US Incarceration Rates

Combine an incentive for politicians of both parties to demagogue for "tough on crime" legislation with an over-broad approach to legislating anything seen as bad behavior by the majority as a crime, and you get the highest incarceration rates in the world.  Scary charts, with incarceration rates growing entirely out of proportion to crime and population.


  1. Dr. T:

    The majority of citizens of the USA retain a wide Puritanical streak and support criminalization of consensual paid sex, gambling, and recreational drugs other than alcohol. Whenever people are found guilty of these "crimes" (particularly drug use and especially 'crack' cocaine), our courts manditorily impose lengthy prison sentences. We imprison more people for nonviolent, victimless crimes than any nation in the world.

    Due to the lucrative War on Drugs (lucrative for law enforcement agencies, that is), we have the highest rate of SWAT team-led arrests in the world. Everyone who smokes a joint must be armed and dangerous, right? Therefore, a SWAT team must stage a no-knock entry at 4 am to bring in that dangerous criminal.

    We also do one other thing better and more often than other nations: confiscate money and property when there is only a suspicion of criminal activity, even when no charges are filed against the owner of the property. The Supreme Court let stand this trashing of the fourth amendment that allows police departments to seize property without adequate justification and with no way for the owner to reclaim it.

    Welcome to the Banana Republic States of America.

  2. Michael:

    We also do one other thing better and more often than other nations: confiscate money and property...

    We're one of the few nations that let people have it in the first place.

  3. Dr. T:

    Oh, so that makes it OK. Why don't you give me your name and address, and I'll make an anonymous tip to the police that you deal drugs from your home and your car. That way you can be one of the contributors to our law enforcement agencies' financial enrichment program. After all, since our great nation let you gather wealth, it's only fair that your wealth should be confiscated without any evidence that you committed a crime, right? Perhaps you'll change your tune after your home, auto, and savings have been seized.

  4. Roy:

    Prisons, which the bottom of the 3 graphs show if the top 2 did not, don't work. But that doesn't stop people from choosing politicians who promise them.

    Most folks (at least initially when asked in a poll that provides the options and makes them something other than obscure) favor restitution over incarceration.

    But that system also involves at least two hard choices folks don't want to face: 1) servitude for someone without liquidity to restore; 2) stepwise more serious response up to and including death to those refusing to restore.

    When they think thru these sorts of outworkings, the initial response often wavers and changes.

  5. John Moore:

    I disagree with Roy. Prisons do work, by keeping criminals off the streets. Correctional institutions don't correct, but prisons do imprison. The falling crime rates are at least partly due to the high incarceration rate.

    OTOH, I agree with others that we over-imprison for certain categories of crimes.

    Remember, the high incarceration rates were due to a citizenry demanding that the government do its job: protecting the citizenry against predators. The result were laws such as the 3-strikes laws, which indeed took lots and lots of predators off the streets. I know one of those 3-strike predators personally, and while it is very sad that he is locked up, I have little doubt that he would go back to victimizing folks if he were on the streets.

    IOW - government is doing what is demanded of it, and being government, not doing it with much grace - in this case, erring on the side of over-incarceration after decades of liberal policies based on flawed premises that caused under-incarceration and lots of crime.

  6. mark:

    Very misleading chart. They make it look like crime rates are flat and always have been since 1960. In reality, the violent crime rate in 1960 was 160.9/100,000 capita and increase to a peak of 758.1 in 1991. That is an almost 5 times increase in the 30 years that this graph clearly does not show.

    Since 1991, the level of violent crime has dropped almost in half, from 758 violent crimes to 454 violent crimes/100,000. And, if you look at a properly scaled graph this is highly correlated with the increased rate of incarceration that started in the 1980's.

    GETTING TOUGH ON CRIME PAID WITH VIOLENT CRIME AND PROPERTY CRIMES ALMOST HALF THEIR PEAK RATES. The reason is that people who are jailed for crimes, mainly drug crimes, are not free to commit other crimes.

    Don't be fooled by this idiotic, pure libertarian propaganda.

  7. mark:

    And, if you cannot see the correlation between decreasing crime rates and increasing incarceration, you are missing the story. Incarceration rates were flat from 1960's until the 1980's. Then we did the "War on Drugs" and started locking people away. You can see that in the chart. And "coincidently" the crime rates started to decrease.

    If this post is redundant then so be it. THis type of graphical trickery is uncalled for and not much different than the data supplied by climate alarmists.

    If you want to see the trends in crime rates in the US here is a simple wiki:

  8. Frink:

    LOL -- Yes, this reminds me of James Taranto's recurring segment "Crime rates fall despite increased incarceration."

  9. GaryP:

    I think the commenters have won this argument.
    However, I do think that legalizing small amounts of drugs and shifting the focus onto property and violent crimes would make sense.
    Prisons are the only reasonable solution for violent offenders. Property crimes are more difficult to balance (i.e. punishment vs. restitution).
    How to avoid creating criminals in our society is the real issue. Poor public education, welfare, and lack of personal responsibility are possible culprits.
    Good insurance (for property losses) and being able to defend yourself (weapons, self-defense training, etc.) seems the best approach for individuals. Ending our "it's not their fault" mantra from our so-called leaders would at least make victims feel a little less insulted (on top of injured).
    Widespread use of the death penalty, for murder, in a expeditious manner (years, not decades) would at least eliminate repeat offenders--and might deter some. (I know that comment will attract a lot of hate, but we have never really tried to develop a reasonable, fair (not perfect--but fair) system to destroy the murderous amoung us.

  10. Dr. T:

    The above commenters have won no argument. The bulk of the changes in the violent crime rate reflect changing demographics: the number of teens and young men in the inner cities directly correlates with violent crime rates. The rates fell after the large baby boomer generation passed thirty years of age.

    If you look at prison statistics, you'll see that the greatest increase among the incarcerated is those convicted of nonviolent drug crimes: drug users and small-time dealers. The number of nonviolent drug offenders in prison increased 12-fold from 1980 to 2010. Fewer than half of the men now in state prisons committed a crime of violence.

    Our prison population now exceeds two million. Nearly 2% of adult males live in prison, the highest of any nation in the world. Yet, despite all these incarcerations, our violent crime rate still ranks among the top 15 in the world. We rank #24 in murder rate among all nations, and we rank #1 in murder rate among wealthy, industrialized nations. If all this incarceration is designed to keep violent criminals off the streets, then it isn't working. What happens instead is that nonviolent prisoners, once released, have little hope of reintegrating into society and turn to property crimes and robberies. (Skills they learned from their cellmates.)

    We need to legalize victimless activities and stop imprisoning people for using recreational drugs, renting their bodies for sex, or betting on a sports event. We need alternatives to prison for nonviolent criminals. (Prolonged parole with work and reparation payments makes more sense.) We need to stop pretending that imprisoning an ever-growing proportion of our population is the best way to increase our safety.

  11. mark:

    "Changing demographics" my ass. You are claiming that from 1991 to the present that the demographics "changed" in the opposite direction tat they changed from 1961-1991. BS.

    The problem is that the people who are incarcerateed for "non-violent" crimes also commit other crimes. When they are incarcerated for these "non-violent" crimes they are no longer free to committ the other crimes they routinely committ. People using and selling "recreational drugs" have a high correlation of committing other crimes to support their habit or create their drug market. The fact is, they are incarcerated for the "non-violent" drug crimes more often because it is usually easier to obtain evidence of these crimes than others.

  12. Roy:

    Mark, Gary, John: you've missed the obvious. You have allowed the partial correctness of your response to blind you. You fellows argue that prisons do work since folks inside can't do bad stuff. You are correct. But does that define "workable"?

    Suppose the 'system' put 10X as many folks inside. Surely crime would decrease. How about 100X, either via making a whole bunch of bad stuff felonious, or by the 'system' imprisoning a significantly larger percentage of those caught under present laws. Surely you realize where this reasoning leads.

    By "prisons don't work" I did not mean they have zero effect on crime rates. I meant (my first point) that the coefficient of corelation between incarceration increase and crime decrease is something very different than one. (Dr T provides data supporting this point in his 3rd paragraph two posts above.) That result happens because incarceration faces the same sort of marginal utility characteristics that any human enterprise faces. At some point, costs increase faster than do benefits. The charts (and a host of internet available data)show that the 'knee of the curve' has been passed. (BTW, Mark, Dr T's demographic point hinges on the 'greying of the population'. This has had a very significant impact if not the most significant impact on crime rates. You mistakenly insist that demographics did not change directions in the '60s and '70s. Consider, for example, what Roe accomplished. Consider what looms ahead for the ponzi scheme aka Social Security.)

    Nonetheless (and this was my second point), people (eg, Mark, Gary, John) still favor prisons over any other solution. Other solutions do exist.

    Contra Gary, for example, death rather than prison is a solution for (at least some) violent crime. Since drug usage is, as a matter of fact, very dangerous to lots of people other than the one taking drugs (drugs are most defintely not a non violent crime), making the user *responsible* is a solution. Legalize drugs. Fine. But you kill somebody when you're high (including on alchohol), you die. Wipe out somebody's assets (eg, a car wreck), then restore that asset. (Can't pay? Sure you can. You've just sold yourself into servitude until the debt is paid.)

  13. John Moore:

    Do you deny that there would be a big spike in crime rates if we let a significant proportion of inmates loose?

    I never argued that our prison policy is perfect. It is not.

    It may very well be that we could cut out prison population in half and not see a big increase in predatory (as opposed to victimless) crime, if we could *magically* identify which prisoners would not predate. Do you have that magic wand? Furthermore, mark is quite correct: many imprisoned on victimless crime charges are in fact predators - either by desire or because they need to steal to support their drug habits. Do you deny this?

    Personally, I'd like to see the end of marijuana criminalization, and likewise for some other drugs. But nobody in their right mind would want crystal meth or crack cocaine legalized unless they can provide a compelling argument that decriminalization would REDUCE their usage - a very dubious assertion.

    You also mention that the correlation coefficient is different than one. So what? Every additional crime victim who is spared, by incarceration of people who violate the law, is a win for incarceration.

    As for the hand waving that "it's the demographics, stupid" - it is indeed hand-waving. Yes, demographics has an impact, but it doesn't explain the 2X drop in crime rates coincident with increased incarceration any more than it explains the 5X increase in crime rates coincident with an exceptionally lax attitude towards law enforcement and criminalization. That's two strong correlations you have to defeat. Obviously demographics is a cofactor, but not a full explanation.

    The hypothetical you raise - the curve of diminishing returns - is not applicable. We won't lock up 10x or 100x people, because we lock up criminals, not percentages. Whether you and I agree or not, people are incarcerated for committing what are indeed viewed as crimes by the vast majority of the population.

    As for what Roe accomplished... it is partly responsible for the epidemic of fatherless inner city children - since, with Roe, it is even easier for the biological father to claim no responsibility - "hey, she should have aborted the kid". Beyond that, it certainly changed the racial balance, by disproportionately killing black babies, but it is merely a hypothesis that this led to reduced crime.

    As to why people still favor prisons... how else do you propose to keep the criminals away from their prey? Prisons are extremely effective at this. What other solutions do you have? I am genuinely interested.

    "Legalize drugs. Fine. But you kill somebody when you’re high (including on alchohol), you die. Wipe out somebody’s assets (eg, a car wreck), then restore that asset. (Can’t pay? Sure you can. You’ve just sold yourself into servitude until the debt is paid.)"

    Do you really think that drug addicts are going to be deterred by this? Many drug users have pre-existing character flaws or mental illnesses that make them irresponsible and short-sighted, or they get into it without realizing the consequences or the power of addiction. So your approach is asking for a holocaust.

    But most drug addicts don't kill people, they just suck up societies resources and steal lots of stuff. Since law enforcement can only correlate the victim and the addict in a small percentage of the cases, folks still end up poor, wile you have created legalized slavery. Sorry, that isn't too appetizing.

  14. mark:

    1. The population is not "greying" enough to make that big of an impact on crime. "Correlations" in statistical modeling are not proofs of causation. I can specify a model any way I want and get the coefficients to come out the way I want them. Only people without risk actually believe these models (I have over 15 years of real world statistical modeling experience, I know what I am talking about).

    2. The real "problems" with prison is that the liberals who have dominated the academic world for the past 50 years have never, ever, developed a true rehabilitation plan or drug rehabilitation program that works. The only marginally effective addiction program is 12-steps approach which is a conservative, relgious approach to overcoming addictin.

    3. I agree with you that prisons are simply punishment by removal from society. But the vast majority of people who are incarcerated, particularly for drug crimes, have absolutely no interest in being rehbilitated. They are long term criminals and most are gang affiliated, which means that crime is not just a "career" cor them, it is their entire lives.

    4. Rather than surrender to the supposed diminishing marginal returns, lets move the other way. Lets make the consequences for crime significant. Right now, the vast majority of crimes have absolutely ZERO consequences on the offender. Again, the VAST MAJORITY OF CRIMES HAVE ABSOLUTELY ZERO CONSEQUENCES ON THE OFFENDER. You steal someone's coat from the employee locker room no one will even call the police because they will not do anything about it. Shoplifting. Auto theft. Drug dealing. These are crimes that you have to be tried and convicted over and over again before you will suffer anything other than minor inconvenience. Lets build the prisons and create a network of incarceration were there will be consequences for your actions. Steal something from the store, you are going to a minimum security type of camp for a month. Steal again, it is 3 months, etc, etc. I bet people would reconsider their actions if there were REAL consequnces.

  15. Roy:

    No, John, I do not suggest we not deal with bad folks in prison by releasing them. Nor do I think you argue that the prison policy is perfect.

    But I do think there exists an excellent, not used response to predatory crime. I mentioned it in my second post: responsibility for restitution backed up by a)servitude, b) death for refusal. (Gal steals a car. Becomes responsible for all costs leading to her identification, capture, trial, and restoration of stolen car. Suddenly theft becomes other than profitable.) Predation self selects itself, and supports its elimation. Predators *rather than* victims pay.

    Druggies don't care? If they don't predate, I don't care either. But if they predate to support their choice, then they select themself for rehabilitation (at their expense) or death.

    You think, btw, that prisons don't legalize slavery? Tell that to inmates who work at 7 bucks a month. And can only buy at the company store. Read the 14th Amendment, which specifically allows state run prisons as an exclusion to the prohibition of slavery.

    What I propose removes people from rather than encourages the perpetuation of slavery.

    John, you have correctly identified one of the more immediately obvious (albeit regularly denied) results of Roe: fatherless children (and disintegration of, nay, war against the family, with generations of poverty resulting). Not merely race wars, but gender wars loudly proclaim those denials.

    Don't joint the deniers by blinding yourself to some other significant long term effects of being able to walk into a 2010 elementary classroom and note that demographically fully one third of the kids who would be there sans Roe aren't. That *reality* (ie, not my opinion) certainly does mean the population age makeup skews upward. Not only will Roe train wreck with OASDI (aka Social Security). It also means, since most crime (especially of the violent predation type about which you are properly concerned) gets accomplished by, well, those not grey.

  16. Roy:

    Mark, you have too much statistical expertise to deny that 'greying' has no more than a minor effect on crime rates. Get past the banter on (thanks for correcting my spelling) correlation vs causation, and simply acknowledge the obvious. (One of the ironies: greying is already making prisons vastly more expensive, as those inside require more than 3 hots and a cot, namely significantly expensive medical treatment that sometimes accompanies aging.)

    Neither liberals nor anyone else will develop a program of rehabilitation if that program denies rather than incorporates responsibility. The present prison policy has, among its ironies, exactly the opposite of rehabilitation as its most likely result. I think you already recognize and agree with this, so I need not defend the point in detail.

    I completely concur with your zeal for making crime have significant, easily identifed and understood consequences. Where those who victimize rather than the victims pay.