Bryan Caplan  

How Many People Does the War on Drugs Put in Prison?

PRINT
Two Tullock Stories... Is George Selgin defending Key...
There are over 1.5 million people in American jails and prisons.  Why are they there?  Take a look at the latest numbers from the Bureau of Justice Statistics. 

Here is the offense breakdown for state-level incarceration for 2012, which continues to outnumber federal incarceration by a factor of 6:1 or so:

incar1.jpg

Here's the federal breakdown for 2012:

incar2.jpg

Most libertarians blame the massive U.S. incarceration rate on the War on Drugs.  At the federal level, this is hard to deny.  But the federal level remains a small slice of the pie.  At the state level, drug offenses are only 16% of the story. 

In absolute terms, of course, that's tons of people.  Drug offenders outnumber murderers.  They outnumber rapists.  They outnumber robbers.  Indeed, they are only slightly outnumbered by all property criminals combined.  But at first glance, blaming the War of Drugs for mass imprisonment runs afoul of basic facts.

What about at second glance?  Econ 101 warns us against using mere accounting to resolve causation.  In the absence of the War on Drugs, many non-drug offenses would never have been committed.  Without prohibition, gang-related violence - and related weapons charges (subsumed under "Public-order" at the state level) - would plummet.  Habit-related property crimes would probably do the same, albeit to a smaller degree.  Theoretically, drug offenders might simply switch into other illegal activities once drugs were legal, but it's hard to believe this effect would be sizable.

My best guess: Five years after the end of the War on Drugs, half of the prison-years handed out for non-robbery violent crime would be gone, along with 20% of the prison-years for robbery and property crimes, and 75% of the prison-years for weapons charges.  That's roughly half of all prison-years at the state level, and two-thirds of all prison-years at the federal level.

Got a better guess?  Please share in the comments - and please show your work.


Comments and Sharing






COMMENTS (23 to date)
Carl Shulman writes:

If the OP analysis is right, and drugs aren't a major cause of rape/sexual assault, then non-sexual assault, non-robbery violent crime must fall by 72%. Maybe that is what Bryan is predicting if he thinks weapon offenses would fall by 75%. That seems like a very big fall.

The BJS suggests that a much smaller portion of violent and property crime is drug-related in this way.

http://www.bjs.gov/content/pub/pdf/DRRC.PDF

"According to BJS national surveys, the most comprehensive information available, an estimated 17% of 1991 State prison inmates and 13% of convicted jail inmates in 1989 reported committing their offenses to get money to buy drugs
(tableā£3)."

[26% of property offenses, 27% of robberies in the 1991 state figures, along with 22% of drug-related offenses.]

"BJS examined homicides in the 75 most populous
counties in the United States in 1988. Many of the homicides involved drugs or drug trafficking, including the following: drug manufacture, dispute over drugs, theft of drugs or drug money, a drug scam, a bad drug deal, punishment for drug theft, or illegal use of drugs. One of these circumstances was involved for 18% of defendants and 16% of victims."

Carl Shulman writes:

"weapon offense"

Offenses, of course. Curse the lack of edit functionality in this comment section.

[Fixed. Sadly, if we had an edit function, some people would abuse it ex post to curse more literally.--Econlib Ed.]

Carl Shulman writes:

72% would be using only the state figures, incorporating the federal ones would slightly reduce the needed change.

Brian writes:

Quite a lot of drug incarceration happens when police know their neighborhoods and know who the bad kids are. They know the bad kids are committing property crimes and moving up to violent crimes, but they don't have enough proof of serious property crimes to put them in jail yet, so the police frisk them regularly until they're caught with some amphetamines or marijuana.

Then the police can put them in jail for drug crimes and the kids might get the idea from their parole officers eventually of getting a steady job and some discipline.

But the kids aren't going to jail over the drugs; they're going to jail for being bad kids and the drugs are just the easiest thing for the police to prove.

In the absence of drugs laws, they'd still need to go to jail but the police would have to spend days documenting some property crimes and minor assaults instead of just finding the pot in the kids' pockets. Or the police might have to fabricate evidence to put the bad kids in jail.

So a lot of those drug prisoners were going to jail for something and it just turned out that the drugs were easiest to prosecute. If drugs were legal, those drug prisoners would still be in prison, but for something else.

Roger Sweeny writes:

I suspect Brian (not Bryan) is on to something.

This is wildly speculative but I fear ending the war on drugs would not have too much effect on convictions. Criminals are, to a large extent, stupid people with impulse problems. They find it difficult to hold regular jobs. Ending the war on drugs won't change that.

Radford Neal writes:

Perhaps Brian and Roger are right, and these kids would still go to jail if they weren't convicted of drug charges.

Or perhaps not. Maybe these kids would turn out better if we didn't tell them that the justice system is fair, with a judge and/or jury deciding whether to convict based on a presumption of innocence, and then permit selective enforcement of drug laws that effectively allow the police to act as judge and jury, deciding on their own who to put in jail.

Lies have consequences. Generally not good ones.

awp writes:

Or, maybe in these neighborhoods cops "know" all these kids are going to be criminals because 40 years ago half their fathers and uncles were busted on b.s. possession charges and denied any chance to find meaningful work to provide a proper upbringing for their family. I mean it was easier back then, we "knew" all blacks were degenerate criminals, and they were all then confined to the given neighborhoods. Or, maybe things haven't changed enough.

Tom West writes:

I have to say I'm always a bit nervous at the idea of complete legalization. Unfettered by legal restrictions, I keep wondering what large scale industrial production techniques to lower costs along with modern advertising campaigns and pharmaceutical research to make it "safer" could do to consumption.

It would certainly be a grand experiment, and at least we know the very high costs of prohibition that legalization would be weighed against.

But I could imagine it going off the rails pretty badly. I could easily imagine opiates on the old being used as commonly as Ritalin on the young, for example.

Christophe Biocca writes:

Roger Sweeny:

Criminals are, to a large extent, stupid people with impulse problems. They find it difficult to hold regular jobs. Ending the war on drugs won't change that.

Let's just grant the premise, for the sake of argument. It doesn't change the outcome. People with poor impulse control aren't completely irrational, they do respond to incentives.

Right now we have enormous incentives for those people to enter the drug trade. There's no reason to believe a-priori that they'd resort to bank robbery instead of just living poor but law-abiding lives. If they all go on welfare, well that's a lot cheaper than keeping them in jail.

We know the incentives are enormous. They destabilize entire South American governments, they power criminal cartels that can kill and maim with near impunity. Why is it such a stretch to believe it pushes a lot of people over to law-breaking?

Glossing over incentives is a big mistake.

Bostonian writes:

In America today, what is not forbidden soon becomes a civil right that should be backed by federal enforcement. Soon after homosexuality was made legal by the Lawrence Supreme Court case came demands for homosexual marriage and for non-discrimination laws to cover homosexuals.

If we legalize marijuana, how soon will it be before employers, even private ones, are forbidden from testing applicants and employees for marijuana and other drugs?

Daublin writes:

+1 Radford.

I am sure it happens, but the police should still tell the truth about the actual crime they suspect has happened. Selective enforcement has many problems, including:

- It jails a lot of innocent people. Even if the cop is *usually* right, up to 49% of the time they made a mistake.

- It unravels a general trust in society, which leads to more crime. If the police are just going to beat up on people they don't like, without much regard to actual evidence, then where is the motivation to be an upstanding citizen?

- Cops are not ideal as judges. We are much better off if policemen are the muscle, and the final binding decisions are made by cooler heads.

floccina writes:

One lonely data point.
A guy that I worked with killed a guy who sold him something other than crack cocaine (I do not remember what it was) for the real thing. He was wrong to give the guy the death penalty for running that scam on him and effectively stealing a little money from him, but he was basically a hard working decent person other than that. I doubt that he would have done any other crime had crack been legal.

Floccina writes:

BTW: At $40,000/year to keep a person in prison I think that we should do more to keep other non-violent criminals out of prison even also maybe even some violent ones. I hear that more police can actually reduce crime. Restitution and ankle monitors might help.

Mark Bahner writes:
Five years after the end of the War on Drugs, half of the prison-years handed out for non-robbery violent crime would be gone, along with 20% of the prison-years for robbery and property crimes, and 75% of the prison-years for weapons charges. That's roughly half of all prison-years at the state level, and two-thirds of all prison-years at the federal level.

Got a better guess? Please share in the comments - and please show your work.

I don't see any way that half of imprisonments for murder, manslaughter, rape/sexual assault, aggravated or simple assault, or "other violent crimes" would go away. That's saying half of all those crimes are caused by drugs being illegal. Half of all rapes and sexual assaults are because drugs are illegal?

I can believe a greater-than-50-percent reduction in weapons charges, and a 20 percent reduction for robbery and property crimes.

Calculations: 12.7 percent...carry the 2...add swag factor = 17.3456% reduction in non-robbery violent crime. At most.

sam writes:

The problem is not the WoD. The problem is the combination of WoD, electronic criminal records, and age predisposition.

WoD arrests are primary young men with high risk tolerance. They may be impulsive as well. They're going to smoke pot, drive at twice the speed limit, and perhaps get into fights.

Half a century ago, most of this would end up in a stern talking to, or at worst a beating by, the cops. Most of these young men would age out of risk-seeking behavior, become productive members of society, and look back on their past with a little laughter and a little shame.

That's still what happen now if you're in the socio-economic class that can afford a good lawyer.

But what of the poorest third of the male population?

Now they get a drug conviction that follows them for life (or perhaps felony speeding, assault for a bar fight, etc.)

You've now created a class of millions of men that don't have much capital, don't have the connections to get a job, don't have the temperament for formal education, and due to a criminal record have been banned from the vast majority of blue-collar work, itself a shrinking category.

What do you think these millions of alienated young men will turn to?

Brian C writes:

You are leaving out one critical variable even assuming all the areas you claim that might fall will.

I did not have time to look up the number, but well over half of all crimes are committed while under the influence of drugs or alcohol. People do stupid things when they are high or drunk.

If your argument as I read it, is to legalize all drugs include hard drugs like LSD, and Meth (which the stats you provide include these type of offensives), then you would also have to assume substance abuse would also increase. You would also have to conjurer that cases were one is moderately to seriously mentally inebriated would also rise, I would argue, quite sharply over the long term. This would be primarily be to a higher use rate from use of more extreme mind altering "hard" drugs.

Therefor, Domestic violence would absolutely go up.I would also expect rape, and DUI's with their associated accidents and fatalities would also rise.

Therefor your argument's merits have serious issues because you completely ignore the obvious crime rate increase that would happen in other crime stats.

Nathan W writes:

How much property crime basically originates from people getting a record for drug possession then not being able to get a job and then feeling like they got kind of screwed by society and then they don't feel so bad about property crime? How often does an initial incarceration for some drug offense effectively direct the person (by network effects, among others) into a life which involves other forms of crime?

These numbers are interesting, but I don't think they reflect the scale of the number of people who suffer from a criminal record for drug possession who may have had a suspended sentence or a very short sentence.

Bostonian writes:

"You've now created a class of millions of men that don't have much capital, don't have the connections to get a job, don't have the temperament for formal education, and due to a criminal record have been banned from the vast majority of blue-collar work, itself a shrinking category."

We need a zero or at least lower minimum wage, so that employers are more willing to take a chance on applicants with a criminal record. There should also be a safe harbor allowing employers to pay employees with criminal records a lower wage for some time (say one year) and to fire them without having to show just cause.

ThomasH writes:

But without a war on Drugs, how do we finance guerrillas and terrorists that we can send our veterans-to-be to fight?

sam writes:

Bostonian: Good luck with that. It's far too rational.

The vast majority of the population operates under the belief that pay goes down when mean employer's don't care and pay goes up when employees empathize.

Thus, there should be minimum wage laws to force them to care. Places with low minimum wage laws obviously don't care enough, especially about children.

Breaking people's hunter-gatherer economic superstitions is going to be even harder than breaking their hunter-gatherer sexual superstitions, and that has taken millenia.

Nathan writes:

I'm a criminal defense attorney. My estimates from personal experience:

Almost 100% of theft and property related crimes are due to addicts trying to get money for drugs.

Maybe half of robberies/ murders/ agg assaults are drug disputes. Most of the rest are alcohol or mental health related.

Not many rapes have anything to do with drugs, although most are alcohol related. If you lump in child molests as sex assaults, even fewer are drug/alcohol related.

Low-level public order stuff like disorderly conduct is almost all alcohol/ mental health. Not really sure what % of felon in possession cases are drug dealers vs. guys who just want to carry a gun regardless of what the law said. But of course, if felony convictions drop, that's a lot fewer felons to violate the law by having a gun.

Around a third to a half of my clients were drunk when they committed their crimes. 10-15% were high on some illegal drug.

One thing I'd be interested in seeing with legalization is how the relationship between police and poor communities changes. Right now, the police are seen as an occupying force. Everyone is afraid of getting busted because they have drugs on them or in their car, or because they have warrants out. And they see the police as the people who took away their family members for years for being addicts. Change that, and you might get increased cooperation, which might actually lead to an uptick in incarceration for real crimes due to greater closure rates, even if the absolute rate of those crimes drops.

Keith K. writes:

@ Brian C.

"If your argument as I read it, is to legalize all drugs include hard drugs like LSD, and Meth (which the stats you provide include these type of offensives), then you would also have to assume substance abuse would also increase. "

Actually you should expect the opposite to occur. As drug use is commercialized the incentive structure is to sell lower potency types of the drug in question. Look at the change in structure of the kinds of alcohol sold during prohibition vs. after prohibition. It primarily went from spirits to beer & wine.

This is also what happened in Portugal when they decriminalized everything 10 years ago. Marijuana use went up, but the use of Harder drugs went down, despite the now greater availability and lower effective cost of taking them. The various pathologies related to illegal drug use also plummeted.

Also, since use is no longer criminalized the users are now much more open to seeking substance abuse treatment. Again, this has been the Portugal experience.


The biggest change I think from ending the war on drugs is simply the subsequent change in police culture that would occur over time. Refocusing police back on to actual crimes will make them appear much less hostile to poorer communities and I imagine that they would be much more willing to cooperate with investigations. Then there is the whole issue of effectively destroying the financial basis of most criminal gangs and thereby undercutting their power.

Looking just at the incarceration statistics does not sufficiently cover all the dynamics that would change if the war on drugs were to end.

Jacob A. Geller writes:

What percentage of drug-related charges add time to non-drug related sentences indirectly, by constituting prior offenses during later sentencing for non-drug related convictions?

My hunch is that drug offenses might add more to prison time served indirectly by lengthening sentences for other charges, than directly, but I would love to see some data.

If prior drug offenses added on average 20% more time to all other crimes committed, I think that would do the trick.

Data?

Comments for this entry have been closed
Return to top