Lord of the Flies—that massively influential staple of high school English classes—concludes with a scene of mutual incomprehension. When the schoolboys are finally rescued from the island where they have been marooned, the naval officer who discovers them can’t fathom how they descended to fratricide. It seems to him that they had a great adventure. Meanwhile one of the boys, Ralph, weeps for “the end of innocence” and “the darkness of man’s heart.” He can’t see the glamour in any part of what he and the other boys have been through. The reader is on Ralph’s side, of course; the naval officer’s perplexity underscores the book’s central lesson. Exactly what that lesson is remains controversial, but it plainly has something to do with violence and how a group of children—or at least a group of boys—can turn savage. The officer’s bafflement is connected to the novel’s perennial appeal. It’s not incidental that the castaways are young people. There is something about violence among juveniles that seems particularly shocking, or particularly telling, or both.1
Just as the legal response to rape and domestic assault reflects ideas about violence as well as ideas about sex and gender, the legal response to youth violence has been shaped both by beliefs about children and adolescence and by ideas about the nature, significance, and definition of violence. The legal response to youth violence, and to juvenile offending more generally, has changed significantly over the past several decades. Those changes and the lessons they offer are the subjects of this chapter.
For most of the twentieth century, minors who committed crimes were tried and sentenced not in regular criminal courts but in separate juvenile courts, and they were incarcerated, if at all, not in regular prisons but in separate juvenile facilities, and typically for less time than adult offenders. Juvenile courts were invented in Illinois at the close of the nineteenth century, and the innovation soon spread not only throughout the United States but around the world. These courts aimed not at retribution but at rehabilitation, or at least at giving young offenders a chance to mature out of delinquency. In the 1980s and 1990s, though, many juvenile offenders, especially those who had committed “violent” crimes, were prosecuted as adults and sent to adult prisons, for decades or for the rest of their lives. At the same time, schools around the country announced “zero tolerance” for violations of disciplinary rules, particularly rules against violence. Instances of physical bullying that were once dismissed as age appropriate, at least for boys, now became grounds for suspension. Like the prosecution and punishment of minors as adults, the more aggressive use of school disciplinary sanctions was part of what law professor Barry Feld calls the “Get Tough Era” of juvenile justice. The policies implemented in that period were fueled in large part by an increase in violent crimes carried out by minors, and by the fears this fed of “superpredators” and a coming “bloodbath of violence.”2
In one sense that era has never really ended: many of the policies enacted in the 1980s and 1990s are still on the books. But beginning in the early 2000s, courts and legislatures began to sour on the get-tough approach to juvenile offending. This happened for several reasons. First, and probably most important, rates of youth violence declined sharply; the generation of superpredators never materialized. Second, advocacy groups brought increasing attention, not just to the harshness of the new policies, but to the glaring racial disparities they had produced. The tougher sanctions were imposed disproportionately on minority boys, and to a lesser extent on minority girls. Zero-tolerance polices had created a “school-to-prison pipeline,” and the children and adolescents sent into the pipeline were, again, disproportionately students of color, particularly boys. Third, scientists found new evidence that human brains are not fully developed until early adulthood, and that the parts of the brain responsible for higher-order thinking and self-control are the last to conclude their maturation. Fourth, partly relying on these developments in brain science, the Supreme Court ruled in a series of cases that the Constitution did not allow the imposition of the death penalty for juvenile offenses, did not permit sentences of life without possibility of parole for non-homicide crimes committed by juveniles, and prohibited the mandatory sentencing of juveniles convicted of homicide to life without possibility of parole. Feld and other scholars of juvenile justice credit these decisions with sparking a renewed appreciation, at the state and local levels, that juvenile offenders are kids, and that “kids are different.”3
The embrace of get-tough policies for juvenile offenders in the 1980s and 1990s, and the partial retreat from those policies in the years since, is partly a story of ideas about children and adolescents. The superpredator scare fed on anxieties about monstrous children, long a staple of popular entertainment. Later, scientific research about changeable adolescent brains rang true with courts and legislatures because it resonated with common experience. When, for example, the Supreme Court was considering the constitutionality of the death penalty for juvenile offenses, Justice Breyer asked at oral argument whether the new scientific evidence “simply corroborated something that every parent already knows”—namely, that teenagers’ personalities are not yet “fully formed.” Solicitor General Seth Waxman agreed that the research “explains, corroborates, and validates what we sort of intuitively” knew about adolescents.4
But if the shifting responses to juvenile offending have been shaped by ideas about childhood and adolescence, they also have reflected and reinforced particular beliefs about violence. The get-tough era was fueled in part by the notion that violence changes everything: that the juvenile justice system was designed for “truants, vandals, and petty thieves” and was not suited for “the violent juvenile offenders of today.” There was a sense, too, that “the violent juvenile offenders of today” weren’t really juveniles, or at least they weren’t children, regardless of their age. Defending the decision to prosecute an 11-year-old murder defendant as an adult, a Chicago prosecutor explained in 1977 that “he may be 11 chronologically,” but “he’s a violent, violent person.” That idea—that violent crimes should be handled in criminal court, not in juvenile court—remains common, although not as common as it once was. In 2019, Oregon made it easier for juvenile court judges to keep jurisdiction over teenagers charged with violent crimes, instead of sending them to the adult system. Prosecutors opposed the reform, asking rhetorically—in the words of one district attorney—“which murder, which violent rapist, which violent robber should be dealt with in the juvenile system”?5
As elsewhere in the law, the more significant violence becomes, the more it matters how violence is defined. In the same way that the stiff mandatory sentences targeted at adult offenders gradually expanded their footprint as more and more crimes were labeled violent, there was a steady expansion of laws allowing minors to be prosecuted and punished as adults as mounting categories of juvenile offending were moved into the “violence” box from the “truancy, vandalism, and petty thievery” box. Something similar happened, too, with disciplinary codes in schools, as the scope of zero-tolerance policies expanded.
Changes in approaches to juvenile offending have also reflected changes in ideas about the nature of violence, and in particular the nature of juvenile violence. The superpredator scare of the 1980s and 1990s was an extreme example of a theory of violence as dispositional rather than situational. The retreat from get-tough policies in juvenile justice has been driven in part by renewed appreciation for the ways in which adolescent violence, like all adolescent offending, can reflect circumstances rather than deep-seated character. The impulsivity of teenagers, and their susceptibility to peer influence, make situational explanations for violence more plausible, as does the fact that so many kids age out of offending.6
Race, ever present in discussions about violence, has played a particularly large role in the shifting understandings of youth violence. Children and adolescents of color, particularly when they are implicated in acts of violence, are far more likely than white youth to forfeit their status as minors. This happens in a literal, legal sense when juvenile offenders are prosecuted and punished as adults—something that happens much more often to minority youth. But it also happens at a more fundamental, psychological level. Race plays a troubling, outsized role in determining whether a minor who commits an act of violence is seen first and foremost as a minor or as a violent offender: as a child or as a superpredator. And on a larger scale, the degree to which juvenile violence is understood as a Black or brown problem has strongly influenced when the problem has been seen as a societal failing, and when as a threat to be contained. In the 1950s, for example, when four white teenagers in Brooklyn were prosecuted for a series of shockingly brutal crimes—which including killing two men for the thrill of it, and kidnapping and whipping two girls—the predominant public reaction was to wonder about the social roots of youth violence. As we will see, attacks attributed to Black teenagers in later decades elicited a very different reaction.7
This chapter will focus largely on the legal response to violence committed by minors, but we will give attention as well to ideas about violence against minors. Corporal punishment of children, at home and at school, is less common than it used to be. The tide of opinion, both popular and elite, is running against these practices. But physical beatings of children, whether by their parents or by school officials, are hardly things of the past. Corporal punishment in public schools remains legal in a significant minority of states. Every year tens of thousands of children are paddled or beaten by educators, with the full sanction of the law. The subjects of these beatings are disproportionately minority or disabled or both. And while more than sixty countries have outlawed spanking by parents, no American jurisdiction has followed suit, and none appears likely to do so. As a rule, only the gravest assaults on children by their parents or guardians—assaults that result in death or serious bodily harm, assaults that appear sadistic in their infliction of pain or degradation, or assaults that are sexual in nature—are treated as crimes in the United States.8
Police officers, in contrast, have no special dispensation to use force against children and adolescents. But neither are there any targeted legal restrictions on officers grabbing, striking, or otherwise assaulting minors. The only rule, for the most part, is the general constitutional ban on seizures that seem under all the circumstances to be “unreasonable.” As we saw in Chapter 3, the law of criminal procedure does not distinguish, in general, between violent and nonviolent forms of police misconduct; violence does not have the same significance here that it does in substantive criminal law. One could imagine treating violence against children as different, either for humanitarian reasons or to avoid turning the children themselves violent. But just as the law draws no general distinction between violent and nonviolent police conduct, law enforcement officers who use physical force against minors are not subject to any special rules or any heightened scrutiny. If anything, physical force against minors is likely to be more readily approved by reviewing courts, partly because young people can seem more threatening—stronger and less restrained—than their elders, and partly because of the centuries-old assumption that children and adolescents often need physical discipline.
Youth violence and violence against youth are typically treated as two separate problems, but that may be a mistake. One of the reasons spanking at home and at school has become less popular around the globe is the strong consensus that has emerged among experts on child development that children who are physically disciplined are more likely later to be violent themselves. Scholars have warned that police violence directed at minors can operate similarly—that young people who are physically assaulted by law enforcement officers are more likely to use violence against others.9 That is one reason to study, side by side, the way the law thinks about violence by young people and the way it thinks about violence against young people. The other reason is more basic, and it is a set of beliefs that runs throughout this book: that ideas about violence are more complicated and more contingent than we often realize; that ideas about violence, when employed inconsistently, can mask racism and other toxic biases; and that it therefore behooves us to be as thoughtful and reflective as we can be when we talk about violence.
Juvenile justice, as a separate field of jurisprudence and a separate system of social institutions, is more than a century old. It was invented, more or less, at the close of the nineteenth century when progressive reformers opened the Cook County Juvenile Court, directly across the street from Hull House, Chicago’s first and most influential settlement home. There had long been correctional institutions designed especially for minors, but the juvenile court was something new; it reflected the belief that children and adolescents who violated the law should be held to different standards and be handled with different procedures than adult offenders. The creators of juvenile courts couched the mission of the new institution in the language of rehabilitation. Instead of punishment, juvenile justice would aim at “saving” children and “making them honest citizens” instead of criminals—or, as Judge Julian Mack put it in an early manifesto for the new courts, “not so much to punish as to reform, not to degrade but to uplift, not to crush but to develop.” The law professor and criminologist Franklin Zimring argues that, in practice, juvenile courts rarely placed much emphasis on rehabilitation, in the sense of affirmative efforts to improve an offender’s character. Instead, he suggests, the raison d’être of juvenile courts has always been simpler: to keep juvenile defendants out of the adult criminal justice system, where they would likely harden into more serious offenders.10
Either way, the notion that juvenile offenders should be adjudicated by separate courts applying separate standards proved remarkably popular. It quickly spread across the United States and around the world. Zimring claims plausibly that no American legal innovation has achieved greater global popularity.11
In the last decades of the twentieth century, however, the United States dialed back its commitment to this innovation. Juvenile courts were never abolished, but they were given less to do; many of the cases they previously would have handled were sent instead to regular criminal courts. From their inception, juvenile courts always had the ability to send particular cases to the adult system. This authority typically was reserved for extreme cases, involving especially serious crimes committed by adolescents on the threshold of adulthood. Most violent crimes did not qualify. In the early twentieth century, even homicides were often handled in juvenile courts, at least in some jurisdictions.12
This changed in the closing decades of the twentieth century, particularly in the 1980s and 1990s. New laws throughout the United States made it easier to transfer juvenile offenders to regular criminal court, or to handle them there from the outset. Some states lowered the age at which juveniles could be transferred to the adult system, some states—a majority, eventually—mandated transfer for certain offenses, and some states adopted “direct file” laws giving prosecutors the discretion whether to charge certain crimes in juvenile court or adult court. Furthermore, some states—again, eventually a majority—declared that once a juvenile had been convicted in adult court, any future charges against that defendant would also be handled in adult court. Partly as a result of these changes, the number of offenders under age 18 admitted to state prison more than doubled between 1985 and 1997. The rate at which arrestees under 18 wound up being sent to state prison also nearly doubled. And many of the juveniles tried in adult court were punished pursuant to harsh new sentencing laws adopted in the 1980s and 1990s, the laws we examined in Chapter 2.13
The major impetus for redirecting so many juvenile offenders to adult court at the end of the twentieth century, and for the mandatory sentencing laws enacted during the same period, was nationwide alarm about increases in violent crime, and youth violence in particular. In the late 1980s, rates of adolescent homicide in urban areas skyrocketed, at a time when homicide rates among Americans aged 25 and older were declining and the number of older adolescents in the United States was also declining. Several academics concluded that a new kind of juvenile offender had materialized and that the situation was likely to get much worse as the aging of the “echo boom”—the children of baby boomers—increased the raw number of adolescents in the United States. The political scientist John DiIulio warned in an influential article that we were about to see an explosion of “juvenile super-predators”: “super crime-prone young males,” “remorseless” and “radically self-regarding,” who “place zero value on the lives of their victims” and who, “as long as their youthful energies hold out” will “murder, rape, rob, assault, burglarize, deal deadly drugs, and get high.” The criminologist James Fox predicted that when the young children of the 1990s became adolescents, the United States could face “a bloodbath.” Americans, he said, were “sitting atop a demographic time bomb.”14
Politicians latched onto these fears. Florida congressman Bill McCollum, for example, warned in 1996 that the “predators out there” weren’t “children anymore”; they were “the most violent criminals on the face of the earth.” Campaigning for her husband’s reelection that same year, Hillary Clinton said—in remarks that later came back to haunt her—that “the kinds of kids that are called super predators” had “no conscience” and “no empathy,” and that before worrying about “why they ended up that way … we have to bring them to heel.”15
By 1996, though, rates of youth homicide actually had begun to decline, and the rates continued to plummet without interruption for the next ten years, falling by roughly 75 percent. The time bomb never went off. In 2001 a report by the US surgeon general called superpredators a “myth” and concluded that “there is no evidence that young people involved in violence during the peak years of the early 1990s were more frequent or more vicious offenders than youths in earlier years.” DiIulio confessed error and regretted that the superpredator theory had proven difficult to retract. “Once it was out there,” he lamented, “there was no reeling it in.”16
How had the predictions been so wrong? Part of the story might be changed circumstances: scholars have variously credited the end of the crack epidemic and its notoriously violent turf battles between rival dealers, the success of environmental regulations that lowered levels of lead in the environment, and the decline in unwanted births that followed the legalization of abortion. Another part of the story, though, was sloppy statistics in the first place. DiIulio and Fox ignored reasons to suspect that the abnormally high rates of youth homicide witnessed in the early 1990s did not, in fact, reflect any underlying change in the characteristics of young offenders, and were likely to regress to historical averages. One of those hints was that the increase in youth homicide in the 1980s and 1990s was not accompanied by rises in other measures of youth violence, and it was entirely an increase in gun homicides. Rates of youth homicide by other means stayed relatively constant, and so did juvenile arrest rates for rape and robbery. Juvenile arrest rates for assault did increase in the late 1980s and early 1990s, but so did the corresponding arrest rates for adults, and in both cases the change appeared to be a statistical artifact: police departments appear to have lowered the thresholds for what counted as simple or aggravated assault.17
The superpredator theory proved so popular, and so difficult to reel in, because it resonated with what people were ready to believe. It felt right. Rates of juvenile homicide had, in fact, soared in the 1980s and early 1990s, and as always the killings that received the most attention were the most sensational. It was easy to believe that the killers, regardless of their chronological age, were not really “kids.” Doubtless this was all the easier because the spike in juvenile homicide, as DiIulio and Fox pointed out, was particularly dramatic in Black, inner-city neighborhoods. Calling the perpetrators “superpredators” turned them into—as Zimring puts it in his retrospective critique of this period—“a separate (and essentially nonadolescent) species.” There is something of a tradition of this kind of linguistic transformation. As the criminologist John Laub has pointed out, it is unclear how, if at all, the “superpredators” of the 1990s differed from “the ‘serious, violent juvenile offenders’ of the 1970s and 1980s and the ‘chronic offenders’ of the 1960s and 1970s.” Zimring agrees. Whenever youth violence appears to be increasing, he concludes, “there is worry that a new, more vicious type of juvenile offender is the cause.”18
In effect, the superpredator theory distinguished violent juvenile offenders taxonomically, not just from normal children and adolescents, but from juvenile offenders of the past. We were dealing, explained a Chicago prosecutor in 1994, with “a whole new breed.” The case for taking them out of juvenile court seemed obvious. Governor Pete Wilson of California voiced the new common sense in 1997: “Juvenile court was not designed to deal with youths who commit serious and violent crime, gang murderers, rapists, carjackers, or kids who carry assault weapons.”19
“Adult crime, adult time” went one popular slogan; another formulation was “old enough to do the crime, old enough to do the time.” Jerome Irons was tried as an adult in Missouri in 1998 and sentenced to fifty years for a burglary and armed assault committed when he was 16; the prosecutor said he wanted to “send a message … that if you are going to act like somebody old, you are going to be treated like someone old.” The following year Anthony Laster, a mentally disabled 15-year-old, was charged as an adult in Florida for grabbing $2 of lunch money from a classmate; the state attorney defended the decision on the ground that any “violent juvenile crime” should be treated as an “adult crime committed by a juvenile.”20
“Adult” crime thus was often understood to mean “violent” crime. A spokesman for the state attorney who charged Anthony Laster made this explicit: the office drew the line, he said, “where there is a threat of physical violence.” The following year, when the same state attorney charged a 13-year-old with first-degree murder, he defended his decision on the ground that “murder is an adult crime just being committed by someone in a 13-year-old body.” Similarly, when legislators in the 1980s and 1990s began to make it easier, and in some cases mandatory, to prosecute juveniles as adults, the principal focus was on violent crimes. Inevitably, though, the category of “adult” crimes expanded, and by the end of the 1990s many states had either categorically removed from juvenile court any minors charged with certain property and drug offenses or had given prosecutors discretion to charge these cases in criminal court. Nearly 40 percent of all juveniles tried as adults in the nation’s largest counties in 1998 were charged with property, drug, or public order offenses. More than three-quarters were Black or Latino.21
The understanding of children and adolescents as superpredators did not only provide the basis for prosecuting minors as adults; it also influenced the sentences handed down. Moving minors to adult court exposed them to adult sanctions, including, in some cases, long mandatory minimum terms, life sentences without possibility of parole, and the death penalty. Judges typically had discretion in sentencing juveniles who had been charged and convicted as adults, and it might be expected that youth would be treated as a mitigating factor. That undoubtedly happened in many cases. Overall, though, there is considerable evidence that in the last decades of the twentieth century, youth became an aggravating factor in criminal sentencing. Juveniles tried as adults received, on average, significantly longer sentences than older defendants convicted of similar offenses.
Some of this may have been due to a selection effect: when judges or prosecutors had discretion whether to keep a case in juvenile court or transfer it to adult court, the cases transferred tended to be the ones that elicited less sympathy and more outrage. But even after controlling for case characteristics, social scientists found that juveniles tried as adults in the 1980s and 1990s received harsher sentences than older defendants. “By the mid-1990s,” conclude two legal scholars who have studied juvenile homicide prosecutions in this period, “youth had ceased to be a mitigating factor in adult court, and instead had become a liability.”22
Youthful defendants were, and often still are, treated more harshly for at least three reasons. The first reason is actuarial: young offenders simply have more years of potential offending ahead of them. This is why the sentencing guidelines in some states assign young offenders extra points toward the scores that are then used to calculate their presumptive sentences.23
The second reason is that when adolescents commit crimes, they tend do so in groups, not acting alone. That can be taken as a mitigating factor. In fact, when the Supreme Court decided in the early twenty-first century that crimes by adolescents called for special treatment under the Constitution, part of the Court’s reasoning was that because teenagers were more susceptible to peer pressure, especially harsh sentences made less sense for them. Not coincidentally, the three cases in which the Supreme Court drew this distinction all involved crimes committed by more than one offender. (More about these cases later.) But criminal law has a long tradition of treating groups of offenders acting together as especially threatening and warranting special sanctions. Moreover, much of the panic about youth violence in the 1980s and 1990s focused on frightening combinations of teenagers: either organized gangs, or “wolf packs” engaged in chaotic “wilding.” Kamala Harris, then district attorney of San Francisco, spoke for a generation of law enforcement officials in 2009 when she called gangs “our most serious threats”: a “ruthless, toxic … form of social cancer.”24
There is reason to doubt whether many gangs were as organized or as powerful as law enforcement—and often gang members themselves—made them out to be. It is hard here to separate myth from reality, and hard to know when “gang related” crimes were actually caused by gangs—when, for example, gangs were making their members violent, as opposed to attracting members who were already violent. Nonetheless, in the closing decades of the twentieth century, anti-gang legislation was passed in all fifty states and at the federal level. Led by California, most states adopted special, heightened penalties for crimes committed by gang members or affiliates—who were, disproportionately, adolescents or young adults. Many, if not most, of those laws are still on the books. In addition to the formal sentencing enhancements available in gang cases, the mere introduction of evidence that the defendant was associated with a gang biased juries in favor of conviction, and may have biased sentencing judges as well. That, too, remains true today. Furthermore, when juveniles convicted of gang crimes, or suspected of gang affiliation, were released on probation or parole, they often were barred from associating with gang members or affiliates; violating those rules could send them back to jail. Again, release conditions of this kind are still commonplace.25
There is a final reason youth has been and often still is an aggravating factor in criminal sanctioning. It is the same intuition that made the superpredator theory seem so plausible: the idea that minors who commit serious acts of violence must have something fundamentally wrong with them, that youth violence is particularly likely to be characterological rather than situational. It was clear to Governor Wilson that juveniles responsible for “serious and violent crime” weren’t just making mistakes, they were “mak[ing] mistakes a lifestyle.” When the state of Missouri sought the death penalty in 1994 against 18-year-old Christopher Simmons for a murder he had committed when he was 17, the prosecutor scorned the suggestion that youth was a reason to spare the defendant’s life. “Think about age,” he urged the jury. “Seventeen years old. Isn’t that scary? Doesn’t that scare you? Mitigating? Quite the contrary I submit. Quite the contrary.”26
The jury may have agreed: Simmons was sentenced to death. A decade later the Supreme Court invalidated his sentence on the grounds that capital punishment for offenses committed by juveniles, even for murder committed by juveniles, was unconstitutionally “cruel and unusual.” That decision, as we will see, reflected a significant shift in the understanding of juvenile violence—a return, in a way, to older ideas. Why did it take so long to arrive? One reason is school shootings.
In the middle of the 1990s—right around 1994 or 1995—rates of youth violence in the United States began to decline sharply. By the end of the decade they had returned to the levels of the early 1980s. During the same period, though, there was an increase in what the FBI calls “active shooter” incidents at elementary and secondary schools: episodes in which someone with a gun fires indiscriminately in a populated area in an effort to kill people. Most of the shooters were students or former students at the schools in which the rampages took place. Incidents of this kind remained relatively uncommon, and they were a very small fraction of juvenile homicides, but they received a great deal of media attention—partly because the school setting and the mass, indiscriminate nature of the killings were particularly horrifying, and partly because the rampages occurred in places, and among populations of young people, not usually victimized by gun violence. All of this was true of the most notorious school shooting of the 1990s, the 1999 massacre of twelve students and a teacher at Columbine High School, in the suburban, middle-class, largely white community of Littleton, Colorado. At the time, Columbine was the deadliest school shooting in the nation’s history.27
One thing the school shootings of the late 1990s did have in common with the surge in youth homicide in the late 1980s and early 1990s was the involvement of guns. Virtually all of the increase in juvenile homicide in the 1980s and 1990s, and virtually the entire drop in juvenile homicide in the late 1990s, involved gun homicides. The killings at Columbine sparked a nationwide debate about gun control, but no major legislation emerged from that debate. Instead, the largest impact that school shootings had on public policy in the late 1990s and early 2000s was to reinforce the kinds of concerns that characterized the superpredator scare: concerns that there was something new and different about contemporary adolescents, that we were seeing a new, more violent breed of juveniles. As a result, the drive to prosecute more minors as adults continued well into the 2000s, long after rates of juvenile violence had plummeted and the superpredator theory had been discredited.28
The school shootings of the 1990s also prompted new worries about the physical safety of schools. Those worries, in turn, were addressed in two main ways. The first was a “hardening” of schools: new metal detectors, armed guards, security cameras, and locked gates.29 The second was a crackdown on violations of school disciplinary rules, including bullying.
There is a very long tradition of treating assaults and physical threats among schoolchildren and adolescents, particularly boys, as an inevitable part of growing up, and possibly even salutary. Part of growing up was learning to take care of yourself, learning not to be a sissy. (That is the idea that the radio host Laura Ingraham was invoking when she taunted a reporter who had complained about being slammed to the floor and punched by a Republican congressional candidate: “Did anyone get his lunch money stolen today and then run to tell the recess monitor?”30) Beginning in the 1980s, though, researchers and policymakers began to take bullying seriously. They began to treat it as harmful—among other things, bullied children have higher rates of suicide and a range of mental and emotional disorders—and as a problem to be controlled. These research and policy initiatives began in Scandinavia and were slow, initially, to reach the United States. But concerns about school bullying in the United States rose sharply after Columbine, driven by widespread reports (later contested) that the killers had been bullied.31
Not all bullying involves violence. Researchers and policymakers also apply the term to patterns of verbal or “relational” aggression among children or adolescents, in person or over the internet. Physical bullying is more common among boys; girls are more likely to engage in verbal or relational bullying. The good news is that all forms of bullying have decreased over the past twenty years, in the United States and overseas. Much of the success may be attributable to anti-bullying curricula and targeted counseling strategies that have been widely adopted in schools.32
Concerns about bullying in the wake of the Columbine shootings also led, though, to policies that proved less successful and in many ways counterproductive: zero-tolerance approaches to school discipline. The language of zero tolerance was first applied to narcotics enforcement strategies in the 1980s, in schools and elsewhere. The idea was that any violation of drug laws, no matter how minor, would be punished. Depending on the context, the punishment could take the form of criminal prosecution, impoundment of sea vessels, or suspension or expulsion of students. Federal legislation in 1994 borrowed the concept of zero tolerance and applied it to school violence, presumptively requiring schools to adopt policies calling for the automatic expulsion of any student found with a gun on campus. By the early 2000s, schools throughout the United States had committed themselves to automatically expelling or suspending students who violated a significantly broader range of disciplinary rules, often including bullying and physical aggression in addition to possession of guns or drugs. Once tolerated as something of a rite of passage, physical fights could now be grounds for mandatory suspension.33
Some of these disciplinary policies are still in place, but many have been repealed. Beginning sometime around 2010, a consensus began to form among educators and policymakers that zero-tolerance policies were a mistake. These measures had caused suspensions and expulsions to surge without any detectable effect on school safety. And once students were removed from class, even temporarily, their chances of graduating plunged and the odds dramatically increased that they would find themselves entangled in the criminal justice system. Zero-tolerance policies helped to create, in effect, a school-to-prison pipeline, and a heavily disproportionate number of the students sent into the pipeline were Black or Latino.34 In 2011 the Obama administration launched a “Supportive School Discipline Initiative,” explicitly aimed at disrupting the school-to-prison pipeline by addressing “inappropriate school discipline.” In 2014 the US Department of Education directed schools to suspend or expel students only as a last resort, and to guard against racial discrimination in the use of those measures. By then there was a broad, nationwide trend toward reducing the number of suspensions and expulsions, with only a few jurisdictions moving in the opposite direction. Even charter schools—many of which had made zero tolerance for disciplinary violations part of their brand—began to abandon that approach.35
That trend appears likely to continue, notwithstanding the political developments of the past several years. In 2018 the Department of Education under President Trump repudiated the guidance it had provided to schools in 2014. The new administration suggested that lax school discipline and excessive concerns about avoiding racial discrimination might be responsible for school shootings, including the massacre of seventeen students and staff members at Marjory Stoneman Douglas High School in Parkland, Florida, in February 2018—even though most school shooters are white, and the shooter in Parkland was a white student who had been expelled. The Trump administration’s position underscores the continuing draw of a characterological focus in understanding violence, including juvenile violence. The administration’s interdepartmental report in response to the Parkland shooting stressed the importance of “maintaining order in schools” and recommended relying on teachers to watch the “behavioral patterns” of their students and “correct—and where necessary, discipline—those behaviors that are unwelcome or unsafe for the school community.” Notably, though, the report did not so much as mention “zero tolerance,” nor did it call for greater use of suspensions and expulsions. There is no indication that zero-tolerance policies for school discipline will regain popularity, and most states and localities remain committed to minimizing suspensions and expulsions.36
Juvenile violence, like adult violence, is largely committed by boys. The superpredator scare was largely about boys, and the overwhelming majority of school shooters are boys. But since the early 1990s, girls have constituted a rising proportion of juveniles arrested for violent crime, and this is particularly true for assaults. In 2015 girls accounted for more than a quarter of all youth arrests for aggravated assault and close to 40 percent for simple assault. The figures in 1990 were considerably lower: close to 15 percent for aggravated assault and 25 percent for simple assault. Most of this change occurred between 1990 and 2005; since then the percentages have stayed roughly constant. Driven largely by changes in arrest rates for assault, the female share of arrests for crimes included in the FBI’s Violent Crime Index—homicide, rape, robbery, and aggravated assault—roughly doubled between 1980 and 2005, from close to 10 percent to close to 20 percent.37
The growth in the female proportion of juvenile arrests for violent crimes, and for assault in particular, does not appear to reflect any actual increase in violence by girls. It is entirely explained by two other factors: the plunge in the corresponding arrest rates for boys, and a lowering of the threshold for arresting and charging girls with assault. The lowered threshold, in turn, appears to be partly a reflection of the reduced tolerance in general, across genders, for physical bullying and other low-level violence between minors, and partly an unanticipated consequence of efforts to treat domestic violence more seriously. Mandatory arrest policies for domestic violence were adopted largely to protect women from violent partners, but there is evidence that these policies have also resulted in more arrests of girls who get into physical altercations with their parents or siblings. There is no evidence, though, that girls actually became more violent at any point over the past several decades, either inside or outside the home.38
Nonetheless, the story about girls’ arrest rates for violent offenses offers some important lessons, and those lessons reinforce themes we have encountered repeatedly in this book. The first lesson has to do with the magnetic pull of certain narratives about violence, and particularly, in the modern era, narratives that emphasize characterological explanations of violence and resonate with fears of moral decline. Just as the superpredator narrative spread and proved difficult to dislodge in the 1990s because it seemed so plausible, so did a similar narrative about violent girls in the early years of the twenty-first century. When the arrests of girls for assault began to climb, a good number of observers took this as confirmation of what they already believed: that violence committed by girls was “a burgeoning national crisis.” “The change in girls’ behavior is overwhelming,” said one expert, the author of a book entitled Sugar and Spice and No Longer Nice. Girls today, he warned, “are not what people think they are.” A decade later—long after girls’ arrest rates for violent crimes had begun to fall at the same rate as the corresponding rates for boys—a columnist, voicing what still passed for conventional wisdom in many quarters, claimed without evident fear of contradiction that “female violence is on the rise in a big way,” and that “there’s something wrong with America’s girls.”39
A second lesson has to do with the slippery definition of violent crime. We often tend to think of violent crime as though it is a category with sharp boundaries; this is part of why it can seem plausible that people who commit violent crimes are themselves in a category apart, that they have something fundamentally wrong with them. As we saw in Chapter 2, though, what counts as a violent crime is often fairly arbitrary. One reason for that is the open-ended ways in which the law defines simple and aggravated assault. Many, probably most, physical attacks aren’t charged as assault, and they therefore are not treated as crimes. Furthermore, most attacks that are charged as assaults are charged as simple assaults, not aggravated assaults, which means that they do not count as felonies, let alone “violent felonies,” and they are not included in the annual FBI Index of Violent Crimes. Police, prosecutors, and juries typically have great discretion in deciding whether something should be treated as a simple assault, an aggravated assault, or not an assault at all. That is why arrests for simple and aggravated assault could spike in the late 1980s and 1990s, even though the number of physical attacks does not appear to have increased. And that is why there could be a surge in the number of girls arrested for assault in the 1990s and early 2000s, without any rise whatsoever in the underlying incidence of violent behavior among girls.
In a series of three cases decided between 2005 and 2012, the Supreme Court ruled that the Fifth Amendment ban on “cruel and unusual punishments” places special limitations on the sentences that can be imposed for offenses committed by juveniles. The number of defendants directly impacted by these decisions is relatively small, but the decisions may be more important for their reasoning than for their holdings. Collectively, they have helped steer the legal treatment of youthful offenders away from the excesses of the get-tough era and back toward the traditional values and assumptions of the juvenile justice system, even in cases of violent crime.
The first of these cases, Roper v. Simmons, was decided in 2005. The defendant, Christopher Simmons, was convicted and sentenced to death for a brutal, senseless killing he had carried out with a friend when he was 17. We encountered Roper v. Simmons earlier. This was the case in which the prosecutor urged the jury at the sentencing phase of the trial to treat the defendant’s youth as an aggravating rather than a mitigating factor. The Supreme Court viewed that argument as obviously improper, but the Justices went considerably further: they used this case as the opportunity to declare that executing a defendant for any crime committed before the defendant reached the age of 18 was “cruel and unusual” and therefore unconstitutional. Five years later, in Graham v. Florida, the Court said that a life sentence without possibility of parole was similarly “cruel and unusual,” and therefore similarly unconstitutional, when it was imposed for a non-homicide offense committed as a juvenile. Finally, in 2012, the Justices decided in Miller v. Alabama that the reasoning of Graham applied in homicide cases as well: the Constitution barred a sentence of life without possibility of parole for any offense committed as a juvenile, just as it barred a death sentence for any offense committed as a juvenile.40
The direct reach of these three decisions was significant but limited. When the Supreme Court decided Simmons, seventy-one convicted defendants—about 2 percent of the death row population nationwide—were awaiting execution for crimes they had committed as juveniles. The Court’s decision granted them a reprieve, and it prevented any new death sentences for crimes committed by juveniles. It had no direct application, though, to noncapital cases. For their part, Graham and Miller mandated only the possibility of parole. States were required to provide defendants convicted for juvenile offenses with “some meaningful opportunity” to seek eventual release, but they didn’t need to actually let anyone out. By their terms, moreover, Graham and Miller applied only to a life sentence, not to a sentence of years, no matter how long.41
The reasoning of these three decisions, though, swept wider than their holdings. All three rested on the idea that, for purposes of criminal punishment, youth mattered. As Justice Kagan put it in her majority opinion in Miller, “children are different.” Justice Kennedy, who wrote for the Court in Simmons and then again in Graham, emphasized three particular differences: the impulsivity and poor judgment often shown by juveniles; their greater susceptibility to peer pressure and outside influences; and the fact that their personalities were not yet fully formed. In combination, Justice Kennedy wrote, these factors weighed strongly against any conclusion that a crime committed by a juvenile, no matter how heinous, was proof of an “irretrievably depraved character.” In Graham Justice Kennedy quoted with approval the conclusion of a lower court that “incorrigibility is inconsistent with youth,” and Justice Kagan relied on the same reasoning when she wrote for the Court in Miller. In Simmons the Court treated the differences between adolescents and adults as largely a matter of common sense, something that “any parent knows.” But Justice Kennedy noted in Simmons that new research by scientists and sociologists provided additional support for those impressions, and in Graham and Miller the Court placed even more weight on that research, especially on the findings by psychologists and neuroscientists.42
The Supreme Court was sharply divided in Simmons, Graham, and Miller. Simmons and Miller were 5–4 decisions. The vote in Graham was 6–3, but one of the six was Chief Justice Roberts, who concurred in the judgment without joining the majority opinion. All three decisions triggered angry dissents. Among other things, the dissenters accused the Court of imposing its own value judgments, and they stressed the artificiality of the cutoff at age 18, or any other sharp line between adolescence and adulthood. They complained, too, about the use of the word “children” to describe older adolescents, especially older adolescents who committed serious crimes.43
There was force to some of the objections. It is plainly true that young people mature in different ways and at different rates; even Justice Kennedy’s majority opinion in Simmons acknowledged that any categorical age threshold would be imprecise. And even some scholars who applauded the Court’s decisions in Miller objected to Justice Kagan’s use of the term “children” to describe older adolescents. Some of those scholars also warned about the hazards of trying to draw normative conclusions from the scientific research highlighted by the Court in Simmons, Graham, and Miller. Nonetheless, there is broad support among legal scholars and criminologists for the reasoning at the heart of these decisions: the idea that the impetuosity, suggestibility, and incomplete maturation of adolescents calls for treating them differently when they commit crimes, even serious crimes. This was the basis for creating a separate juvenile justice system in the first place, and there is broad support for these ideas, once again, among judges and policymakers. One sign of that support is that several state legislatures and lower courts have extended the reasoning of Simmons, Graham, and Miller beyond their strict holdings, giving all juveniles with long custodial terms a right to periodic reconsideration of their sentences, or invalidating the application of mandatory minimum sentences to young offenders.44
If Simmons, Graham, and Miller stand for the proposition that “kids are different,” they also stand for the position that violence isn’t different, or at least that it isn’t as significant as youth when the two categories overlap. Simmons and Miller were murder cases, and the defendant in Graham was convicted of armed robbery and then had his probation revoked because of his involvement in a home invasion robbery. The Supreme Court concluded, in essence, that when a minor commits a violent crime, even murder, it is still first and foremost a crime committed by a juvenile, and it needs to be treated that way. This was part of what the dissenters found so wrong with these decisions. For dissenters, violence could trump youth. It could even cancel out youth. The dissenters in Simmons argued, in part, that at least near the boundary between adolescence and adulthood, an especially brutal crime can justify the conclusion that the defendant had acted with “sufficient moral culpability” to justify execution. Some murders by juveniles, Justice Scalia argued, “are not just the acts of happy-go-lucky teenagers.” The dissenters in Graham and Miller made essentially the same argument about life without possibility of parole. Justice Alito, dissenting in Miller, mocked the suggestion that “a 17½-year-old who sets off a bomb in a crowded mall or guns down a dozen students and teachers is a ‘child’ and must be given a chance to persuade a judge to permit his release into society.” Seventeen-year-olds commit lots of murders, he pointed out, and “some of these crimes are incredibly brutal.” It just stood to reason, then, that “many of these murderers are at least as mature as the average 18-year-old.” Justice Alito was appealing to the same intuition that had led to sending so many juvenile offenders to regular criminal courts beginning in the 1980s and 1990s: serious violence is inconsistent with true juvenility, and therefore juveniles who commit serious violence should be treated the same as adults.45
There is some basis for the view that violent juvenile offending differs from other forms of juvenile offending, at least if we reserve the label “violent” for homicide, aggravated assault, and rape. Age-specific rates of arrest for robbery, burglary, theft, and arson all peak somewhere between ages 13 and 18, and they decline rapidly after that. Arrest rates for homicide, aggravated assault, and rape follow a different trajectory: they peak in very late adolescence or early adulthood—the late teens or early twenties—and then decline more slowly. These patterns are consistent with the view that juveniles who commit property offenses are more likely to age out of criminality than those who commit violent crimes. But other evidence suggests that it is artificial to distinguish between violent and nonviolent juvenile offenders, because although some juvenile offenders “specialize” in this way, most do not. Overall, the major connections between juvenile and adult offending are twofold: the vast majority of adult offenders begin offending as juveniles, but the majority of juvenile offenders do not continue to offend as adults. Furthermore, aggravated assaults and homicides by adolescents show the same distinctive pattern as other adolescent crimes: they are committed predominantly by groups of adolescents, not by offenders acting alone. This was true, notably, of the crimes committed by the defendants in Simmons, Graham, and Miller. That suggests that at least some of the observations the Supreme Court made in Simmons, Graham, and Miller—the observations about the suggestibility of adolescents and their susceptibility to peer pressure—are no less applicable to violent juvenile offenders, or even to juveniles who commit particularly brutal crimes. And there is no reason to think that the other distinctive characteristics of adolescents stressed by the Court in these cases—their impulsivity and poor judgment, and their unfinished maturation—are any less applicable to violent juvenile offenders.46
The law’s ambivalence about violence by juveniles—whether it is ordinary or pathological, whether it is more forgivable than violence by adults or less so—is matched by an ambivalence about violence against juveniles. On the one hand, protecting juveniles from violence—even from being exposed to violence—is sometimes treated as being of paramount importance, not just because children are “innocents,” but because there is good reason to think that children who live with violence are more likely to grow into violent adults. On the other hand, violence against juveniles, at least if it stays within bounds, is often treated as fully excusable, if not actually laudatory. “He that spareth his rod hateth his son,” as the Bible has it.47
Physical punishment of children is less popular than it used to be, partly because there is mounting evidence that it does little or nothing to improve children’s behavior and, over the long run, is apt to make them more aggressive, more antisocial, and more violent. Whipping children and beating them with sticks has long been out of fashion, and elite views have turned against milder forms of physical discipline as well. Expert opinion is solidly against spanking; the American Academy of Pediatrics, for example, recommends against any physical punishment of children. More than sixty countries around the world have outlawed the practice.48
No American jurisdiction has followed suit, however, and none appears likely to do so. The Model Penal Code excuses parental assaults on children, no matter how young, as long as there is no substantial risk of “death, serious bodily injury, disfigurement, extreme pain or mental distress or gross degradation.” Many state laws go further and allow parents to beat their children as long as there is no danger of grave physical injury or death. (Sexual assaults against children are, of course, a different matter. They are treated as serious crimes. But that is because of their sexual nature, not because of the violence involved.)49
Around 70 percent of American adults agree that “it is sometimes necessary to discipline a child with a good, hard spanking.” That figure is down from over 80 percent in the mid-1980s, but it has held steady for the past decade. Slightly over half of American parents claim that they never spank their own children, and 28 percent say they do so only “rarely.” But fewer than one in five American adults favor outlawing physical punishment of children. And views about spanking vary according to geographical region, educational background, and race. African American parents, in particular, are more likely to report that they spank their children, and more likely to defend physical punishment, even serious beatings, as a form of “tough love,” necessary in particular for African American children for whom straying from the straight and narrow, or disrespecting authority figures, can mean winding up dead. Now himself a parent, Ta-Nehisi Coates remembers and understands his father repeating “the old mantra—‘Either I can beat him or the police.’ ” (This is a reminder of the ways different forms of violence that seem unrelated—like police violence and corporal punishment—can in fact be linked; one form of violence can prompt and help to justify another.) Addressing the NAACP in 2009, and receiving a warm response, President Obama spoke fondly of the days when, if parents “saw some kid fooling around and—it wasn’t your child, but they’ll whup you anyway. Or at least they’ll tell your parents [and] the parents will.… That’s the meaning of community.”50
There is considerably less support in the United States today for corporal punishment inflicted by teachers and school officials, a form of discipline that is rarer than it was a generation ago. More than seven out of ten Americans oppose allowing teachers and school administrators to inflict corporal punishment on students, and most states have banned the practice in public schools. In private schools, though, the practice remains legal everywhere but in Iowa and New Jersey. Moreover, corporal punishment in public schools is still lawful in nineteen states, and more than 106,000 students were paddled, swatted, or otherwise physically punished in American public schools during the 2013–2014 academic year, the last year for which statistics are available. That was a dramatic decline from a decade earlier, though, when the annual figure was more than twice that high.51
Whether the figure will continue to decline so sharply is unclear. The trend toward state bans on corporal punishment in public schools has slowed in the past two decades; no state has outlawed the practice since New Mexico did so in 2011. At this point corporal punishment in public schools is very much a regional practice. Over 96 percent of all children in public schools who were spanked or paddled during the 2013–2014 school year lived in one of nine states: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, Oklahoma, Tennessee, or Texas. Half of all children in Alabama, Arkansas, and Mississippi attend schools that use corporal punishment. Mississippi is the most enthusiastic user of the practice. In 2013–2014, close to 25,000 schoolchildren received corporal punishment in Mississippi: a quarter of the national total, and about one in twenty students statewide.52
The Supreme Court last considered the legality of corporal punishment in public schools in 1977, when it decided Ingraham v. Wright. James Wright was an eighth-grader in Miami, Florida, who was paddled so hard by his middle school principal that he missed half a month of school and had to make two trips to the hospital. Many observers expected the Supreme Court to use his case to ban corporal punishment in public schools. Instead the Supreme Court upheld the practice, ruling that the Eighth Amendment ban on “cruel and unusual punishments” applied only to criminal punishments, and that the due process interests of students were fully protected by the rule, long imposed by common law, that corporal punishment could not go beyond what was “reasonably necessary.” The Court was influenced, in part, by the absence of any emerging consensus against corporal punishment in schools. At the time only two states had banned the practice, and polls indicated that most adults supported it. The Justices could “discern no trend toward its elimination.”53
That trend is more apparent today, but in the intervening decades the Supreme Court has become a much more conservative institution, and it has lost whatever appetite it had for regulating school disciplinary practices. The Justices have declined invitations to reconsider Ingraham v. Wright. Lower courts have found that corporal punishment in public schools can violate due process if it goes too far, but it has to go very far. The leading decision says that physical punishment of schoolchildren violates the Constitution only when it causes such severe injuries, is so grossly disproportionate, or is so malicious or sadistic that it amounts to “a brutal and inhumane abuse of official power literally shocking to the conscience.”54
Before leaving the topic of corporal punishment in schools, it is worth noting that the practice disproportionately impacts particular subpopulations of students in troubling ways. That is notably true of minority students. James Wright was African American, as were all the other students at his school, which was notorious for its harsh physical discipline. Today African American children constitute roughly 15 percent of the students in public schools but 38 percent of the students who are paddled, spanked, or struck. Some of this is because Black children are more likely to live in states where the practice is permitted, but even within those states, and even adjusting for frequency and severity of misbehavior, African American children are disproportionately singled out for physical punishment. Corporal punishment is also visited disproportionately on children with disabilities, particularly in some places. Analyzing data collected by the US Department of Education, researchers found that in one-third to one-half of all school districts in Alabama, Arkansas, Georgia, Louisiana, and Mississippi, children with disabilities were more than 50 percent more likely to receive corporal punishment. Statistics of this kind convinced lawmakers in Tennessee and Louisiana to ban corporal punishment for children with disabilities, but not for other children. Boys are spanked, paddled, or struck at school four times as often as girls, but there may be particular reasons to be concerned about the manner in which corporal punishment against girls is carried out, and the way it may appeal to conscious or unconscious sexual fantasies.55
Public schools are virtually the only governmental agencies still licensed to mete out corporal punishment. Correctional facilities do not have this authority. Neither do police departments. Strictly speaking, however, this means only that correctional agencies and police departments cannot use force on a person for the sake of inflicting punishment; it is far from a ban on violence. We examined the legal rules about police uses of force in Chapter 3, and Chapter 6 will explore how the law thinks about violence in prisons, jails, and other penal and detention facilities. What is worth noting here, though, is that almost none of those rules are specifically targeted at the use of violence against juveniles. There are no special restrictions, for example, on when the police may use force against children or adolescents, or what kinds of force they can employ.
The law sometimes pays a good deal of attention to the way that violence can beget more violence, particularly among children and adolescents. In family and dependency court, parents can lose custody of their children—to their former partners or to the state—not just for treating their children violently, but for allowing their children to witness violence, including violence directed at the partner whose custody is at issue. Advocates for abused women have complained that the desire to protect children from exposure to violence, however well meaning, can be carried too far, with the result that abused women can be unfairly punished with the loss of their children, something that often winds up harming the children as well. Some states now specifically bar taking a child away from a domestic violence victim solely on the ground that the child was exposed to the parent’s victimization. The consensus view of child welfare professionals is now that the developmental consequences of being exposed to violence vary widely; some children are impacted much less than others. But guarding children against exposure to violence, not just violence itself, remains a central goal of family law and child welfare, in part because of the widespread recognition that being around violence, let alone experiencing violence, can damage young people in a range of ways, including by making them more prone to violence themselves.56
That connection often seems lost, though, when the criminal legal system addresses the intersections of violence and youth. There is a set of rules for responding to violence committed by juveniles, and there is a set of rules that govern violence against juveniles, and by and large these two sets of rules are mutually independent, shaped by discourses that do not overlap. This itself may reflect certain assumptions about violence, about where it comes from and how it should be understood, and certain assumptions, too, about youth. When it banned capital punishment for juvenile offenses, and later did the same for life without possibility of parole, the Supreme Court stressed the impulsivity of adolescents, their susceptibility to peer pressure and outside influences, and their unfinished maturation. The Court said that the law has to treat these differences as though they matter. If the Court is right about that, it may mean that the law will need to stop treating violence against juveniles and violence by juveniles as two separate problems. It may mean that the intergenerational cycle of violence—the ways in which children subjected to violent abuse can themselves become violently abusive—should play a larger role in assessing appropriate sanctions for youthful offenders, and in formulating rules about corporal punishment of children in schools and at home.