Youth crime: patterns, trends and disparities

Bell begins the second class with a question: 'Is youth crime getting worse?' Most of the students say yes. Several mention news stories about teenagers and car thefts, a video of a fight outside a mall that went viral, and a sense that young people today are more violent. Bell nods. 'That is what most people believe, in almost every generation,' he says. 'Let's look at what the numbers actually show, and then at what they can't show.' This lesson covers how youth crime is measured, what the long-term trends are, the kinds of offenses young people commit, and the disparities that run through youth justice systems in both countries.

How youth crime is measured

There is no single count of youth crime. Each source captures a different slice. Arrest statistics, collected by the FBI from local police through the Uniform Crime Reporting program, count arrests of people under 18; they count arrests, not people, and they depend on which agencies report. Juvenile court statistics, compiled nationally by the National Center for Juvenile Justice for the Office of Juvenile Justice and Delinquency Prevention (OJJDP), count cases referred to juvenile courts and how they were handled. Custody censuses count young people held in facilities. Self-report surveys ask young people directly what they have done, and the National Crime Victimization Survey asks people what has happened to them.

Each source is shaped by decisions made by people. Police-reported numbers depend on what victims report and what police record. If a police service decides to deal with more incidents informally without recording them, the numbers fall even if behavior has not changed. If schools start calling police about fights they once handled themselves, the numbers rise. Self-report surveys consistently show that most adolescents admit to some minor offense, such as shoplifting, vandalism or a fight, at some point, which reminds us that the youth who come to police attention are a small and not entirely random fraction of those who offend.

Long-term trends

Arrests of juveniles in the United States peaked in the mid-1990s and then fell dramatically. OJJDP reports that the number of arrests involving youth fell by about 84 out of every 100 between 1996 and 2020, and that youth arrests for violent crime in 2020 were about one-third of the number in 2006. Some categories, including motor vehicle theft and some firearm offenses, rose in particular cities after 2020, which drew wide media attention, but the long-term decline in youth arrests is one of the largest changes in American criminal justice. Juvenile court caseloads and the number of youth in custody fell along with arrests.

Bell tells the class to be careful with short-term changes. A single year's increase in one offense, in one city, can be real and important, but it does not reverse a twenty-year trend. And a rate is more informative than a count, because the number of young people in the population changes. Desjardins adds that the fall in youth crime has many proposed causes, and researchers disagree about their relative weight: changes in policing and diversion, reduced lead exposure, changes in drug markets, more time spent online and less unsupervised time on the street, better school programs, and changes in how incidents are recorded.

# Rates per 100,000 let us compare places or years with different populations.
# Port Ardent figures are invented for teaching.
youth_pop = {"2016": 11200, "2021": 11850, "2025": 12400}
accused = {"2016": 512, "2021": 398, "2025": 341}
for yr in youth_pop:
    rate = accused[yr] / youth_pop[yr] * 100000
    print(yr, "rate per 100,000:", round(rate))
change = (accused["2025"] / youth_pop["2025"]) / (accused["2016"] / youth_pop["2016"])
print("2025 rate as a share of 2016 rate:", round(change, 2))

2016 rate per 100,000: 4571
2021 rate per 100,000: 3359
2025 rate per 100,000: 2750
2025 rate as a share of 2016 rate: 0.6

In Port Ardent's invented figures, the number of young people accused fell from 512 to 341 while the youth population grew, so the rate fell even faster than the count: the 2025 rate is about 0.6 of the 2016 rate. Bell notes that the same calculation is the first thing he does when a council member says youth crime is 'out of control' in a particular neighborhood.

What young people are accused of

The offenses that bring young people to police attention are mostly the less serious ones. Theft under a modest value, mischief and vandalism, minor assaults, possession of stolen property, breaches of court orders and disturbances make up a large share of youth police contacts in both countries. Serious violence, including robbery, assault with a weapon, sexual offenses and homicide, is a small fraction of youth offending, though it receives most of the public attention and rightly most of the system's resources.

Several patterns distinguish youth offending from adult offending. Young people offend in groups much more often than adults; a large share of youth offenses involve two or more young people, and peer influence, as lesson 1 explained, is a powerful driver of adolescent risk-taking. Youth offending is concentrated in time and place: after school hours on weekdays, especially between about 3 p.m. and 6 p.m. when young people are unsupervised, and near schools, transit hubs, malls and parks. Many youth victims are other young people. Ben Arsenault and the two teenagers who robbed him all lived within a kilometer of Eastgate Park and had at times attended the same school.

Breaches of court orders, called probation and court order violations, and in some states violations of valid court orders by status offenders, have been a particular concern. A young person released with a long list of conditions, such as a curfew, a requirement to attend school every day and a ban on contact with friends, may breach one of them without committing any new harm, and end up back in custody. The federal Juvenile Justice and Delinquency Prevention Act restricts detaining status offenders, and after 2018 limits detention for violating a valid court order to seven days with specific court findings.

Disparities in youth justice

Desjardins turns to the most difficult part of the lesson. In both countries, young people from some groups are much more likely than others to come into contact with police, to be charged rather than diverted, to be detained before trial, and to be placed in custody. Black youth are the clearest example in the United States. National data show that Black youth are arrested, detained, formally processed and placed in residential facilities at rates several times higher than white youth, and the gap is often wider at the deeper stages of the system. Native American and Latino youth are also overrepresented in some states and stages. The federal Juvenile Justice and Delinquency Prevention Act requires states to address racial and ethnic disparities, and the 2018 reauthorization requires them to use data to identify disparities, set measurable goals to reduce them and report publicly.

Researchers have identified several causes, and they are not mutually exclusive. Some disparities reflect differences in exposure to risk factors that are themselves the product of history and policy, such as poverty, unstable housing, under-resourced schools and, for Black youth, the legacy of segregation and concentrated poverty. Some reflect where police are deployed: more officers in a neighborhood means more minor offenses come to official attention there. And some reflect decisions at each stage, including whether an officer warns or charges, whether a young person is seen as having a supportive family, and whether a court finds a suitable community placement. Small differences at each decision point can add up to large differences at the end.

Ohio courts and the Ohio Department of Youth Services participate in the federal disparities requirement, and several Ohio counties, including Lucas County, have taken part in the Juvenile Detention Alternatives Initiative, which uses data to reduce unnecessary detention and the racial gaps within it. Bell gives an example from his own unit. A few years ago Port Ardent reviewed its youth files and found that, for similar first-time shoplifting incidents, white youth were warned more often than Black and Indigenous youth, who were more often referred to a formal program or charged. The difference was partly explained by officers' assumptions about which parents would 'deal with it'. The unit changed its practice so that every first-time minor theft starts with a warning or referral, and officers must write a reason for any departure.

The limits of the numbers

Desjardins closes with a caution. Statistics tell us about patterns, not about the young person in front of us. They can help a service see where its decisions are uneven, where resources are needed, or whether a program is working. But an officer who treats a young person as a statistic, assuming that because they come from a particular neighborhood or group they are likely to reoffend, commits the very error that produces disparities. Good youth policing uses data to examine the system and judgment, informed by the facts of the case, to deal with the individual.

Girls, gender and youth offending

Boys make up most young people accused of offenses in both countries, and the gap is largest for serious violence. But girls' involvement has its own patterns that officers should understand. Girls who come into contact with police are more likely than boys to have experienced sexual abuse, family violence and running away from home, and some of their offending, such as assaults on family members during conflicts at home, grows directly out of those circumstances. Research has also found that girls have sometimes been held in custody or detention for behavior, such as breaching curfew or running away, for which boys were less often detained, on the theory that custody would protect them. The federal rules on status offenders and Ohio's juvenile rules were written partly to stop that practice. Bell's advice is simple: ask what is happening at home, and treat a girl who has run away from home as a possible victim first.

What the numbers miss

Some youth offending is badly measured by every source. Offenses that happen online, such as sharing intimate images without consent, threats on social media and online fraud, are underreported, and when they are reported it may be hard to identify the offender's age. Offending within families and among friends is rarely reported to police. And many of the most serious harms to young people, such as sexual exploitation, are offenses committed against them by adults, which youth crime statistics do not count at all; lesson 10 returns to this. Desjardins tells the students to treat every number in this lesson as a floor for what happens, not a ceiling, and to remember that an increase in reporting can be good news if it means victims trust police more.