« More on the Economics of Capital Punishment-BECKER | Main | Further Comments on Capital Punishment--Posner »



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


Would you be so sanguine if you, or some one
you love, were the person unjustly being
executed? Would you be able to argue with
the same rigor the benefits to society at
large despite its small margins of error?
I am genuinely curious.

Donald L. Dawson

I am an avid reader of your weekly comments and almost never disagree with either your analyses or conclusions. While I have no religious bias against capital punishment and certainly not a scholar's familiararity with all of the studies, I do seriously doubt the conclusive demonstration of deterrance cited by Erlich and other more recent studies. It seems to me that the very citation of 59 executions in a year when there was 7000 murders leave a very serious mathmatical doubt about the value of deterrance. Moreover, law enforcement organizations and prosecutors are under enormous pressure to "solve" and punish crime. I would suggest that pressure and the competative structure of our judicial process leads to more frequent error than is generally admitted or recognized.


Everyone dies eventually. If the money that rich people spend on such frivolous pursuits as landscaping their yards was instead spent on scientific research to engineer an "immortal" vessel for human consciousness then billions of lives would eventually be saved. Even if that money was merely spent on eliminating the worst effects of world poverty in the short term then tens of thousands of people every day would be prevented from dying premature deaths.The usual argument that rich people use to justify frivolous personal spending is that there is a difference between taking a human life and not saving a human life: that it is not murder to fail to act to save another human's life. They would further argue that even if their money was spent on world poverty then it is not completely certain that lives would be saved.Richard Posner's argument in favor of the death penalty seems to be that by killing someone with complete certainty there is a chance that society could save some other people's lives. It is not, however, at all clear that killing one person is exactly balanced by the indirect possibility of saving someone else.In a democratic society, the existance of capital punishment will be closely tied to that society's views on killing people in order to achieve some objective. It is not scientifically meaningfull to discuss a deomcratic society in which these two things are not tied together.A democratic society that had both the death penalty and also had the view that taking human life for any reason was wrong would have a very low murder rate but such a society would, by definition, not exist.A scientifically meaningful question would be: "Does the choice of a society on whether to have the death penalty correlate with it's homicide rate?" Similarly, it would be scientifically meaningful to ask: "If a leader, who is already in power in a democratic government, chooses to advocate for or against the death penalty, how will that affect the murder rate?"It is not, however, scientifically meaningfull to ask: "Does the death penalty increase or decrease the murder rate?"Ultimately, the death penalty is merely a chance for a society to express its aesthetic preferences. If my personal aesthetic preferences were aligned with that of the United States there would be the death penalty for leaders (both corporate and political) who abused their power (eg. using their authority to mislead or take freedom from their constituents) and also for reckless and aggressive drivers. On the other hand, the down and out who grew up in broken homes and killed someone out of pure stupidity would merely get life in prison. Essentially, the death penalty would only be used on those who were smart enough to be deterred by it.

Mr. Econotarian

Unfortunately, cohort comparison with many countries that do not have the death penalty seems to show that one can have a much lower murder rate while still not having a death penalty.

I think it is far more important to determine why the murder rate is so high in the U.S., despite the existance of the death penalty here.

I must admit that I am not a fan of giving the state any additional power over life or death than absolutely required, although I certainly do not think that the death penalty, when applied to actual murderers, is abhorent in and of itself.


The study of Hashem, et al, leaves a great deal to be desired. Simply put, we don't know to any meaningful degree of certainty how much deterrent effect the death penalty has because none of the studies, theirs included, are good enough to tell us. We know that places with the death penalty have higher murder rates than places without the death penalty, and we also know that a great many factors other than the state of the law have far more powerful impacts on murder rates than the laws in a particular state. Being in an inner city or not, and a host of demographic factors are far better predictors of murder rates. New York City, for example, has seen dramatic changes in its murder rates without adopting the death penalty.

We also know that a great many murders that could factually qualify for the death penalty are not charged as capital murder, and that there is very strong statistical evidence to show that the single biggest factor in determining whether murder will receive the death penalty is that the victim was white and that the perpetrator was a black man. Intraracial murders and murders committed by a white man (or any woman) against a black man are far less likely to result in the death penalty.

Because the impact of the death penalty is, at best, a third or fourth order effect in terms of predicting the number of murders committed, it is insanely difficult to make any econometric estimate whose error bars when the data is fairly considered, doesn't render the actual prediction meaningless.

We also know far more about actual innocence as well as "legal innocence" than the post would suggest. Between 1973 and March 2005 a total of 119 individuals have been exonerated and freed, and more have had their sentences commuted from death to a lesser sentence. From 1976 to 2004, 944 people have been executed. When the error rate in the initial death penalty conviction is about 10%, it is hard to have confidence that a meaningful number of wrongful executions of actually innocent people aren't carried out. Wrongful convictions follow a well established profile. They typically involve eye witness misidentifications and/or jailhouse snitches and/or incompetent defense attorneys. But, juries generally aren't informed about the degree to which this kind of evidence has been historically unreliable, and indeed, frequently give it greater weight than other evidence. Until steps such as those taken in Illinois are taken, the reliability of conviction is far lower than the post suggests, and equally important, there is a subclass of those convictions that present a particularly high risk of wrongful conviction which the appellate review process and habeas corpus process that endures so long does little or nothing to address, because every step of those processes relies on the principal that a jury's findings of facts as to disputed issues of witness credibility may not be disturbed on appeal. Coupled with the fact that most death penalty counsel are grossly undercompensated (many people on death row have counsel who were paid under $10,000 in current dollars to represent them, while it isn't ususal for an insurance company to spend $30,000 defending a simple automobile accident) and underqualified in death penalty states, this is a huge issue. While the appellate courts are focused on procedural errors by defense lawyers in the courtroom, the usual problems that cause the innocent to be executed are a lack of factual investigation prior to trial and outside of the court process by the defense (often appallingly little), and bad decision making by juries who believe witnesses who are among the least reliable. Neither is addressed in appeals other than an ineffective assistance of counsel claim which is involves a very high standard of incompetence.

The 59 executions in a year number is also deceptive. There are more than 3,500 people on death row. More than 2,000 of them have been in prison for at least seven years. Tookie Williams was in prison for twenty-five years, if I recall correctly, prior to his execution. With or without death penalty appeal reforms a deluge of executions is approaching.

It is also worth noting that while many states have the death penalty, most executions have taken place in just a handful of states (which have among the highest murder rates in the country). More than half of executions since 1976 have taken place in Texas, Virginia and Oklahoma. Add Missouri, Florida and Georgia and you have accounted for about 70% of all modern American executions. This leaves a very small sample size upon which to base any econometric model and also doesn't account for facts like whether the relatively recent pheomena of making life without parole available in cases that would otherwise be capital has changed the deteurrent effect of the death penalty.

Right now the death penalty costs considerably more than imprisonment to implement per case. Generally speaking, the more important deteurrent factor has been the likelihood of being convicted of something, rather than the punishment that will be received. People who don't think they will get caught are unlikely to care about the punishment.

Dan Cole

Judge Posner's presumption that rational individuals would prefer life imprisonment to death counters Beccaria's 18-century presumption that rational individuals would prefer death to life imprisonment. Arguably, both presumptions are historically correct. In the 18th century, prisons were not nearly such comfortable and well managed places as they are today. Life in prison was unpleasant to say the least, and life imprison would not have been vastly preferable, or much longer, than immediate death.

Today, as Prof. Becker has pointed out elsewhere, prisons are far more comfortable places. Indeed, he has argued that the deterrent effect of the death penalty could be further increased by making prisons even more comfortable. See http://minneapolisfed.org/pubs/region/02-06/debate.cfm.


I commend for Posner for a delightfully delicious post. Here he sidesteps psychological or utilitarian critiuqes and explicitly advocates for the swift death of innocents.

Dan C

Does capital punishment deter murders? Do severe penalties for criminals who commit crimes with guns reduce the number of crimes committed with guns? If a state increases the penalties for illegally using a gun, are criminals rational and do they change their behavior? I don't know.

Why and how do crimes escalate to murder? The killer is often impaired, drugs or alcohol. But my gut instinct is that if the potential killer knows that committing a murder means that he will almost certainly be committing suicide has to deter some potential killers. Certainly many spouses have been saved because the anger felt was not enough to commit the equivalent of a murder suicide ñ or so my wife tells me.

Gang killings are often a cost of doing business or part of a desire to advance within the gang. Perhaps increasing the cost of advancement to include a death penalty risk may alter behavior. But gang members have more to fear from gang rivals then the court system and may prefer a trial by twelve then to be carried by six.

I understand the death penalty in some cases. For example, the death penalty for killing a police officer makes perfect sense. You tell the criminal that as bad as their present situation may be, if they kill the arresting officer they face the death penalty. For profit killers must understand that they can be tried and executed at some future date. While not all potential killers can be deterred, clearly some potential killers can be encouraged to take alternative actions.

I favor the death penalty, but I sometimes fear that I am just pretending that the threat of a death penalty gives me some control over a violent segment of the population ñ Like a raised fist swearing at an approaching storm.


Posner might have analyzed the impact on deterence that rehabilitation has, i.e., murderers who convert into saints in prison convince potential murderers that the desire to murder is an abnormal, transitory feeling. In the face of such information, potential murderers recognize that the desire to murder is wrong rather than a just and normal desire, and so reconsider their plots, deciding instead not to murder. Such a decision not to murder in the face of compelling evidence of rehabilitation is rational; obviously murderers are rational if they can be increasingly deterred by increasingly severe penalties. That Posner neglected to analyze the deterrent effect of rehabilitation vitiates his entire argument; one can simply reverse the polarity of his entire argument so that it is an argument in favor rehabilitation.

As Posner himself says,"The utility comparison seems a standoff, and I will ignore it." Why, then, should we not ignore Posner?

Paul N

I don't doubt the deterrent effect of the death penalty on murder rates, but I oppose it for a different reason than has been mentioned here: the existence of a (federal) death penalty makes it easier for this punishment to be abused in the future, even if it is not abused now.

I would consider the illegality of the death penalty to be an important check on the size and power of government.

Tim Holloway

Donald Dawson wrote:

"Moreover, law enforcement organizations and prosecutors are under enormous pressure to "solve" and punish crime. I would suggest that pressure and the competative structure of our judicial process leads to more frequent error than is generally admitted or recognized."

This is an understatement. I suggest that most persons who actually practice criminal law know that there are a good number of police officers who have no problem telling lies to the jury. These officers may justify this practice because they "know" the person is guilty.


"most executions are in southern states--50 of the 59 in 2004--which that year had a total of almost 7,000 murders".

I take it, then, that the 7000 figure was just for "the southern states". The total figure for the US was much higher, meaning that outside the South, we executed nine people in a year when many thousands of murders were committed in those states. Outside the South, the ratio of executions to murders is surely below 1/1000. Do you still think it true that people would pay a substantial price to eliminate a 1/1000 chance of execution? Would that be rational? At what point is the risk too remote to constitute a real deterrent?

When one factors race and socioeconomic status into the composition of the Death Row population, it's hard to imagine a middle-class white person being executed for a single plain vanilla murder. The risk of being executed would surely deter me, a middle-class white person, but I have almost zero risk. Does it really deter the type of person who actually does the murdering? Very unclear.

All this assumes the statistical validity of the methods used by Ehrlich and his followers, at least as applied to the death penalty data. The raw data, as you know, wouldn't lead you to guess at that result:

"In 2003, the South had the highest murder rate in the country, and that appeared to continue in 2004 even as the South carried out 85% of the nation's executions. The Northeast, which had no executions in 2004, had the lowest murder rate in 2003 and that position appeared to remain the same in 2004." (courtesy of the Death Penalty Information Center).

I also think you are giving short shrift to the likelihood that factually innocent persons are being executed. The number of Death Row inmates freed in the past decade just on factual innocence grounds seems to make it a statistical certainty that we have killed some factually innocent people. I would feel much better about this if I thought governors in death penalty states knew the difference between "there is overwhelming evidence of guilt" on the one hand, and "the jury found him guilty and the court of appeals affirmed" on the other, but they obviously don't. Trial lawyers, however, are all too aware of the gulf that yawns between those two statements.

Tim G

A couple of additional factors I would be curious to see analyzed vis. litigation costs for capital defendants. If it is a rational choice these days to prefer life in prison versus execution, then doesn't the ability to have the death penalty on the table encourage defendants who are likely to be found guilty at trial to plead to a life sentence? (Thereby leading to fewer trials and associated costs). Also, if the death penalty is banned, what do we do with the person who might have been executed, but during the course of his life sentence escapes to kill someone else (or kills an inmate, etc.). How would that potential loss of life (which would not have occurred had the criminal been executed) factor into the analysis? Actually, on the inmate issue, what would stop the psychopath from killing others while in prison if he knew that the worst penalty he would receive is more of the same (i.e., more life sentences)?

I would also be interested in Judge Posner's views on the Cory Maye case being discussed at http://www.theagitator.com/. Maye is on death row after shooting a police officer who entered his home during a "no knock" raid. Maye claims that he thought they were intruders breaking into his home, and that he shot the first person who entered his bedroom (where his infant daughter was sleeping). As soon as he heard someone say "police" he put his gun down. In my opinion based on what I have read, there is no way this guy should be executed for capital murder. Tookie Williams, however, deserves what he got.


I daresay that the risk of executing an innnocent person--no matter how slight--is the single most compelling (and unanswerable?) retort to those who favor the death penalty.


Few people would readily trade their own life for the chance to take someone else's.
When people commit murders they do it with some, perhaps naive, notion that they will evade prosecution, or more likely they fail to calculate the utility of the situtation.
Life imprisonment has a major effect also. If someone were to stop and calmly calculate or even briefly consider that punishment and it was certain, they would rarely if ever commit a crime which would result in that terrible punishment.

The important question is why do people make these
significantly irrational choices? Is it because:
a) they don't believe they will be caught, or
b) they assume they will be caught but the crime is worth spending the remainder of their life in prison.

I suspect that "a" is much more likely and it links back to other parts of our legal system were petty crime is not enforced significantly.

I know that speeding on a highway is illegal, however after countless hours of breaking that law I have never been reprimanded. This is one of the reasons that the Bratton police approach of attentativeness to small crimes is so effective. Every time we commit a given crime and aren't punished we think less about it the next time to the extent that we recklessly disregard many crimes like speeding, littering, or even for some more serious crimes like embersslement. *As the underenforcement of petty crimes raises brazeness it probably also contributes to serious crimes.*
Life imprisonment and capital punishment are both less effective as deterents because people are so used to getting away with crimes. If we did a better job with petty crime, both would become potent deterence tools. Alas, we rely on state execution when speeding traps might work.


In a similar vein to what Paul N. wrote, the deciding factor that turned me against the death penalty is concern that it might be expanded and applied more widely in the future. In the 1980s and 1990s Congress passed laws that provided for a federal death penalty for some levels of involvement in criminal drug enterprises, i.e. crimes that involved contraband but not the death of another person.

This led me to a personal turning point of opposition to the death penalty. I fear the public becoming inured to the prospect of applying in cases where it might shock us today, especially in future times of turbulence. Failing strong evidence that it is greatly superior punishment for murder -- evidence I have not seen -- I prefer to take this dangerous tool out of the hands of a government that might abuse it.

This is no merely theoretical concern. The crime bills passed in the 1980s and 1990s and the growing number of people in prison for drug offences are demonstrations of a trend that seems to amount to a competition among legislators to pile ever harsher punishments on some categories of crimes. I think this competition is disconnected from rational calculation of the costs and benefits to society. In such an environment, the legislature must be deterred by ordinary voting citizens when we detect that their zeal has gone out of control. Taking capital punishment away from them seems a reasonable step, and one that many peer societies and indeed many US states have adopted without terrible costs.


Deterrance is irrelevant. Deterrance cannot distinguish between the practice we have adopted now and the far more brutal and public practices our ancestors used.

If lethal injections deter murders, then it would better deter murders to put people on the rack in the middle of the town square and render them limb from limb. It would deter crime to hang the severed heads of criminals from the walls of the White House. However, I doubt even the most zealous among us would advocate doing that. Something else is staying our collective hand, pushing us to seek the most "humane" way possible to murder black criminals... what could that be...

Capital punishment is a moral question. I submit that the most nasty, guilty, evil, sane person on death row today does NOT deserve to be murdered by the state. Yes, the work of the Innocence Project has proven to anyone who will look that "factually innocent" people have been murdered for someone else's crimes, but that obscures the point. Capital punishment is morally wrong even when it is used against the right person. Full stop.

Capital punishment is not an economic question. It is not analyzed economically by criminals, law enforcement, or the American people on the street. Only Governor Ahnold weighed the political cost benefit of murdering Stanley Williams. But in the end, the public rationalle that even he used to justify sending Stanley to die was "he didn't apologize enough." Show remorse or we will kill you... beg for mercy or you die...

Dostoevsky said that "A society should be judged not by how it treats its outstanding citizens but by how it treats its criminals." Well, by that metric our society is the least moral among our peers.

In the face of disapproval and scorn from enlightened citizens of the world past and present, the American academic establishment struggles to find justification for its barbaric social practices. "2% of our population in prison?... war on crime." "Executing a reformed peace activist?... economic analysis of the penal system." Well I call foul, Stanley Williams' appeal took so long he had time to prove the case against capital punishment. The social value of human life does not end when someone commits a capital crime.

Jack Larkin

I am impressed by Posner's analysis, but I have a quibble with your disregard for the disutility of those who are anti-death penalty. While those who are pro-death penalty may be glad to be rid of murderers and rapists, the rest of us become murderers and rapists by virtue of our complicity with the state.

Chris Carbone

Shouldn't an economic analysis of the death penalty consider socio-economic disparities in its application? My concerns with the death penalty have always centered on the disparate impact of its application in relation to the ethnicity of the accused (which is well documented) and an informed supposition that the death penalty is imposed more frequently on those of lesser economic means. I was also somewhat troubled by Judge Posner's almost dismissive statements regarding the "innocence" of death row inmates and all but equating such unjustified, but certainly intentional, executions with mere accidents.

Bruce Moldovan

Isn't it undisputed that states with the death penalty do not have lower murder rates than those with the death penalty. In fact I seem to recall a study a while back that said just the opposite.

Incidentally, I love it when Posner writes a wonderfully sterile, utterly emotionless, and quite excellent economic analysis of a controversial topic (rape, abortion, drugs, capital punishment) and people get upset due to his seeming lack of emotion on the topic which, needless to say, they find quite emotional. Hopefully Posner's next treatise will contain a chapter entitled, "An Economic Analysis of the Use of Aborted Crack Baby Flesh as a Food Substitute for Starving Third World Death Camp Populations." For our general enlightenment and for my amusement at the response.


Truman Capote famously commented on the deterrence of the death penalty in a Playboy interview decades ago.
I do think the letters tend to fall into two groups.1)Those who cricize your cited studies by citing data(but not study design.)The letter by Don Dawson:"there were still 7000 murders,so it wasn't a deterrent"seems not to realize there mat have been more murders-according to Ehrlich-had there not been a death penalty.Please consider reading a crticized study,and commenting on why it's wrong.(Confession,that study has been on my pile"to read" since October).But,always remember,a study may have findings you don't like,without being wrong.
The bulk of the other letters seem to fall in the pattern exemplifed by Corey:It is wrong.(Presumably because Corey believes it ot be wrong.)In Zelaznys famous short,"The Last Defender of Camelot",Lancelot remarks to Merlin,"The burden of proof is on the presumed moralist".
Let me make my point.Asume Prof Ehrlich's study is correct,i.e. capital punishment prevents 18 killings for each person executed.Would you still be anti death ppenalty.Certainly Corey would.It seems to be a quasi religious point of view and such a view is entitled to respect.But It's like some feedback from the rest of the anti Posner group regarding this.
Now,it's time for me to go to work and make a few bucks,but I want to add a little personal information.Without going into detail,I have a lot of experience with the prison system (and a fair amount with people on Death Row)First,prisons are uncomfortable,frightening ,dirty,very often painful places.They are not comfortable.Inmates are miserable.Secondly,I only knew one person on Death Row who wanted to die.


...the death penalty for killing a police officer makes perfect sense. You tell the criminal that as bad as their present situation may be, if they kill the arresting officer they face the death penalty.That only works if they haven't already killed a police officer. Otherwise, you've just given them every incentive to fight to the death and take as many people with them as possible.Without the death penalty, criminals know that the worst that will happen if they give themselves up is life in prison but that if they fight back they may get an immediate "death penalty" by being shot to death.


Its always good to see Mr. Posner arguing that executing the innocent is unlikely and a necessary evil to get deterrence. Of course Mr. Posner as Judge Posner appears to have voted to send at least factually innocent man to his death who was later exonerted and may have voted to execute a man who, despite a plausible claim of innocence, was executed.

Of course, deterrence studies are widely reviewed as bunk and you can get what ever result you want by simply chosing a dataset that gets you the result you desire. Donohue III, John J. and Wolfers, Justin, "Uses and Abuses of Empirical Evidence in the Death Penalty Debate" . Stanford Law Review, Vol. 58, December 2005,

From the abstract ofthe Donohue & Wolfers article:

Does the death penalty save lives? A surge of recent interest in this question has yielded a series of papers that purport to show robust and precise estimates of a substantial deterrent effect of capital punishment. We assess the various approaches that have been used in this literature, testing the robustness of these inferences.. . . We conclude that existing estimates appear to reflect a small and unrepresentative sample of the estimates that arise from alternative approaches. Sampling from the broader universe of plausible approaches suggests not just reasonable doubt about whether there is any deterrent effect of the death penalty, but profound uncertainty - even about its


"Presumably because Corey believes it to be wrong."

Well, yes, me and some cross-section of society that agrees with me... oh, and most of Europe, but if we start asking Europeans for input on our social policies, then where would we be.

One doesn't have to be religious to look to moral absolutes, one merely has to recognize that cost-benefit math is inadequate to this particular task. Capital punishment measures things which are arguably unquantifiable (human life) against deterrant probabilities that are unprovable. While the Posner blog debates the merits of whatever studies that are submitted for its consideration, more human beings are being prepped for execution.

If you want to talk of studies, what of the substantial literature suggesting that certainty of punishment rather than severity is the number one correlation to deterrant effect? You can say what you want about the death penalty in America, but few would say that its application is certain. What if all of the money currently spent on death appeals was instead spent on investigation of murders? Might the increased conviction rate have more deterrant effect than the occasional socially controversial revenge killing?

If deterrance really was the prime justification for capital punishment, then proponents of capital punishment would at least entertain the above questions. However, I think the real reason people like capital punishment is vengeance. As I said over on the Becker side, go stand outside a prison while an execution is taking place. The protesters aren't holding signs that say "set an example." Their signs say "burn in hell" and "say hello to the devil Stanley"

So we can pretend that there is some rational justification for capital punishment and debate the efficiencies of various policies, or we can admit to ourselves that the primary beneficiary of state sponsored murder is the peace-of-mind of the general population.

Are blacks disproportionately killed by the state because they commit more violent crimes? Sorry... Ted Bundy, Dahmer, the Green River Killer... all white. Maybe because they are disproportionately poor and lack the resources OJ had to get themselves off. OK, but why does economic class change the nature of the offense? If the justice system was fair money wouldn't matter to innocence. Or is it because black men are disproportionately cast as criminals in every media we see. What does a prosecutor, a judge, or a jury presume when they see "unidentified black male" sitting in a courtroom?

Capital punishment compounds any disproportionate unfairness left in our judicial system by giving it permanent effect over a life.
Every year, many minority defendents get death sentences for the same crimes that whites get life for. That doesn't deter anything, rather it perpetuates racial disparity and animus.

But, we needn't even reach the question of racism in application of the death penalty, because all application of the death penalty is wrong. Widely held moral absolutes are great that way. Act such that your actions might be willed to be a universal imperative. Do unto others as you would have them do unto you.


If Posner's support for the death penalty stands or falls on its deterrent value, then he is, at bottom, a utilitarian. The ends, if they are noble, justify the means. But most humans are not so antiseptic; life and death are more than an equation. We, therefore, must struggle with this issue, while Posner finds peace in a statistic.

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