Research Report

Reducing Harms to Boys and Young Men of Color from Criminal Justice System Involvement

Boys and young men of color are disproportionately represented as criminal victims and offenders and are overrepresented in all aspects of the juvenile justice and criminal justice systems. This overrepresentation, which can devastate the lives of young men, their families, and their communities, is most acute for African Americans. Recent events have underscored the continuing issue of race and the uniquely fraught position of African Americans in the national consciousness concerning crime and justice in the United States. Late in 2014, public demonstrations were held nationwide after unarmed black men died at the hands of police in Ferguson, Missouri, and New York City. These protests escalated further when grand juries declined to indict police officers on criminal charges.

The relevant research evidence on this topic is most extensive concerning African American youth, although there is also a growing body of research concerning Latino youth. We sometimes discuss the combined group of black and Latino youth here as “youth of color.” Although youth of both groups face some common issues, there are also important differences in both the nature of the problem and possi-ble solutions. There is also an acute problem for Native Americans, especially on tribal lands, although this has different dynamics and is not well represented in either the data or the research literature.1

Overview and Summary

Incarceration in the United States has exploded since 1980; the resulting situation is often referred to as the era of mass incarceration (National Academy of Sciences 2014). This mass incarceration has been particularly concentrated among poor and poorly educated black men. For example, among high school dropouts, almost a third of black men ages 20–40 were incarcerated on a given day in 2000, compared with 6 percent of Latinos or whites; by their early 30s, the lifetime risk of imprisonment of male high school dropouts was an astonishing 59 percent for blacks, compared with 11.2 percent for whites (Western 2006, 19 and 27).

At the justice system level, aspects of the overrepresentation of youth of color include law enforcement practices that focus on and respond differently to young men of color and that increase their juvenile justice and criminal justice involvement. In addition, police seem to apply greater use of force to people of color, particularly boys and young men, as exemplified by episodic incidents of unarmed black men dying at the hands of police.

Increased involvement with the justice system tends to interfere with school completion and employment in a negative cascade. Harsh sentencing policies then exacerbate any disproportionalities at entry in the justice system, and perverse financial incentives may incentivize longer and harsher sentences than necessary.

For juveniles in particular, the justice system is used to address problems best handled elsewhere, such as school discipline issues; in addition, this pathway into the justice system is used disproportion-ately for youth of color. All too often the justice system is seen as a benevolent route to services that are in the child’s best interests, especially when services are not otherwise easily available or funded. But this involuntary route to services delivery comes with high costs, including detrimental labeling effects, along with the risk of further sanctioning and justice system involvement for noncompliance with treatment and service mandates.

At the community level, boys and young men of color are more likely to live in neighborhoods of concentrated disadvantage in which the ability to use informal social controls over crime and disorder are weakened and where perceptions of police illegitimacy interfere with the ability to harness formal resources to control crime and disorder. Children in high-crime communities are exposed to high levels of violence as both victims and witnesses. With young men of color stopped by police, arrested, and incarcerated at disproportionate levels, their children are more often exposed to the trauma and humiliation of parental arrest and incarceration. Similarly, Latino children of immigrants are disproportionately affected by parental arrest and incarceration by immigration authorities. Schools, which have enormous potential as prevention and intervention settings, also sometimes take harsh approaches to discipline, including suspensions. Such approaches are also used more often with youth of color. The increased presence of law enforcement in schools also seems to increase the likelihood of referral to court for infractions that otherwise are managed within the school.

Although it is commonly believed that poverty, community disorganization, and troubled families are root causes for criminal justice system involvement, in the era of mass incarceration the causality often also runs the other way: incarceration and criminal histories interrupt family and marriage and lock people into poverty and low wages while removing them from many public benefits. Once a community reaches some critical mass of criminal justice involvement, the effects of poverty, disadvantage, and crime are mutually reinforcing. This has happened most acutely in neighborhoods of concentrated disadvantage, which are most often predominantly African American.

At the family level, some of the youth who appear in the juvenile justice system have considerable service needs. In the absence of sufficient community resources for treatment and services, the juvenile justice and criminal justice systems then become vehicles for service provision. However, services through the justice systems come with additional financial costs, with detrimental labeling effects of justice involvement, and with the risk of additional penalties for “treatment noncompliance.” Finally, youth facing multiple problems tend to be involved in multiple systems, which often do not coordinate their responses and sometimes work at cross purposes.

Addressing these problems will require a multiplicity of efforts. With the endemic overrepresenta-tion of boys and young men of color in the juvenile justice and criminal justice systems, almost all as-pects of juvenile justice or criminal justice reform seemingly could benefit boys and young men of color.

A natural step is to require explicit consideration of how changes to policy or practices affect youth of color. Racial impact statements, which would make any anticipated disparate impact on youth of color a presumptive ground for reconsidering and revising a policy, are a mechanism for anticipating these effects. Monitoring the race-specific effects of reform efforts is the obvious complement.

One general conclusion: the juvenile justice system should be used sparingly as a vehicle for improving youth outcomes, because any benefits come with possible harms. Much of the work here will involve removing counterproductive policies that lead to unnecessary and detrimental police stops, arrests, detention, and incarceration. For example, promising approaches have been developed for improving school climate and managing misbehavior by using positive methods that minimize justice system involvement.

Within the justice system, this work will involve changing law enforcement and arrest policies and practices, assigning more youth to diversion programs and fewer to detention and out-of-home placement, and reforming sentencing policy. To buttress these reforms, validated, objective risk and needs tools can inform decisions about youth disposition and placement; these tools should attend to possible bias across race and ethnicity. In addition, probation policies could be improved considerably, especially in the juvenile justice system, where philosophy, probation conditions, and policies concerning responses to “technical” violations of probation that do not involve new offenses all vary considerably. Clear probation policies and practices that are developmentally informed and that aim to promote positive youth development have considerable potential.

To reduce the reliance on the justice system for obtaining needed services, it is crucial to develop a robust infrastructure and financing for providing effective primary and secondary prevention for at-risk youth completely outside the juvenile justice system. Though discussion of most of the prevention literature is beyond the scope of this paper, we note that one cluster of effective programs for at-risk youth concerns family interventions, ranging from early-childhood visiting to family therapy to multi-therapeutic foster care. However, for evidence-based programs to be effective, they must be implemented with fidelity, including at the recommended intensity. Here, too, underfunding successful programs tends to undermine their effectiveness.

Successful reentry also warrants continued investment and attention to mitigate the continuing and lasting effects of mass incarceration. There is a growing body of evidence on best reentry practices, along with evidence suggesting that when reentry services are provided far below the level of need (which is typical), they may not prove effective.

As indicated, many of these issues also have financing aspects that warrant attention. For example, because funding for community-based treatment is often insufficient, treatment and services often come through the justice system. Both inside and outside the justice system, potentially effective services (including reentry services) are sometimes not funded at the intensity required to achieve effects. In addition, some local jurisdictions have perverse incentives to impose longer sentences at state expense.

This paper is organized into two large sections, the first discussing aspects of the problem and the second discussing aspects of solutions to the problem. After some initial background, both the discussion of the problem and of solutions are organized around the justice system, community aspects, and family and individual aspects.

The Problem

Boys and young men of color, particularly black youth, are disproportionately represented as criminal victims and offenders and are overrepresented in all aspects of the juvenile justice and criminal justice systems (Piquero 2008). By age 23, 49 percent of black males, 44 percent of Hispanic/Latino males, and 38 percent of white males have been arrested (Brame et al. 2014). Juvenile court referral rates and juvenile detention placements differ significantly between youth of color and their white counterparts. Black youth, in particular, are referred to juvenile court at twice the rate of white youth, and minority youth are twice as likely to be sent to secure detention as white youth (Bernard 2006).

Higher rates of police contact, arrest, juvenile detention, and other factors have culminated in starkly disparate rates of incarceration in the criminal justice system by race. Among men ages 18 and older, the national incarceration rate is 1 in 106 for white men, 1 in 36 for Hispanic/Latino men, and 1 in 15 for black men. Black men ages 20–34 have the highest incarceration rate by race or ethnicity and gender: 1 in 9 (Pew Charitable Trusts 2008). Incarceration is particularly concentrated among poor and poorly educated black men. Among male high school dropouts ages 20–40, the incarceration rate is 32.4 percent for black men, compared with 6.0 percent for Hispanic men and 6.7 percent for white men; by their early 30s, the lifetime risk of imprisonment for black male high school dropouts is 59 percent, compared with 11 percent for whites (Western 2006, 19 and 27).

African Americans have a unique, well-documented role in the history of US crime and justice, including unequal protection under the law and unequal enforcement of the law, the exclusion of blacks from juries, the use of extrajudicial capital punishment—lynching—during the Reconstruction and Jim Crow eras, and the overrepresentation on death row of black offenders with white victims (R. Kennedy 1997; Miller 1996). Incidents of police use of lethal force disproportionately involve black men and boys,2 as in recent high-profile incidents in Ferguson, Missouri, and New York City, which have spurred large and ongoing community protests. Failures to indict the involved police officers are taken by many as indicating that the legal system routinely devalues the lives of African Americans.

Race has also played a particularly important role in the evolution of juvenile justice, both in the construction of a separate justice system for juveniles and in its evolution toward a more punitive system (Feld 1999; Ward 2012). The current mass incarceration of black men that resulted from the drug war and mandatory sentencing (Mauer 1999) has been argued to be either “malign neglect” (Tonry 1995) or quite intentional (Alexander 2010).

Other Minority Groups

Other minority groups are also overrepresented in the criminal and juvenile justice systems relative to whites, but the dynamics of their problem and the promising avenues to amelioration may be quite different from those for African Americans. Two other groups that merit some discussion are Hispanics and Native Americans. Latino boys and young men are overrepresented compared to whites, albeit at not nearly the level of black youth. Some of the reasons for overinvolvement are similar (e.g., poverty, gang involvement, and stereotypes) and are therefore somewhat amenable to similar solutions. The issue of the immigrant generations, however, warrants particular attention. Because linguistic and cultural differences are largest for first- and second-generation immigrants, one might expect that first-generation immigrants would be especially overrepresented in the justice system. Surprisingly, however, first-generation Latinos are underrepresented in crime and the justice system (Wadsworth 2010),3 and they have better health and mortality outcomes than similarly situated whites. But this advantage disappears generationally; by the third generation, Hispanic youth involvement in crime and criminal justice is worse than whites. This has been called the “acculturation paradox,” in which acculturation of Latinos to the United States leads to negative outcomes.

Another group with exceptional overrepresentation is Native Americans, especially in tribal lands. Some reasons are common to other overrepresented groups (e.g., poverty, unemployment, and addic-tion). However, criminal justice functions so differently in tribal jurisdictions, with less local control over both law and its enforcement and a much larger federal role, that the nature of both the problem and the possible justice-specific solutions are considerably distinct from those involving other groups. Moreover, data on crime and criminal justice responses in tribal jurisdictions are considerably deficient relative to other parts of the United States, making it harder to diagnose the problem. As a result, the recent report of the Indian Law and Order Commission (2013), A Roadmap for Making Native America Safer, focuses most of its recommendations on making tribal criminal and juvenile justice much more structurally similar to the rest of the United States, in terms of maintaining much more local control, financing, and other aspects of law and justice administration, as well as more consistent data collection.

Implicit Bias

“Implicit” stereotypic associations between violence and blacks in particular are long-standing in American society, and they have only been exaggerated by mass incarceration. Implicit associations make judgments consistent with those associations quicker and easier to make, and they make counter-stereotypic judgments slower and more difficult, as demonstrated by the Implicit Association Test (IAT) of Project Implicit ( Among more than 2.5 million participants, 72 percent showed the reverse; this association is also strong, albeit somewhat weaker, among black participants (Nosek et al. 2007, 20). Stereotypic associations have also been shown in experiments with small samples of police officers, juvenile probation officers, and judges (see Staats 2013 for a review).

Whether implicit biases can be countered effectively remains controversial. That implicit associations make counter-stereotypic snap judgments relatively difficult is not subject in any simple way to motivation. Various experimental “debiasing” demonstrations have reduced the extent of bias, such as by inducing experiences with people who are counter-stereotype exemplars. But these reductions are typically short lived and small. Essentially, the larger societal context typically overwhelms these efforts in relatively short order (Joy-Gaba and Nosek 2011).

One reason implicit bias is difficult to remove or change is that biases from implicit associations are part of the spontaneous, somewhat automatic, and “fast” judgment system, rather than the controlled, deliberative, and “slow” judgment system (Kahneman 2011). Although some judgments in criminal justice are of the “fast” kind, such as police reactions to rapidly moving street incidents, many judgments in criminal justice are “slow” in this sense, as with most actors in courtroom situations. In these situations, people may work to counter the effects of implicit associations on their judgments. For example, efforts made by judges with juries have potential (Kang 2009). More generally, the experience of one’s own implicit biases, using the IAT or similar paradigms, can be dissonant, especially when these biases run counter to one’s explicit beliefs in fairness and objectivity. Such experiences may motivate attempts in the slow system to identify and counter potential biases coming from the fast system. Many social psychologists aim to produce such enlightenment effects in the classroom; how well these effects actually motivate and produce real change is of course unknown. More research is warranted on whether such experiences by real criminal justice actors, such as police officers, prosecutors and defense attorneys, probation officers, and judges (rather than convenience samples of college students), lead to real change.

Because the stereotypes producing these effects are ubiquitous in American society, we do not discuss countering implicit bias as a potential “solution” in this paper. We do not know how significantly these stereotypes contribute to the larger societal problem of disparate treatment—beyond the structural conditions that produce these biases. Rather, we treat implicit bias as a background aspect of the problem.

The rest of our discussion of the problem is organized in three large categories: the justice system, other community aspects, and family and individual aspects. Possible aspects of the solution are organized similarly.

Justice System


Boys and young men of color are subject to more surveillance by police in their neighborhoods, partly by virtue of more often living in high-crime neighborhoods than their white counterparts. A recent high-profile example of increased surveillance is the stop-and-frisk program in New York City. A related example is the “driving while black” phenomenon. This increased surveillance leads to more justice involvement for blacks relative to whites.
In addition to being subject to increased surveillance, black and Hispanic men are subjected to police use of force far beyond their proportions in the population. When force and lethal force are appropriate for police is a matter of long-standing debate and controversy, and police departments vary widely in both their appropriate use-of-force policies and their frequency of using force. The degree to which racial disproportions in the use of force are a function of differential citizen behavior and demeanor, differential surveillance of communities of color, and/or differential willingness of police to use lethal force with black and Hispanic men remains a topic of considerable controversy and mixed research findings (Eith and Durose 2011; Fyfe 1982).

In practice, police have considerable discretion on the street, which can be exercised in disparate ways. Under conditions of stress, which bring out implicit biases in threat perceptions and which can be compounded by gaps of cultural competence in interpreting behavior in unfamiliar communities, differential exercise of discretion can have lethal consequences.

Beyond implicit biases, policing also involves many instances of explicit bias. Even if law enforcement agents act on racial bases only as frequently as American society at large, police officers’ weaponry and ability to use force means that racism in policing has particularly egregious consequences. Some notorious cases result in successful prosecution, as when New York City police officers were sentenced to prison for the 1997 beating and sodomizing of the Haitian Abner Louima. But for every such successful prosecution, undoubtedly many other instances of explicit police racism are not reported, not investigated, or not prosecuted.

The failures of grand juries to indict police officers in many civilian deaths—notably in New York City’s 2014 Garner case, which was captured on video—further indicate to many community members that the criminal justice system writ large devalues the lives of African Americans. The perception that police officers can treat communities of color and their young men with impunity has undermined the trust that is necessary for law enforcement to effectively partner with communities.


Sentencing policies that have driven the current mass incarceration in the United States have had a particular impact on boys and young men of color. In particular, the War on Drugs has resulted in vastly disproportionate rates and lengths of sentences for boys and young men of color in at least three ways. First, drug offending in poor urban areas, including low-level marijuana offending, is more often conducted in public spaces that are subject to police surveillance and intervention. Second, street-level drug markets have often been in poor neighborhoods of color, with recruitment of low-level sellers in those communities who were subject to waves of police crackdowns and arrest. A focus on supply-side (versus demand-side) responses thus had a disproportionate impact on boys and young men of color. A third reason has been the imposition of mandatory minimum sentences that considerably lengthened imprisonment terms. The racial/ethnic effect was exacerbated by the differential in federal mandatory sentences for cocaine: mandated sentences for possession of small amounts of crack were equivalent to sentences for 100 times the quantity of powder cocaine. These draconian sentences for crack were in response to a perceived crack epidemic, which was associated with enormous street-level violence. Because crack was more prevalent in minority neighborhoods, these different sentencing policies greatly increased the sentence lengths for boys and young men of color.

Finally, a wave of policy change in the late 1980s and 1990s, in response to large increases in violence and the illusory “superpredators,” mandated or facilitated the prosecution of adolescents younger than 18 in adult court (commonly known as “transfer” or “waiver”), mostly to increase sentence length. Importantly, longer sentences have little deterrent effect; in fact, they may increase the risk of reoffending upon release (Loughran et al. 2009; Nagin 1998; Nagin, Cullen, and Johnson 2009). In addition, transfer to adult court has not been found to reduce violence (McGowan et al. 2008). Given the disproportionate involvement of minority youth in juvenile justice, this policy change inevitably increased the number of boys and young men of color behind bars and exacerbated their incarceration length. Some evidence also suggests that when the transfer of juveniles to adult court has been discretionary, disproportionately more youth of color have been transferred (Bortner, Zatz, and Hawkins 2000). As part of a broader societal change, there was a shift in the 1980s and the 1990s away from rehabilitative approaches and toward punitive ones (Allen 2000), with much collateral damage (Mauer and Chesney 2003). Although the pendulum seems to be swinging back toward acceptance of rehabilitation, especially for youth, enormous work remains to undo the harm of those changes.


A growing body of methodologically rigorous research shows that involvement in the juvenile justice or criminal justice system increases the chances of further justice-system involvement (Liberman, Kirk, and Kim 2014), decreases the chances of graduating from high school (Kirk and Sampson 2013) and gaining employment (Bernburg, Krohn, and Rivera 2006), and increases the chances of more offending (Huizinga and Henry 2008; Petrosino, Turpin-Petrosino, and Guckenburg 2010; Wiley and Esbensen, forthcoming). These “labeling” effects strongly suggest that juvenile justice and criminal justice involvement should be reserved for the narrowest possible set of offending and that the justice system is the least-preferred avenue for providing needed services to at-risk youth and their families.


Jurisdictions often have financial incentives to impose longer sentences or keep youth in custody. For example, in some states sentences of longer than one year are served in state institutions, at state expense, but shorter sentences are served in local jails, at local expense. Acting in accord with such perverse incentives tends to raise public costs overall, while the detrimental effects of longer incarceration periods farther from family are imposed disproportionately on boys and young men of color and their families.

Community Aspects


The overrepresentation of youth of color in the juvenile and criminal justice systems partly stems from the structure of the neighborhoods in which the least advantaged of them live (Sampson 2012). Neighborhoods with high concentrations of economic deprivation, residential instability, and family disruption—which are overwhelmingly neighborhoods of color—provide an ecological niche for crime to flourish (Sampson and Groves 1989; Sampson, Raudenbush, and Earls 1997). Racial residential segregation also concentrates poverty in communities of color and increases the number of motivated, would-be offenders while tending to undermine communities’ “collective efficacy” to manage social disorder by exercising informal control (Sampson, Raudenbush, and Earls 1997). At the same time, such conditions undermine residents’ sense of the legitimacy of legal actors, thus impairing residents’ ability to harness formal institutional resources to maintain order (Sampson and Bartusch 1998; see also Goffman 2014).


Much of the violence in low-income neighborhoods stems from conflicts between gangs (or crews), whose membership consists largely of boys and young men of color—a reflection of the population in neighborhoods with concentrated poverty. This has resulted in strikingly disproportionate rates of death by homicide among boys and young men of color relative to their white peers. In neighborhoods afflicted by crime and disorder associated with gang activity, there is mutual distrust between police and residents. In these neighborhoods and among boys and young men of color, in particular, there are strong norms and narratives that do not discourage violence and disorder (D. M. Kennedy 2008).


Discipline. Schools’ disciplinary responses are often entry points to juvenile justice involvement, espe-cially when law enforcement is involved in those responses. Zero-tolerance policies that mandate harsh responses, even to minor infractions, tend to be destructive. Moreover, there is increasing evidence that when disciplinary responses are discretionary, youth of color receive harsher and more formal responses including suspensions, arrests, and court referrals more often than white youth, beginning as early as preschool (Fabelo et al. 2011; US Department of Education Office for Civil Rights 2014).

More generally, responses to status offenses—that is, behaviors that are not crimes for adults, such as truancy or running away—have the potential to ensnare youth in the juvenile justice system, in a way that can play out to the youth’s detriment. Although there has been considerable progress in reducing incarceration for status offenses, considerable variation remains in laws, policies, and practices for responding to status offenses in ways that promote healthy development (Kendall 2007).

School safety. One response to school safety concerns, especially since the 1999 school shooting in Columbine, has been to increase the presence of law enforcement in schools, often called “school resource officers.” Policy responses have also been driven by the horrific, but rare, school-shooting incidents, rather than the much larger problem of everyday school violence that disproportionately involves communities of color.

There is no strong evidence to date that school resource officers have decreased serious violence in schools—and shockingly, almost no methodologically strong studies have been conducted—but there is evidence that the presence of school resource officers increases arrests and court referrals for low-level issues that would otherwise have been handled informally by schools (Na and Gottfredson 2013). However, there is also remarkably little evidence for effectiveness of other approaches to school safety (Cook, Gottfredson, and Na 2010).

Family and Individual Aspects


Exposure to violence. Approximately two-thirds of children and youth have experienced or witnessed at least one victimization, and nearly 40 percent had experienced more than one type of direct victimization in the past year (Finklehor et al. 2009). As boys and young men of color are more likely to live in high-crime neighborhoods, they are exposed to higher rates of violence than their white counter-parts. Children exposed to violence experience physical, mental, and emotional trauma that can lead to attachment difficulties, anxiety, depression, behavioral problems, and the perpetration of violence (Boivin et al. 2012; Buka et al. 2001; Kilpatrick, Saunders, and Smith 2003). Moreover, when victimized, boys and young men of color are often identified and treated as perpetrators of crime, rather than as deserving of victim assistance, as documented in the Vera Institute’s Common Justice project.4

Parental arrest and incarceration. Millions of children have been exposed to parental arrest, and youth of color are at a higher risk of this exposure. Immigration enforcement actions against parents also affect hundreds of thousands of youth, particularly those from Latin American countries. Witnessing parental arrest, through police or immigration enforcement agencies, is a traumatic experience for children, particularly because arrests can include handcuffing, drawing a weapon on, or subduing the parent. Parental arrest could be a temporary experience for children, leading to several hours of parent–child separation, or a more prolonged separation when a parent is detained in jail overnight or longer or, in the case of immigration cases, deported. Children of incarcerated parents are more likely to experience financial hardship, residential instability, changes in caregiver arrangements, and trauma associated with the loss of a loved one, all of which may translate into short- and long-term mental and physical health issues, poor academic performance and achievement, substance abuse, and delinquency (Foster and Hagan 2007; Geller et al. 2009; Glaze and Maruschak 2008; Murray and Farrington 2008; Murray, Janson, and Farrington 2007; Phillips and Gleeson 2007; Phillips et al. 2002; Smith et al. 2007; C. A. Walker 2003; Wildeman 2011).


Communities with high levels of concentrated poverty also tend to have high levels of need in such areas as mental health and substance abuse treatment and family therapy. However, treatment resources are often inadequate and/or unavailable for financial reasons. All too often, therefore, services are not provided until youth are involved in the juvenile justice system, which can mandate (and pay for) services. Because many of these service needs are also risk factors for delinquency and criminal involvement, the perverse effect of financing services only after youth are formally involved in juvenile justice is to increase the net involvement in juvenile justice and to add to the costs of justice system involvement.

Especially when treatment and services are lacking outside of justice systems, their availability through court mandate can lead to net-widening, in which system actors retain youth who would other-wise be diverted or released from juvenile justice and criminal justice supervision in order to provide services. The court mandate to participate in treatment also tends to include penalties for treatment “noncompliance.” In diversion programs, noncompliance penalties can be larger than the penalties that would have been faced because of the initial offense. Poor minority youth are particularly likely to present to the court with such treatment needs, and adolescents are particularly likely to fail to comply.


There is tremendous opportunity for effective policies and practices across the juvenile and criminal justice system to reduce disproportionate entry and confinement of boys and young men of color and to produce better outcomes for those who do enter the system. Our discussion of the solution set focuses on the justice system. In the following section, we discuss relatively little of the growing prevention literature, which includes school-based, family-based, and individual-based interventions that focus on increasing positive youth development and preventing risk generally, along with associated effects on reducing youth engagement in crime and delinquency. Some of these efforts are discussed in our papers on healthy children and families and educational opportunity.

Given the historic pattern of the disproportionate negative impact of justice policies on communities of color, one important set of policies concerns the need to explicitly consider the race-specific effects of policies and practices through racial impact statements, race- and ethnicity-specific validation of risk and needs instruments, and monitoring how reform efforts benefit different racial and ethnic groups.

The solution set discussed below primarily involves removing or reforming policies that put surveillance, criminalization, and retribution at the forefront of responses to misbehavior. Some of these policies were crafted at a time when there was little evidence that other approaches were effective and there was a broad belief that “nothing works” (Martinson 1974; Sechrest, White, and Brown 1979). At present, enough evidence has accumulated to support the strong advocacy of policies that are rehabilitative, and restorative and that promote positive youth development. Because justice system involvement has its own detrimental side effects, the justice system should also be the last resort for promoting these ends.

The Challenge of Evidence

The history of juvenile justice contains object lessons of efforts aimed to improve the lives of youth that instead caused harm. In a famous example, the Cambridge–Somerville study was initiated in 1939, with random assignment. Its multiyear intervention for at-risk youth included after-school activities, tutoring, and a summer camp. After 30 years, participants had worse outcomes, including more convictions, diagnoses of alcoholism or psychiatric disorders, and early deaths (McCord 1978, 2003).5 Well-intentioned and well-grounded interventions may cause harm, sometimes for unanticipated and subtle reasons. Therefore, it is crucial that program innovation and development be followed by rigorous evaluation to ensure that programs are not doing harm and are instead producing good. Today’s declining rates of juvenile crime and arrests create a tempting atmosphere in which studies with pre-post designs may spuriously find benefits that will not hold up under more challenging circumstances.

Yet many of the most important policy changes needed are system-level changes in policy and practice and not subject to rigorous evaluation. Nonetheless, such policy changes and system reform efforts should be implemented as much as possible in ways that are “evidence-generating” (Liberman 2009). System change can and should be evaluated, but inherently must rely on methods considerably more open to challenge than for robust individual-level interventions. The tension between the desire for strong evidence and the need to act to improve the lives of boys and young men of color requires the flexibility to act based on the best evidence available, together with an openness to evidence that might disconfirm expectations. Because many of the recommended solutions below involve the removal or mitigation of factors that are causing harm, the burden of proof may be lessened. But effectiveness in improving outcomes in the long run will require experimentation with those policies and interventions and an openness to learning from failure (Berman and Fox 2010). Foundations often seem to have more flexibility to negotiate these tensions than political actors or federal governmental agencies.

Targeted Policies, Strategies, and Interventions

Institutional Reform and Policy Change


Racial impact statements. A promising policy that the Sentencing Project has promoted is the requirement that racial impact statements be a prerequisite to any proposed juvenile justice or criminal justice policy changes, much along the lines of the requirement for environmental impact statements in large construction projects (Mauer 2009) or the health impact statements that are being integrated into local policy decisions. Iowa, Connecticut, and Oregon have mandated racial impact statements, and Minnesota’s sentencing commission has drafted such statements for new legislation.6 The use of racial impact statements would bring the question of disparate minority impact into the light before policy is made and would put policymakers on record for the racial impact they anticipate. Should policies actually produce disparate racial outcomes that were not projected, such outcomes would be important grounds for reversing or changing those policies.

Local validation of risk instruments with subpopulations of color. One important reform in juvenile justice and criminal justice concerns the use of validated assessment instruments to gauge risk of (re)offending and service needs. When the instruments are appropriate, the use of objective criteria from such instruments to drive probation recommendations and sentencing decisions (referred to as “disposition” in juvenile justice) has the potential to improve public safety and promote positive youth development. The use of objective criteria also has the potential to reduce the implicit or explicit role of racial and ethnic bias in decisionmaking.

However, the appropriateness and adequacy of such instruments for different populations and in jurisdictions other than where they were developed are less clear. For example, there has been considerable controversy about how well instruments developed with men apply for women (Van Voorhis 2012). Similarly, few validation studies have explicitly considered equivalence for subpopulations of differing race and ethnicity (Baird et al. 2013). Equivalence is important because important predictors as well as outcomes (reoffending) are often assessed through official criminal justice records, whether on arrest, adjudication, or institutional commitment. In particular, criminal history and the age of first arrest (or first adjudication or commitment) consistently emerge as strong predictors of reoffending. The possibility that policies and practices producing those official records result in earlier and/or more extensive criminal histories for boys and young men of color makes it important for risk instruments and their classification schemes to be validated locally by race and/or ethnicity. Foundations would seem well-poised to support the revalidation and renorming of instruments in use for different populations in specific local jurisdictions.

Monitoring race-specific outcomes. The consistent overrepresentation of boys and young men of color in the juvenile justice and criminal justice systems means that almost any effort at juvenile or criminal justice reform that aims to reduce juvenile justice and criminal justice involvement, incarceration, or the detrimental effects of justice-system involvement would be expected to particularly benefit boys and young men of color. But this may not always be the result. For example, although drug court has been promoted as a diversion for drug-involved offenders, there are persistent claims that drug courts involve boys and young men of color less often than whites, although research on this issue remains in a formative stage. It is important that the results of such general efforts be monitored and evaluated to ensure that they indeed benefit boys and young men of color.


Law enforcement can be a critical partner in efforts to reduce, mitigate, and reverse both the overrepresentation of boys and young men of color in the justice system and the cascading negative consequences of that involvement. However, this partnership is hampered by the reality and perception of racially biased policing. The enduring nature of the problem suggests that internal mechanisms (e.g., training, supervision, and internal review) alone will often prove insufficient, and that external mechanisms of oversight and enforcement by citizens, courts, and federal civil rights investigations and oversight will remain important and necessary tools in the efforts to achieve racially neutral and racially sensitive policing (White 2010).

Use of force. While lethal use of force is statistically rare in police-citizen contacts, it falls heavily on people of color and is especially corrosive for police-community relations. One necessary step in the prevention of lethal use of force is the development of clear policies that limit its use to incidents of serious threat. Policies alone, however, are insufficient. In the death of Eric Garner in New York City in July 2014, for example, police seemed to use a choke hold that had been banned as a matter of policy. Training and accountability are critical complements to policy. Training police in de-escalation techniques also has potential to prevent incidents from becoming lethal (Police Executive Research Forum 2012).

TASERs (conducted energy devices) are a technology that might reduce the lethal use of force, if used by police officers as a less-lethal alternative to firearms. However, the use of TASERs themselves has been fraught with controversy—about their effectiveness, about how often they are used unnecessarily, and about whether they too are used disparately against people of color. Too little data have been systematically collected, or research conducted, to allow clear conclusions to be drawn to date (White and Saunders 2010).

The use of body cameras in policing seems promising for reducing the use of force, based on early research (White 2014). For example, a randomized study in Rialto, California, has found that the use of body cameras considerably reduced both police use of force and citizen complaints (Farrar 2013).

Legitimacy. In neighborhoods with concentrated poverty and afflicted by crime and disorder, which are disproportionately communities of color, there is mutual distrust between police and residents (D. M. Kennedy 2008). Because of the lack of trust (or perceptions of what is known as “procedural justice”), residents in communities of color are less likely to perceive police officers as legitimate authorities, which can result in unwillingness to report crime, to work with officers to prevent crime, and to follow the law (Tyler 2006; Tyler and Fagan 2008; Tyler and Hou 2002). Trust and confidence in the police is understood to be the foundation by which citizens cooperate with the police (Rosenbaum et al. 2005). The US Department of Justice has recently announced a new National Initiative on Building Community Trust that will include developing and pilot testing initiatives to improve police legitimacy and community relations.

A first step is to change law enforcement policies that inequitably target communities of color. For instance, with support from Open Society Foundations and Atlantic Philanthropies, among others, foundations have supported efforts to roll back the stop-and-frisk policies in New York City that have so antagonized minority residents and had the side effect of increasing the rate of arrest of boys and young men of color for low-level offenses. Once law enforcement collaborates with minority communities as active partners in promoting public safety, rather than primarily as targets of surveillance, then police departments can enhance police legitimacy, perceptions of procedural justice, and public confidence by providing residents with a voice in the policing process, increasing transparency in their processes, and treating citizens with respect (Hawdon, Ryan, and Griffin 2003; Sunshine and Tyler 2003; Tyler 2004).

A second step may be to expand specific policing interventions, such as foot patrols, which have been shown to increase citizens’ perceptions of safety and enhance police–community relations (Kelling et al. 1981; Trojanowicz 1983; S. Walker and Katz 2002) in addition to reducing violence (Ratcliffe et al. 2011). Community policing, which includes problem-oriented policing strategies, the development of collaborative community partnerships, and organizational transformation to support community partnerships and proactive problem solving (Office of Community Oriented Policing 2009), is a promising model for increasing resident satisfaction and trust in the police (Hawdon, Ryan, and Griffin 2003). However, because these approaches can increase surveillance in ways that may contribute to disproportionate police stops and arrests in poor minority communities, it is critical for such law-enforcement initiatives to be mounted in collaboration with the affected communities.

Model arrest policies to reduce deleterious effects on children. Policing reforms should include policies and practices for children experiencing parental arrest, including children whose parents are detained by Immigration and Customs Enforcement. Given the deleterious outcomes associated with parental involvement in the justice system (outcomes theoretically similar for children whose parents are detained or deported by Immigration and Customs Enforcement), there is tremendous potential for the justice system, Immigration and Customs Enforcement, and service providers to identify these children and offer interventions and supports to reduce trauma and reduce their likelihood of entry into the justice system.

Although police departments often have procedures in place with child protective services in cases of child abuse and neglect and domestic violence (Pence and Wilson 1992), procedures are lacking with human services agencies and other nonjustice agencies in all other cases (International Association of Chiefs of Police 2014). Model arrest policies for police officers and Immigration and Customs Enforcement agents that consider the immediate physical safety, security, and psychological and emotional well-being of children are necessary. Additional support could expand the use of innovative model arrest policies that support collaborative partnerships and cross training between police and human services agencies, recommend prearrest planning with the well-being of children in mind, and promote the view of parents as the best resources for identifying the child’s caregiver at the time of their arrest. The US Bureau of Justice Assistance in partnership with the International Association of Chiefs of Police is currently promulgating approaches like these across the United States. Although not yet widely implemented or rigorously evaluated, such approaches have strong potential to reduce trauma for children and increase their well-being.


The drug war’s disparate impact on communities of color has been extreme enough that some have called for a broad reversal of the drug war, including decriminalizing marijuana and other drugs as well as removing the downstream consequences of drug involvement such as restrictions on living in public housing (Alexander 2010). At the very least, especially in jurisdictions with demonstrated racial disproportionalities in arrests for marijuana use (such as the District of Columbia), decriminalization of marijuana possession could reduce future disparate formal involvement in the justice system.

There has been considerable movement toward sentencing reforms, including eliminating mandatory minimum sentences, changing policies toward early release, eliminating or limiting three-strikes provisions, and mandating community-based treatment rather than incarceration for drug offending. Crime rates remain historically low, and juvenile arrest, adjudication, and institutional commitment rates have been falling, thereby opening the political space for policy reform. The past several years have witnessed the rollback of some punitive sentencing policies that have had such devastating effects on the number of boys and young men of color behind bars. The federal crack-versus powder cocaine sentencing disparity was rolled back from 100 to 1 down to 18 to 1 by the 2010 Fair Sentencing Act, and the five-year mandatory minimum sentence for crack possession was eliminated; efforts are under way to make those changes retroactive. In addition, California’s Proposition 36 has required treatment rather than incarceration for first-time drug offenders, and states are rolling back the requirements to facilitate or mandate the prosecution of many adolescents younger than 18 in adult courts.

Presumably, the rolling back policies that have produced mass incarceration would benefit boys and young men of color to the same degree as mass incarceration itself has disproportionately affected communities of color. However, many of the draconian policies removed discretion, especially from judges, and reversing these policies would restore discretion, which has often been associated with disparate treatment of boys and young men of color. Therefore, the implementation and racial effects of such policy changes warrant careful monitoring. This monitoring will require sustained attention and would seem an obvious opportunity for foundation support.


Effective gang intervention and prevention programming would result in fewer boys and young men of color being shot and killed and fewer boys and young men of color entering the justice system for these crimes. Gang intervention and prevention programming, through strong and effective partnerships among police, residents, and social service agencies, has the potential to tackle the problem of gang-based violence, create new norms and narratives about crime and violence, and therefore reduce and prevent gang-based violence and increase social control and collective efficacy, which are associated with lower crime and violence (D. M. Kennedy 2010; Sampson, Raudenbush, and Earls 1997).

Two promising approaches have emerged. Gang- or group-based violence prevention strategies, such as Boston Ceasefire, the federally funded Strategic Approaches to Community Safety Initiatives, and Project Safe Neighborhoods, have been replicated in various cities. Their premise is that a small number of youth, operating in groups, are involved in conflicts that may be linked to the vast majority of homicides in a city (D. M. Kennedy et al. 2001). Focused deterrence strategies can then effectively reduce violence. These group-based strategies are implemented through partnerships among police, community corrections, social service agencies, and community residents. They deliver an explicit antiviolence message to identified gang members in a group setting, deliver group-based and targeted notifications of the resources police will use to suppress and respond to additional violence perpetrated by the gang or group, and inform group members that community and social services providers are available to support and help them desist from crime and find legitimate employment. A systematic review of focused deterrence strategies found significant reductions in targeted violence outcomes (Braga and Weisburd 2012).

CureViolence, formerly known as Chicago CeaseFire, has been implemented in various cities using a public health framework to intervene or interrupt gang violence through a combination of street or outreach workers and service providers. Although the effectiveness of the CureViolence/Chicago CeaseFire model has been contested, studies show that it can reduce gang violence where it has been implemented (Skogan et al. 2008; Webster et al. 2012).


The current climate for juvenile justice system reform is promising because fiscal constraints are motivating stakeholders are all levels of government to revisit costly punitive approaches that may not be buying public safety, while there is now a critical mass of evidence concerning effective alternatives.
Foundations have been critical agents of change for broad juvenile justice reform efforts, each taking a slightly different approach. The Robert Wood Johnson Foundation’s support for Reclaiming Futures has aimed to reform treatment of drug-involved offenders. The Annie E. Casey Foundation had been a strong force in reforms designed to eliminate unnecessary detention of pretrial youth, with a consistent focus on issues of racial equity. Building upon that work, the Annie E. Casey Foundation is now embarking upon its “deep-end” initiatives that aim to reduce the number of youth in commitment or out-of-home placement, while maintaining a focus on minority youth. The Pew Charitable Trusts has partnered with the Bureau of Justice Assistance to move states away from reliance on prison.

The John D. and Catherine T. MacArthur Foundation’s Models for Change initiative has stressed local innovation and reform in many domains of juvenile justice reform, along with providing considerable technical assistance. This initiative followed the foundation’s support for two other projects: the Research Network on Adolescent Development and Juvenile Justice, which developed and synthesized evidence concerning developmentally sensitive approaches in juvenile justice; and the Project on Human Development in Chicago Neighborhoods, which studied the interplay between neighborhoods, adolescent development, and crime.
These efforts highlight critical roles that foundations can play in supporting juvenile justice reform efforts, which often require sustained advocacy, technical assistance, and local stakeholder commitment. These efforts have included public–private partnerships with federal research agencies in the US Department of Justice and the National Institutes of Health, especially in supporting research to undergird reform efforts. Foundations have often taken the lead in advocacy efforts. For instance, the Hayword Burns Institute, with funding from California Endowment, has been working with sites to engage multiple stakeholders, use a data-driven approach to identify grassroots local opportunities to reduce racial disparities, and reduce initial involvement in the justice system.

We turn to several aspects of juvenile justice reform that seem particularly promising.


Restorative justice. Restorative justice approaches aim to restore the harm created by a crime, generally involve both offenders and victims, and tend to involve reparation as well as dialogue involving victims, offenders, and other stakeholders. Restorative justice has varying models including conferencing models, sentencing circles, and restorative boards as alternatives to formal justice system processes. These approaches have considerable appeal because they can divert offenders from the formal justice system while serving victims’ interests. Evaluations have tended to find that both victims and offenders are more satisfied than with standard responses (Sherman and Strang 2007). Because restorative justice requires the consent of both offenders and victims to participate, it is not appropriate for all cases.

Restorative justice is particularly appealing in school settings, where it aims to heal the relationship between participants who know and are in frequent proximity to each other.6 For example, the Alameda County Restorative Justice program has been using restorative justice as an alternative to formal justice system involvement in response to incidents, including at schools, involving youth.

Trauma-informed approaches to working with juvenile justice youth. Given the harmful outcomes associated with childhood exposure to violence, early identification and intervention by families, teachers, criminal justice professionals, and other social service providers are critical. For boys and young men of color, there is great potential for school administrators, service providers, and the justice system to use interventions and practices that alleviate trauma, meet needs, and restore well-being, as well as offset later offending patterns and poor behavioral responses or adaptations. Cognitive behavioral therapies, implemented in group or individual settings, have been found to reduce the psychological harm stemming from exposure to traumatic events, according to a systematic review conducted for the Centers for Disease Control and Prevention’s task force on community prevention (Wethington et al. 2008).

Juvenile probation reform. Probation and parole officers tend to have considerable discretion in whether to send an individual to institutional custody in response to violations of technical requirements of community supervision that do not involve criminal behavior (Jacobson 2005). In addition, community supervision agencies and officers vary in their philosophy concerning punitive versus rehabilitative goals of community supervision, and many juvenile probation and/or parole officers lack specialized training in adolescent development or how to manage adolescent populations. Juvenile probation reform to clarify goals and policies, and training in adolescent development and evidence-based practices, seems to have potential to improve outcomes for the boys and young men of color who are disproportionately found in juvenile justice (Young, Farrell, and Taxman 2013).

Education in juvenile justice settings. Educational deficits are an important risk factor for juvenile justice involvement (Hawkins et al. 1998; Henry, Knight, and Thornberry 2012). Justice system involvement also tends to interfere with high-school completion (Kirk and Sampson 2013). Therefore, when juvenile justice system involvement is necessary, and especially if residential confinement is required, this involvement should be used as an opportunity to remediate educational deficits and to promote high-school completion. For example, the Department of Youth Rehabilitative Services in the District of Columbia has been lauded for its Maya Angelou Academy in its residential facility.


Even with the most optimistic assessment of reforms to reduce entry into the justice system, the legacy of mass incarceration will remain for decades. There is a strong need for evidence-based programming to improve the odds of successful reentry and reduce recidivism rates. Successful reentry has tremendous potential to improve the lives of boys and young men of color and to restore communities who have been harmed by the removal of a substantial fraction of their populations (Clear 2007).

Research strongly suggests that reentry programming tailored to criminogenic risk (of return to reoffending) and matched to areas of need is effective (Latessa, Cullen, and Gendreau 2002). Not every individual needs the same level of services in prison or after release (Andrews and Bonta 1998; Bonta and Andrews 2007; Weibush, McNulty, and Le 2000).

Generally, educational training, transitional employment programs and job training, behavioral health services, transitional and permanent supportive housing, and family-based programs—such as Functional Family Therapy for juveniles, parent–child/family visitation policies for adults, and evidence-supported parenting programs (e.g., Parenting Inside Out)—have been effective with the reentry population (see the What Works in Reentry Clearinghouse for reviews of some of these programs).

Reentry programming best practice suggests that services should begin during incarceration after validated assessments of criminogenic risk and needs have been conducted. Services should include transitional services that continue from incarceration to the community for an extended period (Altschuler, Armstrong, and Mackenzie 1999; Petersilia 2004; Visher and Travis 2011).

However, some evaluation efforts have found that reentry programs that provide services considerably below the level of need, or poorly matched to individual needs, may be ineffective and even counterproductive. Reentry needs are enormous, and programs on the cheap may be ineffective (Lattimore et al. 2012; J. A. Wilson and Davis 2006).

Removing barriers to success. Ex-offenders face considerable barriers to conventional employment. Yet employment is a key element of successful reentry. Rethinking the barriers and removing those that unnecessarily hamper success without any public safety benefit would seem a promising strategy. Several states and localities have been adopting “Ban the Box” legislation or executive actions, which ban various questions about criminal records on employment applications and delay criminal background checks until later in the hiring process (NELP 2014).

Redemption. For ex-offenders who have stayed “clean” for a considerable period, vehicles to remove employment barriers are warranted. Because we know that the risk of recidivism declines with time, researchers have been examining of how long a “clean” period could be sufficient to warrant “redemption” (Blumstein and Nakamura 2009a, 2009b). Experimentation with policies and practices are now needed to allow such redemption in practice. Blumstein and Nakamura raise such possibilities as educating employers about the declining predictive value of criminal records over time, establishing pardon boards, and creating certificates of rehabilitation. Foundations would seem well-positioned to work toward the development and testing of approaches through public–private partnerships or with the private sector alone.


States such as Ohio and Illinois have reformed juvenile justice financing to remove perverse financial incentives to send delinquent youth to state facilities rather than keeping youth in the community or in local institutions. Reclaim Ohio and Redeploy Illinois provide models for carrying out such financial reform for juvenile justice. California’s “Realignment” has been especially aggressive in moving responsibility for incarceration and supervision from the state to counties. The California model, like the models in Ohio and Illinois, reimburses counties for individuals who are kept under local rather than state control.

In a related vein, Justice Reinvestment has attempted to move funds from incarceration to much less expensive community treatment and service alternatives (La Vigne et al. 2014). These efforts have had some success in reducing reliance on incarceration; they have had less success in redirecting those funds toward community treatment and services.

Community Interventions


Discipline and school climate. The US Department of Justice has recently begun engaging with school discipline as a civil-rights matter. This important step for ameliorating detrimental effects on children of color will involve monitoring disparities in disciplinary practices.

Approaches to improving school management and intervening with school discipline problems without juvenile justice involvement would seem an important part of the effort to close the “school to prison” pipeline. The US Department of Education has been promoting Positive Behavioral Intervention and Supports (PBIS), which is a tiered approach to school management and preventing problem behavior (US Department of Education 2014b).

Attention is focused on creating and sustaining primary (school-wide), secondary (classroom), and tertiary (individual) systems of support that improve lifestyle results (personal, health, social, family, work, recreation) for all children and youth by making problem behavior less effective, efficient, and relevant, and desired behavior more functional.8

Such positive discipline approaches have considerable promise as alternatives to zero-tolerance policies for reducing problem behavior and improving school climate. These approaches would also seem to have promise for promoting school safety that warrants additional development and testing. In addition, as discussed earlier, restorative justice models seem to have particular appeal in schools.

Another approach that has proved effective is reducing disruptive behavior is the Good-Behavior Game, which harnesses peer influence to reduce misbehavior. The program reduces the need for disciplinary intervention by adults, and there have been some reports of long-term benefits to students exposed to the program in early grades (Poduska et al. 2008).

Universal school-based violence prevention programs. Schools have enormous potential as settings for prevention activities, for positive social and emotional learning, and for reform efforts to improve academic achievement, all of which should help prevent crime, delinquency, and antisocial behavior. Universal programs delivered to entire school classes or entire schools (also known as primary prevention) have particular appeal because they do not label or target any youth as “at risk,” which may have deleterious side effects. Two systematic reviews have found that universal school-based violence prevention programs are effective in preventing violent and aggressive behavior (Hahn et al. 2007; S. W. Wilson and Lipsey 2007). In addition, one such program, Gang Resistance Education and Training, has recently demonstrated effectiveness in reducing gang affiliation and violence significantly (Esbensen et al. 2013).


Many juvenile justice youth and their families are involved in multiple public systems that are uncoordinated and may work at cross purposes. One particular area that has received attention has been youth involved in both child welfare and juvenile justice. Youth in households with domestic violence often also require multiple agency involvement. Efforts to improve coordination of child welfare, law enforcement, and juvenile justice in order to improve youth outcomes require a determined multiyear effort, with considerable adaptation and flexibility, as demonstrated by an effort conducted in New York City by the Vera Institute of Justice (Ross 2009). They also require considerable attention to local needs and local systems operations. Foundations would seem well-positioned to support such efforts.

Family and Individual Interventions


Programmatic interventions focused on youth who are either at risk for juvenile justice involvement (secondary prevention) or already involved with juvenile justice (tertiary prevention) could reduce the disproportionate entry and reentry of boys and young men of color into the justice system. Efforts to identify effective name-brand interventions have been developed through the Blueprints for Healthy Youth Development at the University of Colorado and the Washington State Institute for Public Policy. Individual-level interventions that have been effective for juvenile justice populations include cognitive behavioral therapies (e.g., moral recognition therapy) and programs that focus on helping individuals improve their social skills, critical reasoning, impulse management, and self-control (Landenberger and Lipsey 2005; Lipsey 2009). Lipsey and colleagues (2010) also find that generic interventions of the same type as some of the name-brand programs (e.g., family therapy versus Family Functional Therapy and Multisystemic Therapy) have similarly beneficial effects, if delivered with fidelity.

The Washington State Institute for Public Policy has also conducted cost–benefit analyses of these interventions that have shown that public safety would be improved by redirecting incarceration costs to those interventions ( The institute has been able to use this evidence to convince the Washington State legislature to redirect funds that would have been used to build additional prisons toward funding these evidence-based practices (Aos, Miller, and Drake 2006). The Pew Center on the States and the MacArthur Foundation have funded the Washington State Institute for Public Policy to expand this approach to other states under their jointly funded Results First initiative.

The same interventions have been found effective in reducing the risk of juvenile (re)offending, whether delivered in the community or in institutional settings (Lipsey 2009). This demonstrated effectiveness bolsters the conclusion that services available for juvenile justice–involved youth should also be funded for community youth without the need for juvenile justice involvement.

The evidence on effective programs generally contains the strong requirement that the programs be delivered with fidelity. When delivered with fidelity, evidence-based programs tend to be cost-beneficial—they save governments more in reduced criminal justice costs than the programs cost to deliver. However, one key aspect of fidelity for treatment programs is dosage. When treatment programs are delivered with too few contact hours, they often fail to deliver the expected effects (Lipsey et al. 2010; Howell and Lipsey 2012). However, in fiscally tight times, all too often service contracts are written without reference to the program requirements, or budget cuts undermine program effectiveness and cost-effectiveness.

Finally, it is important to note that in the juvenile justice realm, black youth are represented in the evidence base for program effectiveness, perhaps because of the very overrepresentation of youth of color in juvenile justice. This stands in contrast to some issues regarding effective health interventions, for example. Wilson, Lipsey, and Soydan (2003) conducted a meta-analysis of whether “mainstream programs for juvenile delinquency” were differentially effective for minority than majority youth, and found that they were equally effective. African American youth are well represented, and Latino youth are represented to some degree, but Native American youth very little.


Many juvenile justice youth come from families with risk. Although we do not review the prevention literature here, we note a cluster of family interventions that has been found effective in reducing juvenile justice involvement while improving positive youth development outcomes.

Early childhood home visitation. A systematic review of the effectiveness of early-childhood home visitation programs has found them to be effective in reducing child abuse and neglect, with larger effects seen in programs with nurses and mental health workers as visitors (Bilukha et al. 2005). The Nurse–Family Partnership visitation program developed by David Olds, which has reported some long-term (15-year) effects on reducing violence by the visited child, has been designated an effective program by the US Department of Health and Human Services, which has been funding efforts to take the program to scale.

Family therapy. One particular cluster of effective interventions involves family therapy. Two name-brand family therapy programs, Multisystemic Therapy and Functional Family Therapy, have garnered considerable evidence of effectiveness. They have established intermediary organizations (MST, Inc., and FFT, Inc.) to provide training and monitoring. More generally, family therapy programs with effective service delivery and fidelity have been found to be effective in reducing offending among juvenile justice youth (Lipsey et al. 2010).

Multidimensional Treatment Foster Care. Multidimensional Treatment Foster Care is a model of intensive foster care that includes training of foster parents and regular meetings. It has also been found effective in reducing delinquency (Hahn et al. 2004). The developers of Multidimensional Treatment Foster Care have also been active in trying to take the approach to scale and in testing its effective implementation across many counties in California.

Mentoring. Considerable evidence has accumulated on the effectiveness of mentoring programs for at-risk youth (e.g., Big Brothers/Big Sisters), including mentoring for children of justice-involved parents (Dubois et al. 2011; Tolan et al. 2008).

Research and Knowledge Gaps

We note here several important gaps in knowledge that, if filled, would considerably increase our ability to improve the lives of boys and young men of color.

Racial and Ethnic Subgroups

Most of the evidence on the disproportionate impact of the justice system concerns African American youth. There is a need for additional evidence on other subpopulations, including the diverse groups of Latino youth of different immigrant generations. Despite the high levels of disadvantage and violence in Native American communities, the differences in the tribal justice system—which generally must rely on federal systems for functions that would be handled by local or state authorities elsewhere in the United States—have impeded our understanding or ability to propose remedies that would help in tribal jurisdictions.

Similarly, beyond black and white (Wilson, Lipsey, and Soydan 2003), few programs have been tested for differential treatment effectiveness on subpopulations defined by race or ethnicity or in different cultural contexts. We also know too little about how issues concerning cultural competence affect the implementation or effectiveness of different evidence-based programs.

Police Use of Force

Incidents of police contact with civilians in general, and incidents involving use of force in particular, are not systematically documented. Thus, we know too little about the conditions under which force is—and is not—used, and even less about the apparently disparate use of force, despite creative efforts by researchers. The paperwork demands on street officers suggests that systematic and uniform collection of such data by traditional means may be unrealistic. One particular technology, body cameras worn by police, seems to have considerable potential to increase the possibility of systematically allowing the study of use-of-force incidents and how they are handled, beyond their potential as an intervention. However, before we can take the resulting camera-generated data at face value, we will need more research concerning how and when such cameras are used, and how to manage the burden involved in monitoring or reviewing the resulting video.

Effective Gang Interventions

Although gangs can generate considerable delinquency and violence, there is a paucity of effective interventions. Gangs have been notoriously resistant to interventions, and many attempts to disrupt or redirect gangs have instead strengthened gang identity and solidarity. Although there are now two promising approaches to intervening with gang violence, as described above, there are as yet no evidence-based individual-level prevention interventions for already-involved gang youth. One promising approach is to take individual-level interventions that have worked with delinquency prevention and adapt them for gang-involved youth. This approach is currently being tested by Terrence Thornberry of the University of Maryland.

Drivers of Disparities in School-Discipline

Although the evidence on disparities in school discipline is compelling, we know too little about how and why the disparities occur. For example, aggregate disparities in discipline for students of different race/ethnicity may result from within-school disparities or from differences between schools with different racial or ethnic makeups. And, within- and between-school disparities may be amenable to different remedies.

School-Safety Approaches That Empower Students

Despite the lack of evidence on the effectiveness of school resource officers in promising school safety, there is a dearth of strong evidence on alternative approaches. In particular, although there is one promising report (New York Civil Liberties Union 2009), we have been unable to find any strong tests of approaches to school safety that empower students and change their roles from targets of suspicion to agents of safety in ways that also promote youth development.7 Foundations would seem well-positioned to work with schools and school systems to develop and rigorously test school-safety approaches that also support developmental goals for students.


The endemic disproportionate involvement of boys and young men of color in the justice system means that numerous avenues for juvenile justice and criminal justice reforms could reduce harms and improve the lives of boys and young men of color. An important first step has been to bring these disproportional responses to boys and young men of color into explicit focus, such as in the federal My Brother’s Keeper initiative and New York City’s Young Men’s Initiative. A natural extension of this work would be to make explicit consideration of the racial impact of proposed policies, as in racial impact statements, into routine tools to help prevent future juvenile justice and criminal justice policies from having similar effects.

The history of disproportionate juvenile and criminal justice response to boys and young men of color is accompanied by a long narrative of disproportionate use of lethal force against young men of color. The recent large demonstrations, in response to the high-profile deaths of unarmed black men in New York City and Ferguson at the hands of police, highlight how even a few such incidents can inflame communities and undermine the trust and cooperation between police and citizenry that is most productive of community safety. Rebuilding trust will require considerable public changes in police attitudes in jurisdictions where it has eroded, strong policies regarding the use of force and the exercise of police discretion, and robust structures to enforce those policies.

The growing evidence concerning the cascade of detrimental consequences of involvement in juve-nile and criminal justice means that reversing the growth of such involvement, which is the legacy of the past 25 years of punitive policymaking, should be a large part of the effort to improve the lives of boys and young men of color. This reversal will primarily involve the removal of counterproductive policies that lead to unnecessary and detrimental arrests, detention, and incarceration. In some domains, it will also require development and rigorous testing of alternative approaches, as for school safety.

More generally, the justice system should be used sparingly as a vehicle for improving youth outcomes because any benefits come with possible harms. Therefore, it is crucial to develop a robust infrastructure of effective primary and secondary prevention for at-risk youth that is completely outside the juvenile justice system. The problem of successful reentry for those who have been incarcerated to reverse the residual effects of mass incarceration also continues to loom large.

We would be remiss not to acknowledge a more radical perspective, exemplified by Michelle Alexander’s The New Jim Crow (2010). Alexander argues that the long-term social stigmatization and exclusion from societal participation of many black men, through current criminal justice policies, has produced a more pronounced system of social exclusion status than even Jim Crow. This forces the conclusion, she believes, that the criminal justice system is designed to maintain blacks as a subservient “caste.” From such a perspective, some of the preceding discussion of potential “solutions” naively approaches the problem of racial overrepresentation in criminal and juvenile justice as a side effect rather than a function of the system. Such a perspective would likely acknowledge that many of the preceding solutions may improve the lives of boys and young men of color at the margins, compared to not doing them. Yet it would also argue that making the serious policy changes to criminal and juvenile justice that are needed to substantially improve the lives of boys and young men of color will require a sustained political effort that is mounted in anticipation of considerable resistance. In Alexander’s words, “Tinkering is for mechanics, not racial-justice advocates” (2010, 230).

Whether or not one entirely accepts this perspective, it highlights the importance of nongovernmental agencies and foundations in this work. The executive branch of the federal government can play a critical role in the kind of needed justice reforms, as exemplified by Attorney General Holder’s recent efforts and the My Brother’s Keeper initiative. Yet the more radical the changes needed, the more important it will be for them to be advocated or supported externally. Foundations can play a critical role in supporting the development and testing of regimes that operate differently from business as usual.


1. Specific groups of Asians and Pacific Islanders receive disparate treatment in specific contexts (e.g., Hawaiian and Samoan youth in juvenile court in Hawaii; MacDonald 2003). Generally, however, Asians are less involved in crime and justice than other groups. Asian are victimized at considerably lower rates than non-Asians (Harrell 2009), and they are sentenced similarly to white offenders (Johnson and Bestsinger 2009).
2. Although killings by police officers have declined dramatically over the past several decades, National Center for Health Statistics data suggest that “the rates for younger African Americans [under age 25] remain 4.5 times higher, and for older African Americans 1.7 times higher, than for other races and ages.” Mike Males, “Who Are Police Killing?” Center on Juvenile and Criminal Justice, August 26, 2014.
3. See also Robert J. Sampson, ‘‘Open Doors Don’t Invite Criminals: Is Increased Immigration behind the Drop in Crime?’’ New York Times, March 11, 2006.
4. “Programs: Common Justice,” Vera Institute of Justice, accessed May 1, 2014.
5. These results have been attributed to the aggregation of delinquent youth for the summer camp intervention; as a result, informal peer reinforcement of deviant behavior swamped any benefits of the formal program (McCord 1978, 2003).
6. Maggie Clark, “Should More States Require Racial Impact Statements for New Laws?” Stateline (blog), July 30, 2013.
7. Patricia Leigh Brown, “Opening Up, Students Transform a Vicious Circle,” New York Times, April 4, 2013.
8. “Tertiary Level,” PBIS, accessed December 31, 2014.
9. The US Department of Education (2002) released a supportive brief based on a program from Youth Crime Watch of America. However, an evaluation found implementation issues and was unable to find effects on school safety (Rich et al. 2008).

About the Authors

Akiva M. Liberman is a senior fellow in the Justice Policy Center at the Urban Institute, where he researches and evaluates crime and justice policy, with a focus on juvenile delinquency and juvenile justice. His current projects include the OJJDP’s JJ Reform and Reinvestment Demonstration Program, an evaluation of a juvenile reform demonstration effort centered on evidence-based practices; the Juvenile Second Chance Act Reentry Demonstration Projects, an evaluation of reentry programs for returning juvenile delinquents; and Early Access to Medicaid as a Reentry Strategy, which researches efforts to enroll inmates in Medicaid before release to the community.

Jocelyn Fontaine is a senior research associate in the Justice Policy Center at the Urban Institute, where her research portfolio is focused primarily on evaluating innovative community-based crime reduction and reentry initiatives targeted to vulnerable populations. In particular, she has directed several multisite reentry evaluations, including four supportive housing reentry projects.


The nonprofit Urban Institute is dedicated to elevating the debate on social and economic policy. For nearly five decades, Urban scholars have conducted research and offered evidence-based solutions that improve lives and strengthen communities across a rapidly urbanizing world. Their objective research helps expand opportunities for all, reduce hardship among the most vulnerable, and strengthen the effectiveness of the public sector.

Funding for this paper was provided by the Annie E. Casey Foundation, the Atlantic Philanthropies, the Open Society Foundations, and the Robert Wood Johnson Foundation. The views expressed are those of the authors and should not be attributed to these foundations or to the Urban Institute, its trustees, or its funders.
Funders do not determine research findings or influence scholars’ conclusions. Urban scholars and experts are independent and empowered to share their evidence-based views and recommendations shaped by research.

Copyright © February 2015. Urban Institute. Permission is granted for reproduction of this file, with attribution to the Urban Institute.