Scholars, practitioners, and policy makers have historically been allergic to theorizing about why the law should punish adolescents who commit serious and violent crimes less severely than adult offenders. Rather than developing a jurisprudence of youth violence, they have instead debated whether juvenile court or adult criminal court should hear these cases and whether judges or prosecutors should ultimately decide where particular cases should be tried. As a result, there has been an outpouring of scholarship and legislation about the transfer of adolescents out of juvenile court, but few critical studies of the legal standards and substantive principles that criminal courts should use to sanction younger offenders (Zimring Reference Zimring1998; Fagan and Zimring Reference Fagan and Zimring2000).Footnote ¹ For this reason, Christopher Berk’s meticulously argued article is a welcome addition to the procedurally thick but substantively thin juvenile justice literature.

Berk provides a timely critique of the adolescent development framework for juvenile justice, which the US Supreme Court has adopted in a string of decisions that began with Roper v. Simmons in 2005. That 5–4 decision, written by Justice Anthony Kennedy, abolished the juvenile death penalty. In doing so, Kennedy cited studies of development research, including the scholarship of Laurence Steinberg and Elizabeth Scott, to support the Court’s conclusion that adolescents could not be considered among the “worst of the worst” category and thus were ineligible for the death penalty (Roper v. Simmons 2005, 569). More recently, in Miller v. Alabama (2012), another 5–4 decision, Justice Elena Kagan constitutionalized the idea that “children are different” from adults, at least for sentencing purposes. Drawing on Roper and its progeny, she highlighted three “significant gaps” between children and adults: (1) children’s lack of maturity leads to “recklessness, impulsivity, and heedless risk taking”; (2) children are more vulnerable to family and peer pressure, have limited control over their environments, and “lack the ability to extract themselves from horrific, crime-producing settings”; and (3) their character and traits are not as well formed as adults’ and may change over time (Miller v. Alabama 2012, 471).

The Court used Miller to abolish mandatory life without the possibility of parole sentences for juvenile murderers, and then in Montgomery v. Louisiana (2016), a 6–3 decision written by Justice Kennedy, the Court made Miller retroactive.Footnote ² Chief Justice John Roberts dissented in Miller partly because he feared that the “children are different” principle had no limits. As he stated, “There is no clear reason that principle would not bar all mandatory sentences for juveniles, or any juvenile sentence as harsh as what a similarly situated adult would receive. Unless confined, the only stopping point for the Court’s analysis would be never permitting juvenile offenders to be tried as adults” (Miller v. Alabama 2012, 501). Yet the Chief Justice did vote with the majority in Graham v. Florida (2010), which abolished life without the possibility of parole sentences for juveniles in noncapital cases, and in Montgomery. Despite the recent personnel changes on the Roberts Court, there are five justices who accept the adolescent development framework.

Berk describes this “dominant” framework as “philosophically confused” because it treats age as a “gross proxy” instead of a “boundary of a democratic pre-commitment to care for young people” (Berk Reference Berk2019). The dominant framework, he argues, conflates empirical findings about adolescent development with normative claims about childhood. He contends that its proponents overstate the scientific evidence and underappreciate the social constructiveness of childhood. Moreover, widespread acceptance of the proposition that “children are different” from adults prevents us from asking hard questions about why a “seventeen-year-old that kills be treated more leniently than his eighteen-year-old counterpart” (Berk Reference Berk2019). “This line of reasoning,” he notes, “puts pressure on the Court’s jurisprudence in other areas of law—ranging from contracts to access to abortion and general medical care” (Berk Reference Berk2019).

Berk provides a crisp reading of the contemporary constitutional moment (c. 2005 to the present) that probes two weaknesses (or dependencies) in the developmental model. He first draws on the scholarship of Terry Maroney and Gideon Yaffe to point out that the model is empirically dependent on using scientific knowledge to make age into a legal category (Maroney Reference Maroney2009; Yaffe Reference Yaffe2018). That scientific knowledge, however, may change, which could corrode the foundations for the legal category. Berk then argues that the developmental model also relies on a narrative about “what it means to ‘grow up’ in a given society” (Berk Reference Berk2019). As he explains, “‘Child’ is not simply a descriptive category, it’s also a normative one. Advocates mistake socially contingent, but widespread features of our political order for the way things must be. Public choices concerning the political status of children are understood as natural, inevitable, and necessary” (Berk Reference Berk2019, italics in original). Much as science can change, so too can narratives about childhood and adolescence.

There are, of course, prominent critics of the adolescent development framework for juvenile justice. Neither the National Juvenile Justice Prosecutor Center (NJJPC) nor the Trump administration subscribes to its tenets. Instead, they promote a restorative or balanced model of justice.Footnote ³ These critics regularly question the adolescent development framework and resist its full implementation.

Berk proposes using democratic theory as the basis for developing a principled and proportional jurisprudence that would limit the severity of the legal response to children who kill. Because children are denied “a full schedule of democratic rights,” he argues, “[w]e ought to ask what adolescents are owed in virtue of the political consensus that they have a different and special status.” This new starting point would put “scholars, policy makers, and the courts in a better position to make more fine-grained judgments wisely” and might lead to fairer youth policies across the board because the public might take its paternalistic responsibilities toward youth more seriously (Berk Reference Berk2019).

Juvenile justice reformers during the 1950s and 1960s also relied on their understanding of democratic theory and contemporary conceptions of fairness to argue for taking paternalism seriously (Langer and Tanenhaus Reference Langer and Tanenhaus2018). For example, District of Columbia Juvenile Court Judge Orman Ketcham used the “The Mutual Compact Theory of Parens Patriae” to explain that a child exchanged his constitutional rights in return for the state’s promise to “provide him with the essentials of parental training, care, and custody” (Ketchum Reference Ketchum1961, 100). “Unless the state is required to make good its promises,” Ketchum argued, “American juveniles will have exchanged the precious heritage of individual freedom under law for the tyranny of state intervention whenever the state considers that its interests are affected” (Ketchum Reference Ketchum1961, 109). Scholarship demonstrating that states were not acting like good parents set the stage for Justice Abe Fortas’s damning statement in Kent v. United States (1966) about juvenile courts, “that the child receives the worst of both worlds: that he gets neither the protections accorded to adults nor the solicitous care and regenerative treatment postulated for children” (556).

A philosophically grounded jurisprudence of youth violence should address the first constitutional revolution in American juvenile justice. These cases, especially Kent but also In re Gault (1967) and In re Winship (1970), can contribute to an analysis of proportionality and punishment of juvenile offenders in a democracy.

The moral panic about youth violence in the 1990s, which Berk alludes to with his reference to John DiIulio and “super-predators” (Berk Reference Berk2019), is also historically and theoretically significant. Berk’s readers, for example, need to know how his argument about narrative dependency fits with his opening assertion/assumption that “few beliefs are as strongly held” as Justice Kagan’s pronouncement in Miller that “children are different” (Berk Reference Berk2019). The idea that children are different from adults certainly has a long history, but it was not a widely articulated belief during the moral panic. According to the 1996 report on State Responses to Serious and Violent Juvenile Crime, “[i]t is clear from conversations with juvenile justice planners, prosecutors, judges, legislators, and corrections administrators across the country that public fear—more precisely the fear of being killed by a young person—was the driving force behind recent changes to stem the tide of violent crime by juveniles.” The report further noted, “[f]requently, legislatures responded to that fear with proposals to get even, punish, or hold juveniles accountable. Quite often the responses were couched in rhetoric such as ‘If they can kill like an adult, they can be treated just like an adult’ or ‘If you do the crime, you do the time’” (Torbet et al. Reference Torbet1996, 52; quoted in Bush and Tanenhaus Reference Bush and Tanenhaus2018, 8)

Despite promising to describe how the Roberts Court “has come to embrace the developmental approach, along with its general appeal,” Berk does not explain why the Supreme Court has taken the developmental turn (Berk Reference Berk2019). That history has not yet been written, but its outlines are at least visible. They include analyzing the relationship between the moral panic over youth violence and the construction of the adolescent development framework as a response to it. Stanford v. Kentucky (1989) also cast a long shadow over this history. That 5–4 decision held that a state could execute sixteen- and seventeen-year-olds who were convicted of capital offenses because there was not a national consensus against imposing the death penalty on minors over the age of fifteen. The abolitionists, who decided to mount a new campaign against the juvenile death penalty during the 1990s, were doing so at the height of the moral panic.

That cultural context complicated making an Eighth Amendment argument about the unconstitutionality of the juvenile death penalty, which required demonstrating to the majority of the justices that evolving standards of decency in a maturing society should end the practice. Advocates for juvenile justice reform, including those representing juveniles facing the death penalty or life sentences without the possibility of parole, were searching for ways to rewrite the public narrative about children (and juvenile courts, too) as part of their reform agenda (Tanenhaus and Drizin Reference Tanenhaus and Drizin2001–2002).

About the same time that juvenile death penalty abolitionists were organizing, in 1996 the MacArthur Foundation launched its Adolescent Development and Juvenile Justice Research Network, which Laurence Steinberg directed.Footnote ⁴ The success of the Network, including the Supreme Court citing Steinberg and Scott’s research, obscures the uncertainty of the social science knowledge at the height of the moral panic. In 1996, Law and Human Behavior published a special issue about “Children’s Capacities in Legal Contexts” that even included a relevant article by Thomas Grisso about the potential of a developmental perspective for responding to youth homicide (Grisso Reference Grisso1996). An overarching theme of the issue was that empirical research needed to be done to test what were largely theoretical models. The editors of the issue ended their introduction by quoting Donald Bersoff’s and David Glass’s assertion, “[s]ocial scientists play on a legal ball field. Their work is evaluated according to the rules the legal system lays down.… But it is hoped that social scientists will continue to develop situation-specific, ecologically valid, legally relevant, objective data that, despite resistance, will help the Supreme Court, as well as others who make social policy, to arrive at empirically justified decisions that match the real world” (Woolard, Reppucci, and Redding Reference Woolard, Reppucci and Redding1996, 227).

Ten years later, the Supreme Court fulfilled this prophecy, beginning with the Roper v. Simmons decision. I do not want to understate the extraordinary vision of the juvenile death penalty abolitionists who litigated Roper successfully. This included drafting important briefs as well as securing amicus briefs from the leading professional organizations such as the American Medical Association and the American Psychological Association. I do, however, want to stress that the Supreme Court neither had to recognize the developmental evidence nor had to use it to make its decision. During oral arguments Chief Justice William Rehnquist repeatedly questioned Seth P. Waxman, the former solicitor general who was representing Christopher Simmons, about why the Court should take judicial notice of the social science literature cited in the amicus briefs by professional organizations. Rehnquist stressed that “if we are to take that as a fact, it ought to have been tested somewhere rather than presented to us in a brief” (Transcript of Oral Argument 2004, 36). He added that if the Court were to rely on such evidence, then it should have been tested “the way most facts are” (Transcript of Oral Argument 2004, 36). Justice Kennedy, who ultimately provided the deciding vote in Roper, also expressed concern about this evidence. And Justice Stephen Breyer even argued that the scientific evidence could be interpreted to merely corroborate “what every parent already knows” about adolescents (Transcript of Oral Argument 2004, 40).

Justice Kennedy’s opinion did specifically cite social scientific literature from the amicus briefs to pronounce that the “diminished culpability of juveniles is recognized” (Roper v. Simmons 2005, 571). Some of this literature was the product of the 1990s research on adolescent development and juvenile justice. But Kennedy also cited earlier work such as Erik Erikson’s Identity, Youth, and Crisis Reference Erikson(1968) and included arguments in Roper about foreign law (575–78) that proved controversial and disappeared from the subsequent juvenile justice decisions (Agyepong Reference Agyepong2010). Kennedy also gestured to the political theory that Berk has called for as an alternative (or perhaps complement) to the developmental framework and included appendices with the minimum ages for voting and jury service across the nation (581–85).

The significance of Roper is twofold. First, the abolition of the juvenile death penalty itself was a remarkable human rights victory. Second, it opened up the possibility of using more developmental research, which was soon complemented by the emerging field of brain science, in subsequent litigation (Maroney Reference Maroney, Zimring and Tanenhaus2014). The internal court history of how Kagan moved her colleagues from Roper’s “death is different” rationale to Miller’s “children are different” principle remains to be written, but we do know that Bryan Stevenson, who litigated Miller, offered the justices two rules for striking down life without the possibility of parole sentences. The less far-reaching rationale would have distinguished cases of younger adolescents (fourteen years old and younger) from older ones. The Court, however, opted for the more far-reaching and categorical rationale. We still do not know exactly why.

In conclusion, Berk has done the field a great service by providing such a searching critique of the adolescent development framework and previewing his ideas about how to incorporate more democratic theory into the jurisprudence of youth violence. I hope his larger project will bring more voices from the past into this urgent conversation about the future of juvenile justice.