be doing is a different moral question.


Northern Observer:

The use of slavery and capitol punishment is closely linked? That is no different than saying cp is closely linked to agriculture. They all go back thousands of years. The US and its 'dark history' have nothing to do with either. They were invented elsewhere, and would continue, with or without us.


I did not post on this blog on capital punishment during week 1 because:

-I am not a lawyer or judge
-it is easy to be misunderstood on a blog
-sometimes it is nice not to think about capital punishment

The book "Crime and Punishment" comes to mind:


Sane Canadian

Capital punishment is a deterrent. Studies have shown otherwise, but studies show a lot of things. Correlating murder rates and the death penalty is especially tricky, because of all the possible factors that could affect crime. If we could randomly introduce the death penalty in half of the world's cities, and ban it in the rest, we might be able to get some useful data.

Until then, we should conclude that capital punishment deters based on some very common-sense assumptions.

1) Criminals prefer life to death

2) Criminals are aware of the penalties they face for the crimes they commit

3) Criminals behave rationally

Most arguments against the deterrent effect of capital punishment focus on criminals' lack of ability to reason properly. ("Criminals don't think that they will get caught" or "Criminals don't think about what they're doing.")

Shooting rival gang members in the leg is very common, since the shooter's offence for that is "assault with a deadly weapon" instead of "attempted murder." Sounds pretty rational to me.

But all of this is so obvious that I can't believe it even has to be said. Living proof of the deterrent effect exists in my rommate, who tests my self-restraint with silly arguments like that all the time, without reprecussion.

- C

The comments to this entry are closed.

Become a Fan

May 2014

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31