What Was Gary Becker’s Biggest Mistake?

The econometrician Henri Theil once said “models are to be used but not to be believed.” I use the rational actor model for thinking about marginal changes but Gary Becker really believed the model. Once, at a dinner with Becker, I remarked that extreme punishment could lead to so much poverty and hatred that it could create blowback. Becker was having none of it. For every example that I raised of blowback, he responded with a demand for yet more punishment. We got into a heated argument. Jim Buchanan and Bryan Caplan approached from the other end of the table and joined in. BeckerIt was a memorable evening.

Becker isn’t here to defend himself on the particulars of that evening but you can see the idea in his great paper, Crime and Punishment: An Economic Approach. In a famous section he argues that an optimal punishment system would combine a low probability of being punished with a high level of punishment if caught:

If the supply of offenses depended only on pf—offenders were risk neutral — a reduction in p “compensated” by an equal percentage increase in f would leave unchanged pf…

..an increased probability of conviction obviously absorbs public and private resources in the form of more policemen, judges, juries, and so forth. Consequently, a “compensated” reduction in this probability obviously reduces expenditures on combating crime, and, since the expected punishment is unchanged, there is no “obvious” offsetting increase in either the amount of damages or the cost of punishments. The result can easily be continuous political pressure to keep police and other expenditures relatively low and to compensate by meting out strong punishments to those convicted.

We have now tried that experiment and it didn’t work. Beginning in the 1980s we dramatically increased the punishment for crime in the United States but we did so more by increasing sentence length than by increasing the probability of being punished. In theory, this should have reduced crime, reduced the costs of crime control and led to fewer people in prison. In practice, crime rose and then fell mostly for reasons other than imprisonment. Most spectacularly, the experiment with greater punishment led to more spending on crime control and many more people in prison.

Why did the experiment fail? Longer sentences didn’t reduce crime as much as expected because criminals aren’t good at thinking about the future; criminal types have problems forecasting and they have difficulty regulating their emotions and controlling their impulses. In the heat of the moment, the threat of future punishment vanishes from the calculus of decision. Thus, rather than deterring (much) crime, longer sentences simply filled the prisons. As if that weren’t bad enough, by exposing more people to criminal peers and by making it increasingly difficult for felons to reintegrate into civil society, longer sentences increased recidivism.

Instead of thinking about criminals as rational actors, we should think about criminals as children. In this light, consider the “Becker approach” to parenting. Punishing children is costly so to reduce that cost, ignore a child’s bad behavior most of the time but when it’s most convenient give the kid a really good spanking or put them in time out for a very long time. Of course, this approach leads to disaster–indeed, it’s precisely this approach that leads to criminality in later life.

So what is the recommended parenting approach? I don’t want to get into a debate over spanking, timeouts, and reasoning but one thing all recommendations have in common is that the consequences for inappropriate behavior should be be quick, clear, and consistent. Quick responses help not just because children have “high discount rates” (better thought of as difficulty integrating their future selves into a consistent whole but “high discount rates” will do as short hand) but even more importantly because a quick response helps children to understand the relationship between behavior and consequence. Prior to Becker there was Becaaria and in Beccarian theory, people must learn to associate crime with punishment. When responses aren’t quick, children, just like scientists, have difficulty learning cause and effect. Quick is thus one way of lowering cognitive demands and making consequences clear.

Animals can learn via conditioning but people can do much better. If you punish the child who steals cookies you get less cookie stealing but what about donuts or cake? The child who understands the why of punishment can forecast consequences in novel circumstances. Thus, consequences can also be made clear with explanation and reasoning. Finally, consistent punishment, like quick punishment, improves learning and understanding by reducing cognitive load.

Quick, clear and consistent also works in controlling crime. It’s not a coincidence that the same approach works for parenting and crime control because the problems are largely the same. Moreover, in both domains quick, clear and consistent punishment need not be severe.

In the economic theory, crime is in a criminal’s interest. Both conservatives and liberals accepted this premise. Conservatives argued that we needed more punishment to raise the cost so high that crime was no longer in a criminal’s interest. Liberals argued that we needed more jobs to raise the opportunity cost so high that crime was no longer in a criminal’s interest. But is crime always done out of interest? The rational actor model fits burglary, pick-pocketing and insider trading but lots of crime–including vandalism, arson, bar fights and many assaults–aren’t motivated by economic gain and perhaps not by any rational interest.

Here’s a simple test for whether crime is in a person’s rational interest. In the economic theory if you give people more time to think carefully about their actions you will on average get no change in crime (sometimes careful thinking will cause people to do less crime but sometimes it will cause them to do more). In the criminal as poorly-socialized-child theory, in contrast, crime is often not in a person’s interest but instead is a spur of the moment mistake. Thus, even a small opportunity to reflect and consider will result in less crime. As one counselor at a juvenile detention center put it:

20 percent of our residents are criminals, they just need to be locked up. But the other 80 percent, I always tell them – if I could give them back just ten minutes of their lives, most of them wouldn’t be here.

ThinkingProblemsCognitive behavioral therapy teaches people how to act in those 10 minutes–CBT is not quite as simple as teaching people to count to ten before lashing out but it’s similar in spirit, basically teaching people to think before acting and to revise some of their assumptions to be more appropriate to the situation. Randomized controlled trials and meta-studies demonstrate that CBT can dramatically reduce crime.

Cognitive behavioral therapy runs the risk of being labeled a soft, liberal approach but it can also be thought of as remedial parenting which should improve understanding and appreciation among conservatives. More generally, it’s important that crime policy not be forced into a single dimension running from liberal to conservative, soft to tough. Policing and prisons, for example, are often lumped together and placed on this single, soft to tough dimension when in fact the two policies are different. I favor more police on the street to make punishment more quick, clear, and consistent. I would be much happier with more police on the street, however, if that policy was combined with an end to the “war on drugs”, shorter sentences, and an end to brutal post-prison policies that exclude millions of citizens from voting, housing, and jobs.

Let’s give Becker and the rational choice theory its due. When Becker first wrote many criminologists were flat out denying that punishment deterred. As late as 1994, for example, the noted criminologist David Bayley could write:

The police do not prevent crime. This is one of the best kept secrets of modern life. Experts know it, the police know it, but the public does not know it. Yet the police pretend that they are society’s best defense against crime. This is a myth

Inspired by Becker, a large, credible, empirical literature–including my own work on police (and prisons)–has demonstrated that this is no myth, the police deter. Score one for rational choice theory. It’s a far cry, however, from police deter to twenty years in prison deters twice as much as ten years in prison. The rational choice theory was pushed beyond its limits and in so doing not only was punishment pushed too far we also lost sight of alternative policies that could reduce crime without the social disruption and injustice caused by mass incarceration.


Great post, thanks! Pretty clear example of how economists critiquing rationale choice is so much more informative/productive than non-econs doing it

Tabarrok is not an economist. He is a joke.

In that case, a very wise joke.

+1, very wise indeed. That was a great read!


Great post, Alex--not just on the content, but on your clear and compelling presentation.

+1 from me also.

My question: with misdemeanors and felonies do we have too few categories? (Leading to category errors.)

Oh, I have a second question: we were supposed to be at the start of a great shift from incarceration to ankle-tracking. How's that going?

That read was pretty abysmal.

"Tabarrok is not an economist. He is a joke."

'Real economists do esoteric math to justify government expansion.'

Why, because you don't like his conclusions?

...declared Bob, who has recently been installed as Dean of Economists.

Great post indeed

I enjoyed it too.

What is wrong with you people?

We're a mess. Maybe you should hang with a better crowd?

The better crowds are too small.

Because folks like you won't go there, they stay small.

This post was tl;dr for me but Buonanno et al. found that the US now has less crime than Europe (except for homicides), and argue that to catch up with America, Europeans should imprison more people.

Maybe because I often think in a chronic disease framework, I am a bit worried about whether the appropriate follow-up time was used for these RCT's . I see most of them detected the "dramatic" effects at 1year of follow-up, but while important, I would like to see whether after 5 or more years the effect still exists (?)

Violent crime is overwhelmingly a young man's game, so getting a 17 year old to suppress his more boneheaded and violent behaviors for a few years could be a big win.

I think the issue is efficiency. "nor does it provide an answer to the question whether such a policy would also be efficient from a cost-benefit point of view". Alex's point is we are spending quite a bit on prisons that would be better spent elsewhere.

"Better spent elsewhere" would be derived in part from value judgments. We know what Tabarrok's are.

You could "spend" it on deficit reduction, for example.

Punishment in America is a grave injustice. It's like beating a child into a bloody pulp for stealing a candy: altogether out of proportion with the crime.

"Punishment in America is a grave injustice. I"

This is a bizarre statement.

It’s like beating a child into a bloody pulp for stealing a candy: altogether out of proportion with the crime.

IIRC, just shy of 40% of all cases disposed of start their course through the system as felony charges. For all that, only 40% of all defendants convicted are remanded to prison. The mean sentence served by these felons is 30 months.

I take it you have never been shot, raped, mugged, or had a family member killed.

Larry, not everyone is driven by hateful vengeance.

It is easy to understand that victims want bad people to pay pay pay, but we're talking about sensible public policy here, not the right to vengeance.

Where's the proof? It seems out of character that the man known for asking that question would hold views about crime and punishment that lack proof. I did not know Becker (other than what was revealed in the Becker-Posner blog), but I suspect that his views about crime and punishment were based on the premise that crime is morally indifferent, and that one would engage in crime based solely on the risks and rewards. Hence, it isn't the risk of being caught that deters but the punishment - being morally indifferent to crime, it matters little if one is caught, but matters a whole lot if the punishment is harsh. I would agree that anti-social types might have this view of crime, but most people who commit crimes aren't anti-social, they just dodn't exercise control in a particular situation. Thus, I agree with Tabarrok. Up to a point. "[T]he police deter." What if the police are viewed not as a deterrent to crime, but as criminals?

"... based solely on the risks and rewards. Hence, it isn’t the risk of being caught that deters but the punishment."
If the perceived risk of being caught is low, isn't it an incentive to criminal behavior? People can die flying, but they continue doing so anyway. If/when the risk seems higher (did happen in the wake of 9/11?), they change their cost/benefit avaliation, I think.

Criminal’s decision-making process is not different from an honest person. Committing a crime is just like playing a game you can win or lose, you will decide to take the risk when you believe you’re good enough to realize the task without being caught. The justice system is there to persuade criminals from crossing the line threatening them with severe punishment but there must be a problem of information.

Criminals misjudge law & order’s ability to get and punish them and/or overestimate their own skills to get “the prize” without any costs so they decide to execute their plan. Intelligent people will act the way Becker expected them to act but like Alex said, criminals are not very good at thinking about the future.

Because bad decisions are linked to faulty use of available information or lack of good information to assist the decision, criminals knowledge of potential punishments is a good way to prevent crime but their ability to fully understand the risks of being punished are fundamental as well, I suspect must criminals, like must dumb or desperate people, are optimists.

Since only an undetermined percentage of criminals are actually caught and punished, the successful crooks feel that they've made the correct choice.

Oh come on! Everyone knows what the long-term solution is.

I have no doubt this would prevent 80% of crime.

What reduces crime is genetic relatedness, see Japan.

Doesn't Somalia have similar relatedness to Japan?

Without endorsing Jim's claim, no.

Pfft. Either way, we can find "related" but high conflict tribes. Pacific islanders in general, not to mention Japan's centuries of strife.

(Short of "crossroads nations" most are similarly "related.")

How much crime do Japanese commit when living in diverse societies such as California and Brazil?

That's a matter of culture, not genes.

Sounds like another case in which multiple categories are being fused. There are clearly crimes of passion, crimes born of mental illness, crimes that are effectively unsanctioned economic opportunities, and perhaps other categories. They cannot be treated similarly. Ultimately, there are also other considerations of justice, however, beyond prevention of crime; and those are often the bright line between Haidt's parties, which approach certain kinds of justice very differently and only meet on metrics of crimes deterred for lack of any common language.

Megan McArdle had a similar piece from a few days ago. In that case it was an idea that punishment for parole/probation violations should be certain and short. You get good compliance when you are sure to be punished for every infraction. It is a little harder to do proactively without a police state though.


You would think that this would be the case with electronic monitoring. But the results show that "certainty" simply catches more people. So you get a false positive that electronic monitoring increases parole violations.

Longer sentences did not lower crime as much as expected because no matter how bad jail is, the street is worse. There are limits to how badly the govt can punish people. Whereas, the brutality of organized crime like MS-13 is limitless. Organized crime will not merely punish you, they will also brutalize your family and friends. Even worse, the jails are corrupt and organized crime has a large presence. You can get drugs and other contraband (including weapons and guns!) in almost every jail in America. The homicide rate *inside* a Maryland jail is twice the national average homicide rate *outside* jails.

The fundamental problem I have with this analysis is that it totally misses opportunity cost. There is an implicit assumption that people are unwilling to take on danger, for money. Coal mining, logging, construction, commercial fishing - these are all very dangerous jobs, that can pay less after taxes than drug dealing. So if I grow up in the hood, with an 8th grade reading level, and there is no longer a steel mill to work at, street pharmaceutical sales might not be a bad gig. And if you work your way to the top, you can be like Tavon White, who ran the Baltimore City jail, making over 200k a year tax-free selling drugs to inmates from the comfort of a jail cell (safe from rivals), with 4 guards as concubines.

So, I would not necessarily say severe punishment does not work. But if it does not, it's because the alternatives are worse yet.

Unfortunately, "the street" is inside the prisons as well. If you can't get rid of prison gangs and rampant prison rape INSIDE prison, what does that say about government's ability to do anything anywhere?

I agree entirely. And if you can't keep drugs out of prison ... etc etc.

Good point. The street and the prison are now overlapping in many areas. The street is hell for some and heaven for others. Depending on where they are in the hierarchies of the criminal world.

If you can’t get rid of prison gangs and rampant prison rape INSIDE prison

I'm sure they could get rid of prison rape if they actually wanted to. The will just isn't there.

I doubt that many prisoners would agree that "the street" is worse than prison. However, if this assertion is at all true, I have a "win-win" solution to the problem.

"the street" is most definitely not worse than prison. "the street" is a big place with a much lower "ms-13" per square mile than the prison. In prison you are forced in to a high density version of the street, complete with an additional armed group - the guards - who, if recent history has shown us, can do anything from beating your visiting family, ripping out your eyeball, and scorching you to death in extremely hot water without any punishment whatsoever.


1.If the fall in crime which followed increased mandatory minimum sentencing was the result of a decrease in lead exposure or an increase in abortions (both contentions resting, at least in my opinion, on ludicrously weak evidence), what accounts for the recent increase in crime now that we are moving away from those policies?

Agreed. Jesus Christ is Tabarrok prone to making popular, but completely unsubstantiated arguments. The idea that crime fell "mostly" due to abortion and reduction of leaded gasoline use is so laughable and totally unsupported by any semblance of fact that I can't believe I'm reading a real economist. Sailer solidly rebutted the abortion-crime idea, and the leaded gasoline idea is just looking at two graphs and matching them up- is there any actual evidence that exposing children to lead increases their risk of becoming criminals later, all other factors adjusted for? And, yes, the imprisonment boom of the 1980s did lead to a strong spike in arrest rates. The U.S. arrest rate only began to decline in the late 1990s. Saying the fall in crime was "mostly" due to a high arrest rate would be better supported than what Tabarrok is claiming here.

He doesn't like cops. Ergo, he's repelled by anything which suggests cops do useful work. Here's one bit of wisdom here


And here's the antidote:


"is there any actual evidence that exposing children to lead increases their risk of becoming criminals later"

Did you even bother to look? It's basically common knowledge that lead exposure reduces impulse control and makes a child end up being more likely to engage in violent behavior. Saying that the crime correlations are just lining up two graphs shows that you haven't even bothered to read about the subject. Why do people make comments that express their own ignorance as an argument?

The notion was peddled by Kevin Drum, who is not a sociologist or public health maven, just an opinion journalist. If you wanted a demonstration of the point, you'd need an ecological or panel study, not a longitudinal study.

While we're at it, until recently the lead-smelting capital of America was Herculaeneum, Mo. The index crime rate there is less than half the national average and the homicide rate is nil.

Lead is very much responsible for crime. Kevin Drum might be an opinion journalist, but Rick Nevin is a serious dude:


It is clear: Lead poisoning is the answer.

Crime is actually increasing among older people, due to blood lead levels, while falling in the youngest:


Explains NYC crime drop by lead:


So Tabarrok's claim that the experiment "failed" is totally unsupported. He's just pulling things out of his ass and expecting us to eat it all up.

"longer sentences increased recidivism."

-Is there any evidence for this whatsoever? If so, why doesn't Tabarrok link to it?

I am shocked that Alex would take at face value the purely correlational study on parenting styles and adult violence. I don't see anything more than rudimentary controls. Even the summary itself says that "causal effects cannot be assumed." In a normal economics seminar, I'm sure that Tabarrok would be the first to point out that greater punishment of children may be endogenous to kids that are less well behaved, hence standard selection bias.

I do agree that quick and clear works, but I'm not so sure why he uses inconclusive evidence against harsh punishments for kids to bolster his case.

Thanks! Didn't even notice that.

Let's give Gary Becker his due; he wasn't arguing with a bunch of libertarians or even classical liberals, he was arguing with the unhinged criminology that was prevalent at the time. They would have been successful at defining crime as a social construct if they could have managed to get home without being mugged.

And a very important point. There are rational actors in this equation. The justice system is purportedly about doing something about crime and criminals, and sometimes it does. But it is mostly about the very interested actors in the system who have well defined pecuniary interests in how the system works. A crime is committed, evidence gathered, a case put together, a criminal charged and a trial held before a jury of their peers. It should in the vast majority of cases be able to turn around in two weeks. A 'good' lawyer will make it last two years. A 'fair' judge will throw away 3/4 of the evidence. An 'effective' prosecutor will add 5 other charges to the breaking and entering, corner the accused and get a plea bargain for something and on to the next on his/her way to a 95% conviction rate. A powerful police union protects and punishes the neighborhood for the sloppiness or plain criminality of the police officer who gathers evidence illegally or just plain lies and fabricates. The incarceration is done by powerful guard unions and contractors providing an expensive service.

So the studies and understandings of the criminal mind is of interest and important, but how the justice system actually works is far more important and probably far more difficult to amend than some twit who in a drunken stupor does something evil.

I'm all for giving Becker his due, he was brilliant. But Alex describes a situation when he was quite literally talking to a bunch of libertarians.

haha all right, good point.

We live in a world of bounded rationality. People have understood this, like forever.

To set one straw-man against another (crime as a social construct vs crime as a rational act) IMO doesn't advance understanding, it apologizes.

I disagree with some points, but on think you're on the money with the general thrust of your argument.

How could a "rational actor" possibly take into account all these additional uncertainties?


2. I note that at the beginning of your post you flatly state that "the experiment . . . didn't work" but at the end you note merely that increased sentencing did not reduce crime as much as expected and give the particular example that doubling a sentence from ten to twenty years did not increase deterrence by exactly the same ratio. Do you really contend that "the experiment" did not work at all or do you wish to retract or alter that initial comment? Did Becker or anybody else ever claim that there would be a mathematical identity between the amount by which a sentence would be increased and the increase in amount of "deterrence" produced thereby? How do we measure the amount of "deterrence" produced anyway? Don't you want to re-think this whole thing? Define your terms a little more rigorously? Pull your head out of your priors?

Yes, he did claim a mathematical identity as per the quotation


No. Properly understood, the quote merely implies, at most, that (given the same risk of a criminal conviction) an increase in punishment should produce an increase in deterrence. In this post, Tabarrok contends that unless the mathematical ratio between increased punishment (stated in years of confinement) and increased "deterrence" (stated in unknown terms) is a mathematical identity, the relationship Becker suggests between punishment and "deterrence" is disproven. Of course, I suspect that Tabarrok does not actually believe that and probably would not assent to the argument if the underlying logic when is stated explicitly. However, he wants to trash Becker's analysis and evidently lacks better arguments to support his claims. Thus, we have "twenty years in prison deters twice as much as ten years in prison" as punch line for the entire post. A "far cry" indeed. But did Becker ever say this? Did anybody?

As I grew up in the 50s and 60s when locking doors happened only when leaving for an extended time, and even cabins were often left unlocked based on the belief that if unlocked no one would break in to trash the place and would find nothing of value to take unless they were in dire need. And kids played until dark out in the woods or seven streets over far out of sight of parents.

Now the idea of letting brother and sister walk a couple blocks in their neighborhood to the subway to travel to the public library, or not locking one's door whether in or out is a serious risk of life and property after decades of very harsh war on crime zero tolerance punishment seems to prove that Becker's policies were total failures.

No to mention women forced to work based on harsh treatment of those on welfare to deter poverty has led to women going to jail for child endangerment because they left their kids outside in the car sleeping while working the nightshift in sight of the car or interviewing for a job while her kids are doing homework within sight in the food court. Or jailing the poor for failing to pay fines levied to raise revenue based on the broken windows theory of ticketing anything in public to keep people from being on foot in public.

Why the jailing?

Because the poor were not getting the message that its a crime to be poor, and its necessary to up the punishment.

"Now the idea of letting brother and sister walk a couple blocks in their neighborhood to the subway to travel to the public library, or not locking one’s door whether in or out is a serious risk of life and property after decades of very harsh war on crime zero tolerance punishment seems to prove that Becker’s policies were total failures."

-Note to mulp: public opinion is not a substitute for actual crime trends (increase in the 1960s and 1970s, collapse during 1990s).

You know that job interview Black fat woman cheated her supporters to go on a shopping spree, right, mulp? Typical criminal. Do you have any evidence against the effectiveness of fines? Seems to have worked pretty well at improving driving behavior in Russia.

Incentives against being poor discourage poverty. That's a good thing.

No collapse. There was a 50% decline in homicide rates over a 25 year period and a 30% decline in composite crime indices. You did see a collapse in New York City and a few other loci under the influence of people trained in the NYC police department.

Yeah, but over the period 1993-2000 the speed of the decline was stronger than over the period 1993-2015. What do you define as a collapse, anyway?

A 'collapse' would be the 81% decline in the homicide rate registered in the City of New York. In the rest of the country, what you saw was a return to the homicide rates which prevailed ca. 1960 (conjoined to higher rates of lesser crimes than was the case then, though lower rates than you saw in 1980).

Collapse, gradual decline, whatever you want to call it - the streets are safe again, as safe as in 1960, outside of a few shrinking ghettos. This has taken place in practically all cities, not just "designer" cities like Washington and San Francisco. The only areas where crime is increasing is where expanding Hispanic barrios, full of young men, interface with essentially zero-crime European American neighborhoods with very aged populations. Even there, the situation is not that bad. It's only bad in the ghettos. It's hard to imagine that mass incarceration has *nothing* to do with this improvement - it's hard to commit a crime against the public when you're in jail.

This has taken place in practically all cities,

It most certainly has not, and an examination of CityData would tell you where it did not (e.g. Baltimore and Detroit). New York cities are a case in point. In New York City, the homicide rate ca. 1990 was fluctuating around a set point of 27 per 100,000. That in Rochester was fluctuating around 20 per 100,000. Around 2010, the corresponding figures for these two cities were 5 per 100,000 (for NYC) and...20 per 100,000 (for Rochester). One place has an 81% decline in the homicide rate and the other has no decline.

Pure stream of consciousness.

Too many guns is also part of the problem.

(Cue irrational defenses of in-vogue interpretations of the 2nd amendment.)

"In vogue"? Those interpretations are perfectly congruent with the plain meaning of that piece of constitutional text. It requires twisting your head into a pretzel to sign on to what Saul Cornell is pushing.

First, I'm not American, so you don't have to take what I say seriously on this. Like most of the entire world, I really don't get the fascination of many Americans with guns. The "in vogue" bit was plain baiting ... :)

But here's the for real part. When I look at the text and consider American history (early times), it really seems to me like they were short on professional army and there were a few too many native Indians about, and they wanted to ensure that every locality had the ability to raise a self-defense militia against natives or other potentially invading forces.

But the argumentations which promote gun culture as safer, well, for those who are not regularly exposed to NRA marketing or who do not belong to communities where NRA-promoted slogans, etc. are "in vogue", it all just seems like crazy talk driven by slogans designed by people who simply want to make money selling guns.

I don't think the founding fathers quite intended it to be this way. My interpretation (again, an "ignorant" foreigner) is that the 2nd amendment was designed to enable local governments to raise a militia if/when needed. The right to carry concealed arms to Walmart on a Sunday shootout? I really think they would be hitting their heads against the wall if they could see how that part of the constitution is interpreted these days.

I've added my additional 2c, but I think most sides of the argument are altogether predictable. Perhaps you feel it is necessary to present some of the FOR arguments?

Did Becker just die or something? Why the past tense?

Gary Stanley Becker (December 2, 1930 – May 3, 2014)

the noted criminologist David Bayley could write: The police do not prevent crime. This is one of the best kept secrets of modern life. Experts know it, the police know it, but the public does not know it. Yet the police pretend that they are society’s best defense against crime. This is a myth

Of course, James Q Wilson was advocating celerity of punishment a generation ago. Notably, Wilson was the issue of political science faculties and not sociology faculties. Maybe the real problem is that wide swaths of sociology amount to gold-plated group think.

Interesting post.

But consider a set of individuals who (1) are cognitively limited, (2) have poor impulse control, (3) are prone to violence, and (4) tend to emotionality. I just described the people in jail.

I doubt the Becker approach would work for this population. I also doubt that the liberal rehabilitation approach would work. For some, they are like this for genetics, for others it may be early childhood environment.

The only approach that would work and has worked is to incarcerate these people so they do no more harm.

For this population, drug and alcohol abuse is commonplace, part of the syndrome. Legalizing drugs may get rid of the cartels, but if it leads to more drug abuse by defective people, we may get more violent local crime.

Finally, social scientists tend to be secular-minded liberals. The vocabulary of the soul and spirit, of salvation, of good and evil is absent. We have free will, and some people just choose evil. Psychotherapy will not solve the problem of evil.

It would probably be cheaper to provide drug addicts with housing in special communities that lock the exterior gates before dark and provide 3 meals a day and drugs. That this isn't even a real consideration demonstrates that what you describe as effective criminal justice is something less than a rational approach: it's emotional for some and profitable for others. Sounds a lot like crime itself, and it probably does compose a crime in itself.

It's 'emotional' when punishment is designed to punish. Thanks for the education.

(Not an accusation but....)

Do you beat your child because you're angry? Is so, then it's unambiguously wrong. Did you spank him because you carefully thought out the pros and cons and determined that physical punishment was likely to be effective in "teaching" children the difference between right and wrong? If so, then you can very well argue your case.

Punishment for the sake of punishment, along the lines of "eye for an eye, tooth for a tooth", is an emotional response drive by a desire for retribution, revenge, etc. It should have little or no place in effective public policy.

Punishment as a deterrent can be defended intellectually. Punishment as a means to "teach" about the difference between right and wrong can be intellectually defended.

But I think a lot of people who would support doubling a sentence from ten years to twenty years are motivated by thinking more along the lines of "make that bad bad man paaay, and make him pay BIG", regardless of whether 20 years is altogether out of proportion to the crime.

Punishment for the sake of punishment is like beating your child because YOU are angry. It should have no place in sensible public policy.

Thanks for your input.

"consider a set of individuals who (1) are cognitively limited, (2) have poor impulse control, (3) are prone to violence, and (4) tend to emotionality."

Geez, I thought you were describing all 4 year olds, 8 year old half the time, 12 year olds 20% of the time, 15 year olds 10% of the time, 17 year olds 5% of the time.

But it seems only when black, disabled, brown, do kids get put in jail when they act that way in public school or on the street under broken windows policing.....

that could reduce crime without the social disruption and injustice caused by mass incarceration.

There is no 'mass incarceration'. The majority of criminal cases are disposed of without remanding the defendant to the state prison and the mean sentence served in state prisons is 30 months (while the mean time served in county jails is measured in weeks). People who complain about 'mass incarceration' are really just complaining about incarceration.

New York has been highly successful in advancing public safety. The prison census in New York grew at a slower rate than the prison census nationally. It still trebled. The notion you can accomplish much without actually punishing anyone is fanciful.

No, the complaint is that, at any given time, over 1% of the adult male population of the U.S. is in prison or jail. You are perfectly free to argue that it's a jolly good thing that America is doing this but you can't deny it is a statistical outlier in international terms. Countries as diverse as Russia, Singapore and Cuba imprison fewer of their citizens than the U.S. does.

but you can’t deny it is a statistical outlier in international terms

Something only of interest to the stupidly other directed. And the correct figure is 0.75%.

Kind of ironic right? "land of the free" my ass...

You commit crimes, you lose your freedom, as it happens for a mean of 30 months if you're one of the 40% or so of convicts actually remanded to a state or federal prison. You want to be free of incarceration, stay on the right side of the law.

Since only an undetermined percentage of criminals are actually caught and punished, the successful crooks feel that they've made the correct choice.

An American cannot live for 24 hours without committing some kind of crime, a crime that's encoded at some level of government. Maybe everybody should get to serve a stretch in the pen periodically, like the obligation of jury duty, to serve as punishment for their undetected crimes.

An American cannot live for 24 hours without committing some kind of crime,

Cannot figure if you drank the libertard Kool-Aid or the William Ryan Kool-Aid.

The United States of massively over-incarcerated.

Almost the entire Western world outside of Republican America agrees on this point.

The United States of massively over-incarcerated.

No it is not, for reasons stated above.

Almost the entire Western world outside of Republican America agrees on this point.

This is a matter of no interest to any serious person.

"the mean sentence served in state prisons is 30 months"

Two and a half years is an insanely long amount of time to be imprisoned for, yet you're presumably citing it as evidence of what a soft touch the justice system is re: imprisonment. How curious.

It's 'insanely long' to be imprisoned for what?

When I had to know these things (ca. 1985), fully 38% of the cases disposed of by the New York State courts started their course through the system as felonies. Of course, a great many were plea-bargained down for the reasons that goes on, but that should tell you that prosecutors are very commonly dealing with serious crimes even if most of their caseload is run-of-the-mill. The 40% or so who are remanded to prison in this country likely do include some dealt with in a draconian fashion, but the vast majority are either recidivists or people who did sorely condemnable things.

What about public transportation in Europe? No turnstiles, but a low probability that an inspector will ask you for your ticket. If you do not have one, a significant fine. This system seems to work well.

Shouldn't we check the data before deciding whether it works well or not? If inspectors conduct truly random inspections for a month or so, they can get an estimate of what percentage of riders are fare evaders and how much money the system is losing.

However, the case for fines is stronger than the case for long prison sentences since fines at least raise money but with two caveats. First, some people can't afford to pay large fines (I believe Alex has written on this subject) and so simply enter the county jail system at great expense to taxpayers. Second, in the U.S., the 8th Amendment bans "excessive fines" but not, as the Supreme Court has noted, "excessive imprisonment." The reasoning was probably that you don't want the government becoming dependent on crime to fund itself but it does risk creating a bias toward imprisonment as the one-size-fits-all criminal punishment.

" the 8th Amendment bans “excessive fines” but not, as the Supreme Court has noted, “excessive imprisonment.” "

This could be the seed of some interesting constitutional argumentation on the matter.

Ignoring (ghetto) economic factors in crime for a moment -- for pure (works everywhere) psychology:
Boys are in the emotionally dependent state until they are eighteen and a half years old -- this seems to run full steam to the very end and turn off over a week's time. My limited Bronx observations anywhay.

The crux of juvenile delinquency seems to be that IF the kids believe nobody cares about them (WRONG ABOUT HALF THE TIME); they literally wont care about themselves -- thus, no penalty can deter them.

Quick cure? Five or six weeks of intensive attention -- you are building a SUPPORTIVE relationship -- the crime doesn't slow down an iota until the very end. One day a different personality wakes up: the invasion off the body snatchers. Crazy experience.

(Speculation: I don't have too much experience with this.) Over eighteen and a half they may have become I-don't-want-to-workaholics from years of not going to school or work. I used to say back in the seventies when jobs paid more -- that a probation officer had to hold their feet to the fire: go to work today or go to jail today. they appreciate work money more than easy-come-easy-go money.

But that was before the minimum wage dropped several dollars below LBJ's $11 (as per capita income bloomed) -- and before unions disappeared. William Julius Wilson could conceivably change the name of his book When Work Disappears : The World of the New Urban Poor to When Pay Disappears or more descriptively to When Unions Disappear.

Which brings us to today's big American crime story (not the story of high union density economies): all pervasive, drug dealing gangs In Chicago 100,000 out of my guesstimate 200,000 gang age minority males "work" in street gangs.

Cure? First let's take a look at another cohort that is no longer in the legitimate job game because of lower and ever lower wages for the same grueling work: guys like me; American born taxi drivers (NYC, Chi, SF).

In Chicago the fare is now half a dollar a mile below 1981 (one 30 cent mile increase between 1981 and 1997!); since 1991 built subways to both airports/opened up unlimited livery/added 40% more taxis. A $15 minimum wage would force an immediate dollar a mile jump in the meter. We old guys aren't in drug gangs -- we are mostly on Social Security. I tried SF before my back went out in 2004 -- since which SF has doubled its number of taxis. Don't know what the new would-have-been American born cabbies are doing.

Cure (for gangs): labor laws are all in place; the issues presumably have been settled. But the American (not any other modern economy's) labor market is the only one where the rules simply aren't ENFORCEABLE. Simple/one-step solution across the country (spreading state to state) make union busting a felony. Make muscling one of the three actors (ownership, labor, consumer) in what should be the free market no longer playable by making it a crime. RICO prosecution auto-invoked for ongoing malfeasance.

As the twig is bent so grows the tree -- will today's gangees change their stripes. But every new generation wants to grow up straight. And lots of the present wanderers off course would welcome good paying jobs -- meaning paying good for the jobs that are already out there in the uniquely labor-powerless American market. Get me Jimmy Hoffa -- not private jails.

Your solution is 100% ass-backwards:


Homicides collapsed during the recession&aftermath.

Or we could just legalize drugs, in so doing destroy most of the business model of organized crime, and plough some small share of the taxable proceeds into dealing with the small portion of drug users of have addiction problems.

Organized crime antedated street drugs and the drug trade never meshed well with the codes observed in Mafia families (which is why mafiosi like Paul Vario banned the drug trade in their gangs).

Proscription on the trade in intoxicants was enacted for a reason, and enacted in a society wherein people were much more under the control of their relatives and in which there was scarcely any recourse to aught outside the labor market.

I'm sorry you feel you have to drive a Chicago taxi but the market is telling you that you're in the wrong business. Customers know that (except for you, of course) a ride in a Chicago taxi means they will be driven in a filthy car by an angry, reckless driver who doesn't speak English and hasn't heard of the airport. The market price for such a service, even if there are exceptions and some drivers are good, should be zero but it isn't. Try driving a limousine or Uber.

If I had a magic wand and could wave it over criminal justice in America this is the waving I'd be doing.

Someone did a study on paroles that came to the same conclusion - come down on minor infractions of the rules as quick as is possible but not like a ton of bricks - in the case I read about the person who broke the rules would end up going back to jail for the weekend. As this was effective.

This line of argument also points to just how ineffective the death penalty is. Maximum punishment that is dolled out a maximum length of time after the crime and for a capricious number of crimes. The time it takes from getting the sentence to having it carried out is of course totally rational - it's the ultimate punishment so it requires years and years of reviews.

I think we should punish people much faster then we do but not so harshly. The harshness in this case also includes the way that a criminal record follows people for a lifetime after the crime. It really should be - commit a crime, start doing a reasonable amount of time as quickly as justice permits and then be able to move on once you're in incarcerated.

Instead it's - commit a crime, nothing happens. commit a crime, nothing happens. commit a crime - wait for 6 months until you're actually sent to prison for 20 years. Get out and then be unable to rejoin society because no one will ever trust you again. Frankly this sounds like the way a dysfunctional marriage works.

We had a dramatic decrease in the safety of Americans but now some people want to return to the high crime era. Weird.

oops, "decrease" should have been "increase".

We had a decrease in crime but an increase in safety.

Within the last decade many states facing budget constraints such as New York and California have cut back dramatically on their incarceration rates and have continued to see a decrease in crime.

No, they have not. There was a 10% fall in the prison census in New York. It was driven by how many people were in the intake pipe, not budgetary considerations.

But violent crimes have dropped to historic lows. Maybe the high level of incarceration is due to a confounding factor - the drug war.

(Not necessarily agreeing with Becker, just saying it's not at all clear that it didn't work).

We should seriously widen our options for punishment. In particular I would explore corporal punishment more. It's cheap and easy to administer and need not do any lasting physical harm, and seems like it would address the points about criminal psychology that Alex mentioned in the post—specifically that it would use pain and—more importantly—humiliation right now instead of long-term incarceration in the hard-to-fathom future to deter criminal behavior. There are those who would think that inhumane but, it has to be compared to what we are doing now, the one-size-fits all lock-em-up approach, and as for me a whipping would have to be pretty severe before I preferred a year in jail to it.

But it doesn't matter whether this is a great idea or not because of the "r"-word.

We seldom incarcerate people for the sort of crimes you'd use corporal punishment. Corporal punishment would be replacing social worky tripe like probation and community service. The social workers won't like it, so lets do it. Caning, strokes of the birch, and, especially, the pillory-and-stocks.

@ laderff ISIS agrees with you.... but I am not sure this is the way to go ;)

Yeah, and Hitler tried to curb smoking.

I'm not sure I LIKE it, but a severe beating is nothing compared to years in prison.

The strange thing is that random, instant, and harsh punishment absolutely does change the behaviors of children.

This is the bee theory of raising children. Most children eventually get stung by a bee, and they instantly become forever highly aware of bees and not to mess with them.

So why do people think that random, instant, and harsh punishment won't work? We have a perfect example of how such punishments do work.

Sorry, the example is bees?


Also, my toddlers are both terrified of bees but have never been stung. Learned behavior.

Random doesn't work. Bee stings are not random.

The solution is to throw bee hives at criminals.

Instant, painful punishment. Lesson learned. No need for prison time.

You almost answered your own question. Instant, harsh punishment convinces the child who is subjected to it to change his or her behavior. I don't think you will get any disagreement from Prof. T by extrapolating this to adult criminals. But the point is that the probability of getting caught is low enough that lots of criminals get away with a lot of crime before they get caught. And they are not necessarily deterred by the punishments that other criminals suffer.

Thanks for this post. I am interested in what Gary Becker would have said about these critiques. Russ Roberts has brought up Becker’s strange theory of the optimum punishment in two Econtalks (one with Edward Lazear and another one with MeganMcArdle). I was left thinking, “This ‘higher probability of punishment deters crime and harsher punishments don’t’ result is empirically pretty solid. How in the world would Becker stick to his guns?”

Maybe Becker studied ancient history, with Babylonian--or was it Asyrian--death penalties attached to giving bad haircuts, and the laws of Draco?

That was my understanding of the matter.

But I think he approaches the question of "optimal" differently. Because high probability of catching criminals costs lots of money, so he prefers draconian punishment for the few who are caught.

I think it is grossly unfair to "make an example" of the few percent who get caught, by giving them draconian sentences far out of proportion with the crime. Better to spend more money to increase the probability of getting caught. Perhaps not "economically optimal", but far more "ethical", so long as policing resources to not unfairly target any specific group.

Point one: I think mental illness, drug additions and conditions like Fetal Alcohol Spectrum Disorder are big contributors to some crimes. Harsh punishment is not likely to help with those folks.

Point two: Many people fear serial killers (they are the bogeyman hiding under our beds). Some killers are highly transient. Some like Ted Bundy are intelligent and sophisticated. If the death penalty is a deterrent then the data should show killers "commuting" from death penalty states to non-death penalty states to commit their crimes.

Very good point #1. I believe there is strong empirical evidence for this (re: mental illness etc.) Punishment might indeed not help. But there's still an argument for keeping some of them -- if they are criminal of course -- off the streets.

Are these people less capable of learning than rats and pigeons?

I have first hand experience with some one with Fetal Alcohol Spectrum Disorder and she has a lot of trouble with judgment and consequences and was in and out of detention as a juvenile. Between fifteen and eighteen she spent about one third of her time incarcerated (violence, mischief, breach of probation).

So far she has managed to stay out of prison as an adult but she has received intensive support from social welfare agencies (who seized her from her drug addict mother when she was two) and relatives. But for the support she has gotten she would probably be in prison or dead by now.

I have some cousins with FASD. Of the 4 of them 3 are incarcerated, 2 are women who have lost their children. The one not in prison is a firefighter. Mildly ironic as he went through intense institutional therapy after being caught lighting fires as a kid. He was also the only one who had a near normal level of intelligence.

The boy in jail is the least likely to be able to consider consequences. He was very very susceptible to peer pressure as it played on his fears of appearing different and stupid. The 2 girls were involved with drugs and generally chaotic, one had been sexually abused beginning at age 2. I really don't know what could have been done for them. They were doing fine until they were taken out of their foster home and put in a group home as teens. I'm glad that one of them were able to get out of that cycle of decline.

These kids needed structure and an environment that could keep them in control. The funding necessary to have kept these kids from harming themselves and others was almost certainly not politically feasible.

I don't believe that the death penalty is a stronger deterrent than life in prison.

Think about it. Would you rather spend life in prison or be dead?

Dylann Roof would rather spend life in prison, as he has offered to plead guilty to the thing he is obviously guilty of to avoid a trial and a death sentence.

Think about it. Would you rather spend life in prison or be dead?
Gary Gilmore was the 1st person executed in the United States after the Supreme Court declined Justice Brennan's suggestion that we all pretend that a practice explicitly regulated in the constitutional text was somehow in contravention of that text. There was a reason he was the first: he did not care to appeal his case up the wazoo. The median lapse of time between conviction and execution in these cases is north of 9 years and there have been cases the appellate courts chewed over in repeated episodes over as long as 25 years. Maybe it's just the har-de-har public interest bar proffering their services and making nuisances of themselves, but some people seem awfully intent on filing briefs if they should prefer to be euthanized like dogs.

Great post on a fascinating topic.

How do you build a rational model for the irrational?

The strongest correlation to crime seems to be IQ. I suspect over the next 100 years crime will once again fall to lows unthinkable today, as various forms of intelligence augmentation become more prevalent.

In this case "IQ" is a proxy for no "mental illness, drug additions and conditions like Fetal Alcohol Spectrum Disorder"

Do higher IQ people have less mental illness and drug addiction?

Regardless of your IQ before, am am pretty sure you will show lower after.

I see you bought into Reefer Madness.

I wonder why you went there rather than say meth related decline.

Come on, John. You really can't think of any other mechanisms by which lower intelligence might lead to criminality?

Of course, but I think we should acknowledge that IQ in this situation is not "all else being equal." It is not a twin study. It is a post hoc measure of IQ for those with the very worst life stories.

What about age and sex? Those, especially when considered together, are excellent predictors, and yet they are pretty much independent of IQ. I think it's probably IQ, time preference, and testosterone/aggression.

Whatever your IQ, the trick is to operate within its bounds. People with any IQ crash and burn, when they think they are a standard deviation higher.

Are those really likely to become criminals, though? I think they're more likely to get failing grades in advanced classes.

All criminals are obviously unintelligent. They just accidentally come up with involved and complicated scams, robbery and burglary scenarios, financial frauds, etc. maybe by watching television crime shows and movies. Perhaps those instructional entertainments should be banned, like guns.

chuck, is this parody? No one can tell on the Internet.

As I say above, I think our two categories (misdemeanors and felonies) are too broad. Who becomes a criminal? "More Than 30 Percent Of Americans Arrested By Age 23, Study Says" http://www.npr.org/sections/thetwo-way/2011/12/19/143947345/more-than-30-percent-of-americans-arrested-by-age-23-study-says

Inspired by Becker, a large, credible, empirical literature–including my own work on police (and prisons)–has demonstrated that this is no myth, the police deter.

That's perfectly commonsensical, of course. The sociologists who were telling you otherwise should have been out the door if anyone in academe was ever accountable for anything other than running afoul of office politics.

Probably common sense is right on this one. But you should always be prepared to let evidence trump common sense, so long as you are on the lookout for poorly specified models, data problems, etc.

Common sense is often proven wrong.

Common sense is seldom proven wrong, particularly with regard to palpable things people encounter every day. There are people who will never forgive Messrs. Giuliani, Bratton, and Wilson for accomplishing something, because it made them look like chumps. They'd spent their careers advancing notions Joe Blow off the sidewalks of Brooklyn would likely never endorse and then discover after decades of talking down to everyone that they were bloody wrong.

In England in the beginning of the 19th century, many relatively minor offenses were punishable by execution. The result was that the juries often opted not to convict, even when the perpetrator was obviously guilty, because they couldn’t opt for a lesser sentence. Banks actually lobbied to *decrease* the punishment for forged bank-notes because the law against such forgeries was effectively going unenforced. This is described in some detail in the book Jury Nullification. I think this is an example of what Tabarrok means by “blowback.” You can’t just make the punishment arbitrarily severe, because people end up subverting the law. The probability of being punished may well drop to near zero. Recently judges and prosecutors have been subverting this feedback-loop by ensuring the jury members don’t know the harshness of the punishment, particularly for drug crimes.
See the paper “Jury Nullification: The Contours of a Controversy.” A google search should find an ungated PDF. Some commentary about the bank-notes is in a footnote on page 72.

The share of cases actually resolved with petty jury trials is in the single digits. The share of the prison census there on a bill where a drug charge was the top count is just north of 20%. This may be a problem worth correcting, but recall that it amounts to fewer than 3% of cases processed.

What percentage of incarcerations are due to police stings? It's OK, and evidently legal, for cops to carry around illegal drugs to sell to naive college students, when they aren't using them themselves. A huge percentage of the Muslims arrested on weapons charges and assistance to terrorism have responded to fabricated cop crime scenarios. In most cases there would have been no crime if the cops hadn't initiated one.

What percentage of incarcerations are due to police stings?

Very few.

My understanding is that, beginning in the 1980s, we did indeed increase the probability of being published. Anybody have numbers on this?

By which of course I mean punished......


Great post. About this passage;

"[...] consider the 'Becker approach' to parenting. Punishing children is costly so to reduce that cost, ignore a child’s bad behavior most of the time but when it’s most convenient give the kid a really good spanking or put them in time out for a very long time. Of course, this approach leads to disaster–indeed, it’s precisely this approach that leads to criminality in later life"

Interesting. I just finished Steven Pinker's "The Blank Slate" and he's pretty insistent that, within limits, parenting styles don't have much effect on children's life outcomes and adult personalities.

You are correct. The literature on spanking is like all literature that claims parenting matters. They simply find a correlation and ignore the overwhelming evidence from behavioral genetics showing parenting has no influence on adult outcomes.

It's true that children who get hit are more likely to grow up to become criminals but it's because

A) They inherit genes from parents who were more prone to hit them
B) Kids who are prone to bad behavior are more likely to provoke physical punishment.

hey simply find a correlation and ignore the overwhelming evidence from behavioral genetics showing parenting has no influence on adult outcomes.

"Overwhelming evidence" would be what ordinary people would call 'fad notion of the last 15 years which contravenes common sense but serves to aggrandize the position of fad pushers".

"Common sense" can tell you what part of parent-child resemblance is due to genes and how much it's due to upbringing? Please explain.

'Commen sense' tells me that facial features and behavior and values are not identified. Common sense also tells me that when an internet chatterer refers to 'overwhelming evidence' regarding some contested intersection of anthropology, sociology, and psychology, that he has never attempted to assemble a bibliography of any kind.

Art Deco, have you read The Nurture Assumption? And do you have any actual evidence for any parenting intervention that has any influence on adult outcomes? Dismissal is a poor substitute for argument.

You fancy a journal article and a trade book derived from it (both composed by a textbook editor with an axe to grind) constitute 'overwhelming evidence'? The Unz world has et your brain.

You're still not providing any contrary evidence. If Pinker and Razib are for it, and it relates to genetics and biology, it seems like a good bet that it's true.

What 'contrary evidence'? Her work is a series of literature reviews. She stated a bizarre thesis (that childrearing has no effect on outcomes) and is an admittedly interested party. It's not my job to prove people wrong when they say cockeyed things, even if you believe them.

"behavioral genetics"

Why don't you suspend your opinions for the time being and come back to us in 100 years when ANYONE actually knows what they're talking about here. Only very limited knowledge on genetic determinants of behaviour has been compiled to date.

I'm not sure what system Tabarrok is describing. "Broken windows" policing is very much about increasing the probability of punishment--from basically zero under liberal, John Lindsay-style regimes, but it hardly features harsh punishments for quality of life offenses. And crime has dropped dramatically under such policing regimes. Tabarrok seems to proceed in a world of pure theory, without any attention to actual crime statistics over the past 20 years.

I strongly disagree with the claim that tough sentencing isn't working. It is clearly keeping criminals off the street. It is thus protecting non criminals from criminals. I call that a huge success.

If the argument is that we could get the same benefits in an even more humane or efficient way (such as lower punishment for victimless crimes), then I am all ears.

The means of doing so would be to invest more in police patrols and less in the prison census, i.e. invest more in deterrence and less in punishment and incapacitation. People who bellyache about 'mass incarceration' tend to be people who dislike cops as well, so they're seldom going to suggest that. I think Mark Kleiman might be an exception.

That's just what's required, more psycho cops:



Cops are useful and generally decent. You are expendable, and generally vicious and mendacious.

+ 1

I was wondering when someone would mention Mark Kleiman. He's an interesting fellow, a seeming exception to the lefty rule. That may explain though why he indulges in Republican-bashing so overtly at samefacts.org, more so than the other contributors. Perhaps he's compensating for his image among peers.

Does your study show that crime is deterred or deferred?

Alex, this is your best post yet. It was a good read, worthy of a link from the Browser. Yours trolly!

Becker's theory, which has much more to it, simply says that there's a trade off to be had between p and f, not that there's any sort of direct linear relationship. I mean, if we compare current punishments to historical punishments (historical prison quality, use of death penalty for more offenses) it's not clear Becker has even been shown to be wrong on this particular front.

Using behavioral economics to come up with practical solutions is a welcome change. However, the would-be criminal is not the only actor on the scene. Politicians want to be seen as tough on crime so they promote policies that don't necessarily have any impact on crime reduction but make politicians look tough. The people want someone punished, preferably an indigent and particularly a minority, especially if they perceive crime to be on the rise, a perception which a sensationalist media and opportunistic politicians are willing to encourage. Businesses want to make a buck which is easier done if prison populations rise rather than shrink.

So addressing the incentives to the would-be criminal will have minimal impact if those of the other actors are not taken in consideration too.

Abso freakin lutely.

Politicians routinely increase penalties with zero regard to any "optimum policy" and with the sole regard of being able to trumpet their "tough on crime" "achievements".

"parenting and crime control because the problems are largely the same. Moreover, in both domains ... punishment need not be severe."

Except that when a criminal makes his '10 minute mistake', the consequences for society at large can be very serious. Often the consequences of a child's actions negatively impact the decision maker (the child), as much as anyone else, and the outcomes aren't usually too serious. Locking up people prone to '10 minute mistakes' for extended periods can prevent crime.

Economic incentives to commit crimes and poor impulse control/limited future thinking certainly contribute to crime, at least for a certain type of individual. But a lot of people struggle with economic incentives and impulsiveness. Debt and obesity run rampant in the general population. How many of us are impulsive at times or fail to integrate all possible variables in decision making? Yet most people stay a reasonable distance away from committing crimes, particularly serious ones. Economic incentives and 'heat of the moment' thinking only lead to crimes when the person harboring them have criminal mindsets. The root cause of criminal mindsets is complex. But I really think that that is the fundamental factor of most crime. Normal people don't have to worry about making economic or impulsive decisions leading to crime. Only those with criminal minds commit serious, repeat crime and arguably most crime in general.

On my way home each day, I walk through a park with people walking their dogs. Generally, they are about as well behaved as their human caretakers, despite being far less intelligent and more impulsive. But recently, a small, but very aggressive dog bit me. It probably had less incentive to bite me then any dog out there since its small size put it at greater risk. And in absolute value, it had no reason to attack me. I wonder that most human criminals are little different. They attack other people, that's what they do, one way or another. Intelligence and impulsiveness play a role, but I suspect are secondary factors.

The problem is the assumption of risk-averse criminal. Maybe they become criminals in part because they are risk-prone. They want to be powerfull and very rich, but do not mind the risk of end dead.

Reminds me of a gangsta t-shirt I saw in Portugal which read "get rich, or die trying"

Judges are content to punish taxpayers and white, male, Christian heterosexuals.

pooooor you. The "victim" has spoken. The monopoly on CEO positions and the right to force your poorly motivated moralizing ways on EVERYONE is in retreat. Ergo, you are the victim. I feel so sorry for you.

Perhaps we should return to the system where Christians can re-assert religion as the primary determinant of all laws, regardless of whether they are ineffective or counterproductive laws, and regardless of how many people these laws discriminate against.

"Poor you".

My bold solution: castrate both criminals as well as parents and children of criminals. The bolder the punishment, the stronger the reduction in crime, AEBE.

BTW, I don't like Becker's approach. It seems to me most criminals don't consider there to be much of a difference between the various levels of punishment, and thus mainly consider the risk of getting caught, not the severity of the punishments.

So if I understand this correctly, what you're saying is that he was a dangerous fanatic. Or perhaps an ideologue, unresponsive to evidence, whose ideas on crime, to the extent that they were taken seriously, greatly increased the amount of human suffering in the world.

One of the problems with a mushrooming number of criminal offenses and draconian punishment for them is that the punishment isn't limited to the offender. His wife, kids, and other members of his family are punished as well, even though they've done nothing wrong.

And rightly so. It should be more like this.

I guess you're right, that's the way it's been in the past. Dropping nuclear weapons on Japanese teen-age girls walking to school is just punishment for the rape of Nanking and the attack on Pearl Harbor. Drone attacks on Afghan weddings helps get even for 9-11. Wounded Knee makes perfect sense when you think about Little Big Horn. Why should the domestic judicial system be any different? How long do you suppose it will be before wives will be forced to testify against husbands in the name of national security? Kids have already been trained as stool pigeons on their parents by the DARE program. If they don't squeal they should get permanent detention and a forever spot on the no-fly list. No high school football, either.

Dropping nuclear weapons on Japanese teen-age girls walking to school is just punishment for the rape of Nanking and the attack on Pearl Harbor.

There was that little matter of the war which had been ongoing for four years and the various alternatives one might pursue for bringing it to a close.

The bomb didn't help end the war, because the Japanese leadership did not care about bomb casualties. But I'm 100% in favor of nuking Turkey to smithereens just to teach them a lesson. Turkey's a democracy, unlike Japan. Every nuclear weapon exploded on every Turkish city would make me cheer with joy.

The bomb didn’t help end the war,


I chuckled, too.

It is a grave injustice when an unemployed man goes to jail for trying to make a few bucks selling weed. Not only is he punished, but the children lose a father, the wife loses a husband.

It is easily shown that single parents have more difficulty raising "well adjusted" children.

Laws that steal a father away from a child for petty crimes largely related to poverty (such as selling small amounts of drugs) are a gross injustice. If not for the father, then for the child. (And assuredly someone will argue that the child is better off sans "drug-addict" dad.)

It is a grave injustice when an unemployed man goes to jail for trying to make a few bucks selling weed. Not only is he punished, but the children lose a father, the wife loses a husband.

1. It's highly unlikely to happen to you unless your trading in a large mass of it or you've been collared umpteen times before.

2. I think you'd have to scrounge to find too many people who trade in street drugs for a living and maintain an orderly family life. The majority of 1st born children in this country are bastards, so your petty drug dealer will seldom have a 'wife' or a child in residence.

"It is easily shown that single parents have more difficulty raising “well adjusted” children."

-Correlation is not causation. Do you actually have any research on this (with expected racial and income adjustments)?

Also, Art Deco is right.

I know there's some research on this, but out of laziness, I appeal to common sense. Two parents is better than one.

One of the problems with a mushrooming number of criminal offenses and draconian punishment for them is that the punishment isn’t limited to the offender. His wife, kids, and other members of his family are punished as well, even though they’ve done nothing wrong.

These are fictions. You have to scrounge through the New York Penal Law to locate an offense that was not incorporated into the code more than 25 years ago, and what you'd find would be some elaborations on cybercrime (with regard to which some offenses were already present in 1989) and some shizzy bits like 'something something something within 500 yards of a school'. The sentencing schedule is the same bar a redefinition of the modal penalty for a class A misdemeanor. Penal codes generally allow too much discretion to judges and prosecutors, but that's not a novel problem and the business of legislatures has been to remove some of that discretion.

No, Alex's best post was on Neil Peart, and that will never be topped given the subject material.

You can easily do as much time in prison for stealing a car (grand theft auto) as white collar crimes which amount to stealing the retirement funds of large numbers of pensioners. This is one of many examples where we can see that the punishment system is not fair.

What about pushing sales of Risperadol for the elderly and children, when a) this was not approved by FDA and b) it was known that there were major health risks? This is far more dangerous to national health and well-being than a million tonnes of marijuana, yet hundreds of thousands rot in prison for crimes related to drugs which have been classified as illegal, while J&J sales executives continue to enjoy not only freedom, but their millions of dollars in illegal proceeds.

Excessive punishment is not the only injustice. The ability of privileged groups to get off easy is another major problem in the punishment system.

Policies which exclude ex-cons from jobs are easily one of the strongest factors behind recidivism. Perhaps second to the forced criminal networking club that is prison.

The real issue has become background checks. People who, for whatever reason, for sure a term in the joint, can't possibly pass a background check and won't be able to get real employment, rent an apartment, get a mortgage, fly on a commercial jet and so on. Having once failed a background check, the individual will be on it forever, regardless of any rehabilitation or change of heart he might have. This means that the number of those failing background checks will continue to grow until a significant portion of the population is unemployable and unhouseable. Yet these people will still get hungry and thirsty, need some rags to wear and a place to rest their weary head. Crime, of course, is the easy answer to this predicament. Also, businesses will be established to exploit and service these people. They will be employed in low-paid casual labor or perhaps even low-paid skilled labor. Landlords will consider them a last resort that can be extorted for higher rents than the unconvicted. At some point these untouchables will become a caste whose numbers will present a real problem for the nation/state. I hope that I'm alive to see the fun.

an’t possibly pass a background check and won’t be able to get real employment, rent an apartment, get a mortgage,

More than a third of the population lives in rental housing. Somehow they manage. Per the Urban Institute, about 0.25% of the population are vagrants, a far smaller number than the population of ex-cons. Somehow, convicts find a place to live. I have in my social circle a young man who was imprisoned from 1997 to 2011. He's had no more trouble than an ordinary man finding work in the four years he's been out of prison. People manage.

"by exposing more people to criminal peers and by making it increasingly difficult for felons to reintegrate into civil society, longer sentences increased recidivism"

I think this is the crux of the matter.

Fortunately, a lot of people are sensible about this. In high school, I once got in trouble for smoking pot. I had to go to a meeting with an addictions counselor. a) she took me for my word that I didn't feel problems with addiction, and b) she said that it would be a bad idea to join any groups for troubled youth, because the troubled youth would probably introduce me to more problems than attendance could possibly solve.

I had a part time job, had good marks, participated in many extracurriculars, had friends, and ... smoked weed. As opposed to what I believe is her typical American peer, she believed that any "cure" was worse than the "disease".

Indeed, they let me "get away with it". Oh, except they gave me a few days off school. A well deserved break, I reasoned, given my hard working morning to midnight routine.

Then again, smoking weed doesn't really reflect a "criminal mindset" does it? So perhaps this is not a good example.

I dispute the assertion that rational choice implies that "twenty years in prison deters twice as much as ten years in prison"; in fact the result that k-percent increases in `p` has higher impact on behavior than k-percent increases in `f` understood naively can also arise from the intuitive observation that, somewhat surprisingly, criminals' response to prison time is sublinear. One year in prison is a long time; it's unpleasant, dirty, it disrupts your life and will likely end your current job and current relationship. Being a felon is a black mark on your record that harms your reputation. Ten years in prison is worse, but arguably much less than ten times worse. Sure, it's the same shit, but you get used to it, and eventually psychologically readjust to the point where it becomes a "new normal". Your job and relationships are almost certainly all gone, you're already a felon, so the 10th year doesn't do all that much. So yes, of course increasing prison time from 9 to 10 years does less than increasing p from 9 to 10%.

Foolish talk. Until you change the hearts of men, you can create as many deterrents as you please. Nothing will change.

"Incentives don't matter".


Comments for this post are closed