A pertinent quote (emphasis added):

The [present juvenile justice and mental health] movement was energized by empirical evidence offered by many reliable studies that provided data about the prevalence of mental disorders among youths in juvenile justice settings. Those studies announced that a large proportion of these youths—as many as two-thirds—met DSM criteria for one or more mental disorders (i.e., mood, anxiety, substance use, conduct, or developmental disorders). This was an alarming message for many juvenile justice administrators. Federal government and juvenile advocates called for the juvenile justice system to respond and many presumed that this meant that they had to find a way to provide treatment for most of the youths in their care.

This presumption, of course, is simplistic. The fact that two-thirds of youths in detention centers meet criteria for a psychiatric disorder does not mean that they are seriously in need of psychiatric treatment. Youths with a particular disorder vary in the severity of their symptoms. Some function relatively well in everyday life and others very poorly. Youths' psychological conditions are more labile than those of adults. Compared with adults, there are greater risks that youths with symptoms of one disorder at one point may, within another year, meet criteria for a different disorder or no disorder at all. Moreover, prevalence rates for mental disorder depend on what one defines as mental disorder. Shall we leave in or take out conduct disorder? How about substance use disorders? Thus, most experts recognize that it is not necessary and is probably unwise for the juvenile justice system to translate the published prevalence rates into a policy that seeks treatment for two-thirds of the youths in its custody.

This, in my opinion, makes a lot of sense. Often times people will quote statistics such as the prevalence of mental illnesses in jails and prisons (discussed here) without giving much thought to the inevitable complexities behind the nexus between crime, behavior, and mental states. As Stephen D. Levitt and Stephen J. Dubner explore in their book Freakonomics, the epidemic of the juvenile "superpredator" never materialized as predicted by other criminal justice experts. As author Grisso rightfully concludes:

The rather sudden appearance and rapid growth of the juvenile justice and mental health movement is both encouraging and troubling.

Indeed, although as someone who has worked as counsel in the juvenile justice system, I'd take some issue with another part of the conclusion:

Juvenile justice policy in the United States has had three eras. The first was the birth of the juvenile justice system at the beginning of the 20th century. The second was the due process reform of the system represented by Kent v. U.S. and In re Gault in the 1960s. Both of those eras arose after decades of growing consensus regarding their necessity.

Perhaps there was a growing consensus regarding the necessity of making juvenile delinquency matters more like adult criminal court in terms of constitutional criminal justice, but that change effectively removed much of the therapeutic nature of the juvenile justice system in favor of one heavy with the adversarial model. Such change may be congruent with the Constitution, but it should be recognized that it is in stark contrast to the original model and purpose of the juvenile court system and frequently impedes the therapeutic aims of the juvenile court.