« "Why Nobody's Talking About the Supreme Court" | Main | "How Did Chicago Get So Violent? Did the effort to eradicate the city’s gangs in the 1990s inadvertently lead to its bloody present?" »

September 27, 2016

When someone focused on criminal justice empirics calls this the "Greatest. Graph. Ever."...

via this tweet, I feel compelled to reprint it:


Those who are familiar with Professor John Pfaff's work on Twitter or elsewhere will surely understand why he views this graph as reflecting so much greatness, and those not familiar with Professor John Pfaff's work should see this post as my recommendation that you take the time to figure out why he things this graph is so great.

Also, to add my two cents (and also throw in another useful discussion point), I think the graph would be even better is it also noted that December 1972 also marked the end of conscription for the military (i.e., "the draft") in the United States.

September 27, 2016 at 08:57 PM | Permalink


Any chance you could post a reply here givning at least a thumbnail of his theories/findings? Some of us are too busy applying the modified categorical analysis constantly, yet are interested in this. Thank you.

Posted by: USPO | Sep 27, 2016 11:23:37 PM

USPO: Here are some prior posts highlighting John's work:





Long story short: John's work repeatedly highlights nationwide that prosecutors starting bringing felony charges for twice as many cases in the 1980s and 1990s, especially violent crime cases.

Posted by: Doug B. | Sep 28, 2016 8:27:16 AM

I hope the author is not advocating a return to the military draft, especially a male-only one now that women have equal rights are no longer barred from combat duty (the chief rationale for male-only conscription). Indeed, male-only conscription would not stop violent crime. In fact, the Richard Speck, Manson clan killings, etc. occurred during very high draft calls. Maybe the male prison population temporarily decreased during war-time, but the military merely transferred the criminal element overseas where they targeted both foreign countries and other U.S. military personnel on bases. The wartime drop in male prison population obviously did not prevent the violent lone-wolf crimes like Richard Speck and Charles Whitman (a former Marine) from going on their killing rampages. Indeed war and increased draft-calls might have contributed to a rise in the especially violent sadistic crimes like the Manson family and Speck.

No, I hope the author is not getting ideas about male conscription being a panacea.

Posted by: william r. delzell | Sep 28, 2016 9:17:08 AM

WRD: I would never assert or even mean to suggest that "male-only conscription would ... stop violent crime" or be some kind of criminal justice "panacea."

But I do mean to suggest that we all recognize and consider that:
(1) the vast majority of serious crimes are committed by men between the ages of 18 and 35, and
(2) forcing men of these ages to either join the military OR stay in school longer likely will generally lead to this crime-prone cohort, relatively speaking, being
(a) somewhat less likely to commit crimes, and
(b) somewhat less likely to be prosecuted for the crimes they commit (e.g., "underenforcement" is much more common, I think, historically and still today in the military and on colleges campuses), and
(c) somewhat less likely to be incarcerated for the crimes for which they might be prosecuted.

Posted by: Doug B. | Sep 28, 2016 12:53:02 PM

I'm with Doug on his thoughts, here. A lot more logical than JP

Posted by: Dudley Sharp | Sep 28, 2016 1:05:27 PM

This will give a better picture of what may be happening here:


That is, the number of people institutionalized in mental hospitals shifted to prisons with the closing of mental institutions during the transition reflected in the graph in this post. Check out figure 2 on page 43 for what I believe a better depiction of criminal institutionalization in this regard. Thank you.

Posted by: will | Sep 28, 2016 1:41:54 PM

The problem is that this data can be spun all sorts of ways depending on one's agenda. Doug's fantastical claim in this comment section about who commits crimes is a perfect example of that fact. One of the great tragedies of American culture today is the way it pits males against males in an endless cycle of dominance and violence, which Doug is only too happy to serve as a cheerleader for. Sad.

Posted by: Daniel | Sep 28, 2016 3:51:53 PM

How many incarcerated inmates are there because of corruption in the Court? If corrupt Courts are in Montague County, how many other counties are also corrupt?

Posted by: LC in Texas | Sep 28, 2016 4:13:37 PM

Daniel, what is "fantastical" about the factually true assertion that most serious/violent crimes are committed by young men? Are objective facts that you do not like to think about magically "fantastical"?

Also, what am I a "cheerleader" for? Modern American culture, which does not have a draft and thus does not think we have to conscript US males to trial to fight and kill non-US males, seems now much more critical of institutions that "pit[s] males against males in an endless cycle of dominance and violence." I certainly do not have an affinity male dominance and violence, and that is one reason (of many reasons) I am troubled to see Donald Trump succeeding by largely preaching these virtues.

Posted by: Doug B. | Sep 28, 2016 5:08:18 PM

Conscription may very well have something to do with the crime rate. I don't know that we'd want to make that hobson's choice however. During the decade long Viet Nam war, we sent about 3,500,000 to combat feeding what seemed to us like endless foreign engagement. We had 60,000 causalities. To put it in perspective, that's about 10X more casualties than our latest endless conflict in the middle east.

I didn't follow will's link, but I'm sure he's also on to something. When psychiatric drugs were developed and began to be prescribed we did empty mental hospitals. These were largely involuntary confinements but became very costly. I thought I read that we have reduced it by 90%.

Of course my solution is simple. End the war on drugs.

Posted by: beth | Sep 28, 2016 7:05:52 PM

"the factually true assertion that most serious/violent crimes are committed by young men?"

I have never seen a single study that provides any proof of this claim. The best I have seen is studies that seek /impute/ the committing of crimes based on arrest or conviction records. But given what we know about the intense and dramatic bias of police towards men, as evidence by the recent massive amount of killing of young black men, as evidenced by the very chart you use above, that connection is spurious. There is no good reason to think that either arrests or convictions represent anything like the real picture of who is actually committing serious crimes.

So yes I think the claim you make is fantastical and not a fact as you assert.

Posted by: Daniel | Sep 28, 2016 8:28:22 PM

Daniel, I am sympathetic to your concerns about how biases might impact whether arrests/convictions are a true reflection of true offending, especially for drug crimes and low-level offenses that a widely under-reported and under-prosecuted. But homicides are not widely under-reported or under-prosecuted, and here is BJS data on homicide offenses from 1980-2008: http://www.bjs.gov/content/pub/pdf/htus8008.pdf

The data as reported by BJS indicate that "Males represented ... nearly 90% of offenders" and also that about 65% of all offenders are between the ages of 18 and 34. Even if you think some bias finds expression in this data, I still think it provides considerable support for the "fact" that men between the ages of 18 and 35 committ more homicides than any other cohort.

Here is some BJS data on rape and sexual offenses: http://bjs.gov/content/pub/pdf/SOO.PDF

That publication details that "Nearly all arrestees for forcible rape in 1995 were male (99%), while about 8% of arrestees for other sex offenses were female" and that about 70% were between the ages of 18 and 39. Again, even if you think significant gender and age bias impacts these data (much of which results from victim reports, I think, and not just police investigations), it still seems pretty clear that men under 40 are in fact the cohort committing the most sexual offenses.

Question this data as you see fit, But I really think it disserves an honest discussion of reality to assert that the claim I have made about offender realities "is fantastical and not a fact."

Posted by: Doug B. | Sep 28, 2016 11:36:35 PM


This is a technique you use frequently in discussions. You start with a grand sweeping claim and when called on it you respond with much narrower claim or set of claims to buttress that sweeping claim. While homicide and violent rape are by all means serious crimes they by no means encompass the universe of serious crimes. I stand by my comment that you still have not produced data that shows that most serious crimes are committed by men, your data only indicates that a tiny subset of serious crimes are committed by men.

Posted by: Daniel | Sep 29, 2016 11:38:29 AM

If the statistics regarding incarceration essentially involve a snapshot of the percentage of persons incarcerated on a given day or during a given year, this may not reflect an increase in the behavior that is now subject to prosecution. Jurisdictions around the county have increased sentences and the type of behavior that can be prosecuted. While I do not have any data in front of me, I also suspect that there has been a trend of judges sentencing more harshly as compared to he 70s or 60s.

Posted by: Tim Holloway | Sep 29, 2016 12:11:36 PM

It also no doubt is reflective of charging decisions. We could have the same behavior demographic, but charging decisions escalate the level of incarceration.

Posted by: beth | Sep 29, 2016 2:50:28 PM

Daniel, my reference to "serious crime" when stating that the "vast majority of serious crimes are committed by men between the ages of 18 and 35" had in mind primarily homicide and rape, as those are the two major crimes historically and still today we consider to be the most serious crimes (at least for sentencing purposes).

What other crimes, Daniel, do you think also comfortably fit into the "serious crime" description. When you give me your list of serious crimes, I will find more data to support my assertion in light of your belief as to what are serious crimes. But just to spice things up and provide still more data for you to consume, here is a study about Robbery offenders in Croatia that notes as to "offenders' personal status, their age generally ranges from 22 to 40 years." https://www.ncjrs.gov/policing/cri359.htm And here is a study from Australia finding that 90% of armed robbers are male and "most ... are less than 30 years of age." http://www.aic.gov.au/media_library/publications/tandi_pdf/tandi328.pdf

If you want more, at least on the gender front, let's use 2011 FBI arrest data with help from Wikipedia:

Males constituted 98.0% of those arrested for forcible rape
Males constituted 89.0% of those arrested for robbery
Males constituted 85.0% of those arrested for burglary
Males constituted 83.0% of those arrested for arson.
Males constituted 81.5% of those arrested for motor-vehicle theft.
Males constituted 81.7% of those arrested for stolen property.
Males constituted 81.7% of those arrested for vandalism.
Males constituted 79.7% of those arrested for offenses against family and children.
Males constituted 77.8% of those arrested for aggravated assault.

Again, you are welcome to try to wish away evidence/data you do not like away (and, as I said, these kind of data on arrests certainly reflect a variety of socio-economic biases), but I think it important to try to make policy decisions and have policy analysis based on what evidence/data shows us, not on what we just wish and hope to be true despite significant evidence/data to the contrary. And there is significant evidence to support exactly what I said: "the vast majority of serious crimes are committed by men between the ages of 18 and 35."

Posted by: Doug B. | Sep 29, 2016 6:22:01 PM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB