When Numbers Lie, Kiddie Edition

Nearly one in three people will be arrested by the time they are 23, a study published Monday in Pediatrics found.


Astounding.  Horrendous. Inaccurate?

The study, published in USA Today, sounds yet another round of alarms that proves that one third of our children are violent predators, or we are outrageously over-reacting to childhood indiscretions, according to one’s predilections. 



The new data show a sharp increase from a previous study that stunned the American public when it was published 44 years ago by criminologist Ron Christensen. That study found 22% of youth would be arrested by age 23. The latest study finds 30.2% of young people will be arrested by age 23.


Criminologist Alfred Blumstein says the increase in arrests for young people in the latest study is unsurprising given several decades of tough crime policies.


So the culprit that has turned our babies into criminals is tough on crime policies?  That makes perfect sense, except for the fact that the numbers don’t seem to bear out.  Via Radley Balko at HuffPo :



Taken together, these studies and anecdotes suggest a troubling trend of putting kids in handcuffs for doing the sorts of things kids have always done. This has spurred concern over a burgeoning “school-to-prison pipeline” problem in which children — particularly poor, minority and at-risk children –are funneled from public schools into the criminal justice system. In response, the Justice Department and the Department of Education launched a joint initiative last July that aims to combat this trend.


One problem with the school-to-prison pipeline narrative, however, is that isn’t clear that it actually exists, much less that it’s getting worse. According to the Department of Justice’s Office of Juvenile Justice and Delinquency Prevention, 1.9 million minors were arrested in 2009. That’s down 17 percent since 2000. The 2.3 million arrested in 2000 was down 20 percent from 1996. Since 2000, juvenile arrests have also slightly declined as a percentage of total arrests, for both violent and property crime.


The problem with empiricism is that it’s all about numbers, even when they fly in opposite directions.  The problem with the adoration of empiricism (not that there’s anything wrong with that) is that it becomes a basis for action and reaction.  As if we needed more than “common sense” to rationalize policies that “everybody knows” are right even though there is no basis in fact for their creation and less obvious unintended consequences that “everybody” won’t notice until decades later.  Hey, thinking is hard and can make your head hurt.

My efforts to reconcile the conflicting numbers failed.  Whether the numbers of children arrested are up or down is a mystery to me.  It’s not that I doubt Blumstein’s study, but that Balko’s stats seem equally well grounded.  But not being slave to empiricism, there are conclusions that can nonetheless be drawn.

In our zeal to protect our children from every conceivable harm that might possibly befall them, we’re creating a new underclass of children whose conduct, a few decades ago, would have been characterized as normal, if occasionally rascalian, childhood behavior.  Now, it’s criminal, and its perpetrators are criminals too.  And this will limit everything they do for the rest of their lives.



State and federal governments have been prosecuting consensual crimes more aggressively, particularly drug crimes. And since the late 1990s, again with federal prodding, most public schools have embraced a “zero tolerance” policy for many offenses (drugs and weapons in particular), treating every infraction as if it were a criminal offense. The policy bars school officials from considering context or using nuance when dealing with an accused offender.


There are also more police in America’s schools. A survey by the Justice Police Institute released last month found a 37 percent increase in the number of law enforcement personnel (called Special Resource Officers, or SROs) employed by public schools between 1997 and 2007, including more than 5,000 such officers in New York City schools alone. The increase in SROs, also driven by federal funding, was in part influenced by media-driven hysteria over a few highly publicized school shootings in the 1990s.


There have also been a number of stories in the news of late about pre-adolescent children arrested for absurdly minor offenses, including a 6-year-old Wisconsin boy arrested for “playing doctor” with a 5-year-old girl, a 12-year-old arrested in Memphis for not wearing his helmet at a skateboard park, a 13-year-old boy arrested in New Mexico for burping in gym class (his parents’ lawsuit also revealed the arrest of a 7-year-old girl who refused “to sit next to the stinky boy” in class), a 10-year-old Connecticut boy arrested for giving a classmate a “wedgie,” and a 5-year-old who was bound at the wrists and ankles, arrested and charged with assault after kicking a police officer in the leg.


Remember those days back when, having some fun, blowing off some steam, risking life and limb playing with sticks?  You are today’s child criminal. Me too.  Thinking back, I feel fairly confident that I would be on death row for some of the shenanigans I pulled.  Of course, we used to call them shenanigans. Now, they’re Shenanigans in the First Degree.

The arguments made here (at great length and with some zeal) that the putative goal of creating a world where no harm, real or theoretical, can go uncriminalized and unpunished, is going to be the ruination of our children.  They need to skin their knee, eat dirt, play jokes and make mistakes.  Just like we did. We cannot assure every child the opportunity grow to adulthood without every feeling pain.

While every parent wants their child to reach maturity with only the happiest memories, it requires a great many other children to pay the price.  Whether this is one-third or less, and whether this comports with our baseless “common sense” adoption of every myth used to manipulate our emotions and limited grasp of reality, the consequence is that we are turning children into criminals over nothing in our efforts to stop every inchoate harm.

I like empiricism as much as the next guy, unless the next guy happens to be a statistician or lawprof in need of tenure.  But I’m more concerned for the welfare of children.  The gravest harm we can do to them isn’t ignore the potential for an occasional skinned knee or hurt feeling, but to construct an environment more inclined to have them arrested in the name of a perfect world than allow them to make the ordinary, normal mistakes from which they can learn to develop into mature, responsible, productive adults.

And that’s why the numbers tell me, whether their up or down. 

6 thoughts on “When Numbers Lie, Kiddie Edition

  1. John Neff

    FWIW the Iowa juvenile court services annual report has a five year comparison table and the juvenile complaints decreased by and average of 6% per year. This is another data point to support Radley Balko’s statistics.

    Arrests by age 23 is a different statistic that I do not think is very useful because about 85% of those arrested are only arrested once. What is so bad about their not coming back?

  2. SHG

    While it’s unclear why they chose age 23 as the cut off, I believe the point is that one third of those who are about to enter the workforce are saddled with a criminal conviction.  It’s not suggest they’re criminals, or bad people, but that they begin their adult lives tainted.

  3. Thomas Stephenson

    I can believe that a third of them will be arrested by the time they’re 23. But how many of those arrests will be for really minor stuff like alcohol violations?

    It reminds me of a statistic I saw that 60 percent of all students in Texas would be suspended from school at some point, as some sort of evidence we’re raising a bunch of future criminals. Except that they were actually referring to in-school suspension, which is normally the punishment for cutting class. 60 percent of high school students have cut class at some point? I’m actually surprised that number isn’t higher.

  4. John Neff

    I think they picked 23 as the cut off age because for adult alcohol related arrests the 18-20 age group is under the legal age to drink and the natural choice for a comparison group would be the 21-23 age group.

  5. Rick Horowitz

    Math is not my strong suit, but isn’t it possible for both these reports to be correct?

    One has to do with the number of arrests in a given year; the other, with the number of those arrested before a certain age.

    Just because the number of arrests — even over the last decade — are down, that doesn’t mean that it’s not possible for the number of those arrested before a certain age can’t be as reported.

    Does it?

    I mean, it might make it more difficult to buy the increase from 22 percent to 30-plus percent — and I already included my disclaimer — but I think it’s still possible for both to be right.

  6. John Neff

    The authors gave ranges for the accumulated percentages.

    16% to 27% by age 18
    25% to 41% by age 23

    When the corrected for missing cases
    they gave

    30% to 41% for age 23

    Newspapers tend to remove the range and give a single percentage that gives the false impression the results are exact.

    It is unlikely that the methods used in the older and newer studies were similar so I don’t think evidence for an increase is very good. It is possible that changes in arrest policies for juveniles could have increased the percentage but the data used in this study is not sufficient to confirm that supposition.

Comments are closed.