The Beauty of Law and Economics

by Don Boudreaux on April 12, 2006

in Law, Prices

Here’s my latest column in the Pittsburgh Tribune-Review — explaining how increasing the severity of criminal penalties would raise the murder rate.

Question for readers: does this logic help explain why drug dealers are violent?

Be Sociable, Share!



Add a Comment    Share Share    Print    Email


Max April 12, 2006 at 7:46 am

I have several problems with this simple statement, because sever criminal penalties can also be good, as long as they are reserved for truly evil crimes like murder, rape and alikes.

The opposite you can see in Germany, where laws on rape (f.e.) are not very harsh (7 years) and thus we have a huge amount of repetitive criminals.

However, your statement is ok, as long as we are speaking of criminal acts like drug traffiking, drug use, theft and so on. Those minor acts of crimes become more violent, when criminalized. This also a reason why the Mafia turns violent, because they have to fear even harsher penalties.

mineavatar April 12, 2006 at 8:51 am

I think this idea probably accounts for a good deal of the violence surrounding the drug trade. Although, I think much is also caused by the fact that drug dealers are unable to resort the legal system to resolve disputes, theft, etc. This leaves only the use of violence as a means of resolving conflicts that arise. I would assume the same applies to the Mafia.

Ben Litchman April 12, 2006 at 8:53 am

I'm very sympathetic to the idea of executing rapists, so that article was quite intriguing. At the very least, they should be imprisoned for life. It seems that the proper disincentives would be in place with such a penalty.

tyson April 12, 2006 at 10:44 am

I would say that the reason the dealers are so violent is because it is cost effective, by showing your propensity for violence you increase your street resume, the marginal risk is small and in fact, given that the majority of dealer don't make that much money, going to prison for a few months to a year for violence would be akin to GMU grad school. It enhances the resume. The major dealers try to avoid violence as it is bad for business, but for a younger dealer the large reward for such a small risk is worth it. So until one becomes a mjor supplier, violence is simply good advertizing.

bbartlog April 12, 2006 at 11:43 am

The sliding scale of punishments for different crimes has a number of benefits for the criminal justice system and society as a whole. I am not sure that the one you mention is very important. Theories of criminality that focus on some utilitarian calculus carried out by the criminal are somewhat lacking, since criminals tend to be even poorer approximations of rational actors than average human beings.
Anyway, here's my list:

- different severities of punishment allow society to signal the heinousness of different types of behavior effectively (not just to likely criminals, but to everyone). This maintains social norms.
- light or moderate punishments for lesser crimes reduce the cost of law enforcement. If many offenses were capital, criminals would have little reason not to resist arrest with deadly force. More SWAT teams and fewer lightly armed sheriffs would be needed.
- lighter punishments reduce the cost of convicting innocents and so reduce the need to perfect the justice system (which would appear to be asymptotically expensive).
- to the extent that prison time is a finite resource, it makes sense to allot more to crimes which we would wish to deter more and less to others less heinous.

To Ben and any other 'hang 'em high' folks here: would you support public crucifixion for murderers? Why or why not?

John April 12, 2006 at 11:50 am

If violence and drug offenses are criminalized in the same manor, meaning offenders receive the same amount of jail time, then drug dealers have no incentive to not commit violent crimes. If a drug dealer is selling drugs and thinks that his business may be more effective with violence, then that person has no disincentive to not commit violent acts.

I am not certain, but don’t most offenses like drug dealing and violent crimes just stack on each other? For example if you were caught dealing and used violence in the process, wouldn't you be charged for two crimes? Therefore, you would receive punishment for drug dealing and violence? If stacking is true, then that would destroy the argument.

Christopher April 12, 2006 at 12:21 pm

Horrid topic, but great point. I never thought about it before.

Doug Murray April 12, 2006 at 12:55 pm

John: How do you stack death sentences?

Joe April 12, 2006 at 1:35 pm

Excellent article. Once again I feel enlightened after reading something on Cafe Hayek. Thanks.

John April 12, 2006 at 1:45 pm

The stacking mainly applies to lower offenses, particularly the drug dealing example. That was Don’s question, and my response was to his question.

I understand you can’t stack death sentences. But, four years for drug dealing and four years for some violent act could potentially be stacked. And I think that happens in some jurisdictions. So with that said, the additional time for committing a violent act could be a disincentive.

happyjuggler0 April 12, 2006 at 5:23 pm

If you simply took the pro-choice position on drugs and legalized them, then drug murders drop to pretty close to zero, just like liquor store owners don't kill each other.

With the stroke of a pen we could decimate the murder rate in our (US) country. This would also free up a ton of prison space for people who commit real crimes like murder, rape, assault, theft etc.

Finally, if we didn't insist on taxing the hell out of drugs when they became legal, then property crime would also fall dramatically as junkies could suddenly afford their fixes, and prostitution would also drop dramatically for the same reason since women would no longer need to resort to it to feed their habit.

Don April 12, 2006 at 5:40 pm

Indeed an intriguing analysis, as usual, but I am not sold. One commentor pointed to the quite plausible lack of rational behavior on the part of violent criminals. To the extent that criminals are rational actors they not only have to weigh the severity of the punishment, but also the likelihood of capture, certainty of punishment and timeliness of the punishment being meted out. I would hypothesize that criminals are more likely to consider the more immediate questions of 'will I get caught' and 'will I do time' rather than consider the difference between a life sentence and a death sentence. Arguably, with the way the death penalty is meted out in this country the difference between the two is a marginal difference in terms of deterrence. I would argue that crimes that society truly wants to prevent through a disincentive structure, should have special police units and expedited judicial procedures. Picture a "Rape Task Force" at police departments and all violent rape cases jumping to the head of the line on the docket. As a father of two beautiful children, I would like to see dedicated enforcement and expedited criminal procedings as a primary tactic to combat violent sexual crimes against children and more severe punishments as a secondary priority.

Patrick R. Sullivan April 12, 2006 at 6:12 pm

There was a time when we did routinely execute people for crimes other than murder, including for rape. Caryl Chessman–The Red Light Bandit–being one of the most notorious.

So, it seems it would be an empirical question. Was the per capita murder rate 50 years ago higher, or lower than today?

colson April 12, 2006 at 6:54 pm

I wish I would have found this article about a day sooner. I just did a paper on the drug war and this would have been some good insight to insert. But I do agree with the previous commentor that it does assume some rationality amongst the criminal to justify such actions. If there were a change, and rape were equaled to murder in the eyes of the court, I think (because I don't have anything to base this on but personal assumptions) the effect would be trailing the actual change in law. Those who may be more inclined to rape(? – for lack of a better definition) would need to be exposed to the punishment either by first-hand experience or through channels of communication – and this assumes that he or she rationalizes or justifies the action of rape and murders the person afterwards because of this justification.

It is an interesting argument Don and one area that I have never really looked at in my own studies (law, crime and punishment) although I'm just starting on that area now myself.

Patrick April 12, 2006 at 8:01 pm

Your analysis is a little over the top here isn't it? It's not relavent if the sentence is harsh, light or in between. The fact is prison sentences have but only one reason, to pay for crimes. The duration, severity, conditions, location and consequences (unintended or otherwise) just don't matter. We cannot take any of that into consideration when sentencing a criminal-they committed a crime and must pay with the sentence that society deems fits the crime.

nonluddite April 12, 2006 at 8:32 pm

If rape was punishable by death, then a rapist has every incentive to kill their victim, as this makes their capture less likely and their punishment the same.

Sam April 12, 2006 at 11:32 pm

In response to John, even if they do stack there is an incentive to commit the violent crime if it is given an equal or lesser sentence than the dealing. Another basic economic principle is to invest in business projects where the payback reward is greater than or equal to the risk (I invest $100,000 in a project with a risk of failure of 20%, so I demand that I get at least $120,000 back in return if it succeeds…). So say the drugs charges net him four years and the violent charge nets him four years. He already has four years on the plate, so if he feels like if he has at worst a 50% chance of committing the violent crime to get away with both charges, logically he would do it every time. However, if you increase the jail time of the violent crime in relation to the drug charges his odds of getting away with the violent crime must be much higher for him to take the risk. I believe that's how these incentives work.

MjrMjr April 13, 2006 at 2:36 am

I wonder how many Cafe Hayek readers have any firsthand experience with drug dealers or buying drugs? That was a great column, btw, Don. I initially didn't agree with your conclusion but the way you fleshed it out has me convinced.

I don't think the same logic applies to drug dealers. Rape is violent, nonconsensual act. A drug deal is a mutually beneficial trade. There is no victim. While a rape victim has a great interest in seeing her attacker caught, neither the drug buyer or seller have any incentive to report the crime. The dealer wants the customer to come back and the customer wants the dealer to be there the next time he wants some drugs. If either party rips the other off then violence could occur but only because there's no legal enforcement mechanisms in a drug transaction.

johngaltline April 13, 2006 at 8:03 am

I've long wondered if the penalties for child rape are already too high for this very reason. Way, way too often, the victim turns up dead for no good reason that I can see.

This has also made me wonder what sort of crimes were committed under the rule of Draco, who believed that all infractions of the law should be punishable by death.

Of course, we could also stack the death penalty by tinkering with the method. A rapist facing death by lethal injection still has some incentives to avoid death by slow torture if you reserve the latter as a punishment for "aggravated circumstances."

johngaltline April 13, 2006 at 8:14 am

p.s. — Sorry for the duplicate pings. It re-pinged when I updated with a link to your article.

bbartlog April 13, 2006 at 8:29 am

Yes, this was the method employed by the Romans, who had various extra-gruesome methods of execution for those situations where mere death didn't seem to fit the bill. Hence my question about crucifixion above (which BTW was not the worst method according to the Roman scale).

beeper April 13, 2006 at 10:45 am

Good insight. But wouldn't there be something akin to "income" and "substitution" effects?

Sure, more the proportion of rapists who also murder their victim would increase, but there would also be a disincentive to be a rapist.

So the overall effect on crime would be an open question.

johngaltline April 13, 2006 at 11:38 am

So do you balance the severity of rape between the reduced incidence of rapes and the increased incidence of rape victims being murdered?

What's the exchange rate? How many murders do are "acceptable" in the meaningful deterrence of rape? Good question.

LJK April 13, 2006 at 1:56 pm

Drug dealers are violent because of the nature of the drug market, particularly on the highest levels. Since there is no legal way to settle disputes, violence becomes the only answer.

Violence also becomes a form of competition in itself. The dealer, or gang, who is most willing to use violence – or take the violence a step further – is more likely to secure more (and better) territory. This might explain why the more violent gangs – specifically the Crips and the Yakuza – have chased the Mafia off their traditional trading grounds.

I can give an example from my own experience. The town I grew up in has a major drug problem. Many high school dropouts entered the drug trade. Violence was rare, and when it did happen was largely assault. In 2000, a violent, Jamaican gang entered the scene. After a few quick murders, they successfully cornered the market. Those who did not go to work for them exited the businesses quickly. Traditional competition would have been impossible. No one could safely offer anything but the Jamaican’s products at the Jamaican’s prices.
As far as I know, they are still there today – though police efforts have hindered them. The only way to compete with them would be to become violent to a level the Jamaicans would not be willing to engage in, and nobody is willing to go that far.

Bill April 13, 2006 at 2:01 pm

The rational decision maker model seems to leave much to be desired when it is applied to criminal behavior. My sociology of law professor seemed to completely discount it. (Not to say his opinion is any more or less valid than a law and econ perspective, but it shows there is some argument among those who study criminal behavior)

Anyway, I guess the sociology and law camp claims the detterent effect is pretty useless because criminals generally don't know the penalties for crimes and they don't think they will get caught (0 prob. of getting caught * 10 yr sentence = 0). I disagree with such an extremem rejection of the detterent effect, because I think criminal penalties do an excellent job of dettering the rest of us from committing crimes, and deter petty criminals from committing expecially serious ones.

It would seem to me that a sociological model would be more suited when trying to perdict the behavior of repeat offenders. So often substance abuse and cultural influences (like a young man refusing to back down in a barfight situation when drunk and with people watching when he would never fight with no audience while sober) seem to confound a rational model. How do you model utility points for saving face, being generally pissed off at a specific time, etc., against some vague calculation for expectation of being convicted and expected sentence of imprisonment. I mean, aren't many if not most murders heat of passion type instances? Nobody is really thinking far enough ahead in many of those spur of the moment situations for the deterrent effect to really take grasp.

Anyway, I do agree with your main point though, and have only this to add. If rape were punishible my death, juries would likely refuse to convict anyone of anything but the most horrific instances, and would likely let some drunk date-raper walk rather than condemn him to death.

Jeff April 13, 2006 at 2:41 pm

This lends insight into a current issue being debated at the University of Virginia. There is a single sanction on honor violations, where you basically get kicked out of school if convicted. There is some debate, but generally unauthorized help on homework is on a different level than plagarism on your thesis. So our the system may reduce overall incidents of dishonesty but increase the severity of those incidents when they do occur.

Wild Pegasus April 13, 2006 at 3:10 pm

I am not certain, but don’t most offenses like drug dealing and violent crimes just stack on each other? For example if you were caught dealing and used violence in the process, wouldn't you be charged for two crimes?

No. You'd be charged with 8 or 9 crimes.

- Josh

Ben Litchman April 13, 2006 at 5:18 pm

"To Ben and any other 'hang 'em high' folks here: would you support public crucifixion for murderers? Why or why not?"

No, I wouldn't. It's both unnecessary and unconstitutional.

To Bill, regarding the rationality of criminals: Thomas Sowell uses the example of Britain and their penalties for burglary in his book "Applied Economics: Thinking Beyond Stage One". Since criminals know they won't risk being shot when attempting to rob a house (as it's essentially impossible to own a gun over there), they are much more likely to do so when the resident is at home in Britain than they are in the United States.

Ben Litchman April 14, 2006 at 1:25 pm

I discussed this article with a friend, and we realized that there is a problem with it. If you start from the premise that rapists should at least get life imprisonment, you are up against the same principle (i.e. equal sentencing for rape and murder) in a state without capital punishment. But we simply should not let rapists out early, as we do so often these days, simply to satisfy the logic of this principle.

Bill April 14, 2006 at 1:38 pm

Oops, I wasn't trying to say criminal couldn't be rational, ony that a rational decision making model is a terrible way to try to predict criminal behavior or come up with criminal sentences for the sole purpose of deterrence of others. I don't think burglary is really a heat of passion crime. Besides, most of those in-house burglaries probably happen during tea-time, when everyone knows the family will be in the dining room drinking tea and eating biscuits. The criminals in Britain also don't plan on getting caught, is my guess.

Mike Z April 14, 2006 at 7:26 pm

(I hope I didn't miss someone else's comment on this.)

One of the reasons why the death penalty is not necessarily a deterrent to murder is that it is almost always carried out decades after the fact -if at all.

The good part is that we need to be able to clear the innocent – something that seems to be happening nowadays. But the bad part is that the very bad guy figures that the worst that can happen is that he gets free room and board (not exactly 4-star) for 30 years or so.

In those few cases where guilt is certain, punishment needs to be carried out quickly. In ancient Rome, and in the Middle East today, that means right away.

Scott April 17, 2006 at 10:37 am

I'm not an economist (and don't even play one on TV!), but I'm not sure that I buy this.

For example, if the penalty for burglary is currently 3 years while the penalty for rape is 5 years, and then the penalty for burglary is increased to match the penalty for rape at 5 years, would it make sense to say that now the people who formerly would have been nothing more than petty burglars are now going to just up and decide to start committing rape, simply because the penalties are now the same?

Bill April 22, 2006 at 5:03 pm

The death penalty doesn't deter people. I think the empirical reserach has pointed that out pretty well. How many people do you honestly think say: The benefit to me of killing this guy who snitched on me/ is cheating on me/ owns this store is greater than the cost of the possibility I will be caught and imprisoned for life, or caught and put to death in twenty years, but less than the cost of the chance I will be caught and put to death very quickly?

Now, it may well make people confess more quickly when they are caught, to avoid the death penalty, which when you are caught you do think about. On the other hand, it is also used to get icconcent people to confess. Seriously, I know a guy who did did 12 years for a crime he didn't commit. DNA evidence cleared him, others aren't that lucky, the evidence was destroyed, lost, or never there to begin with.

Andrew Douris December 15, 2006 at 4:41 pm

Let's not get the economic term "rational" confused with the word's meaning among laymen. When we say rational in economic analysis we mean utility maximizing. Thus, depending on the factors that drive one's utility function, in this case the ability to continuing raping, "rationality" in this case is by no means the same as being rational to lay men.

So, everyone shouldn't get so hung up on saying they have a problem with perceiving a rapist as "rational" because you obviously just don't get it.

Previous post:

Next post: