A decade ago the National Research Council and Institute of Medicine report Juvenile Crime, Juvenile Justice pointed out that there were “major disparities in the extent of involvement of minority youth, particularly black youth, compared with white youth in the juvenile justice system” (2001, p. 228). A number of assessments over the ensuing decade continued to document this overrepresentation of minority youth, especially African Americans, in the juvenile justice system (Engen, Steen, and Bridges, 2002; Bishop, 2005; Lauritsen, 2005; Bishop and Leiber, 2012). Such overrepresentation immediately raises at least two types of concerns. First, this circumstance raises questions of bias, fairness, and legitimacy regarding the functioning of the justice system. Second, it raises questions about the larger life-course trajectories of many youth in minority communities who may become marked by criminal records early in life.
In part for these reasons, the question of disproportionate minority involvement has been an explicit federal policy priority. Congress first gave attention to racial disparities in 1988 when it amended the Juvenile Justice and Delinquency Prevention Act (JJDPA) of 1974 (P.L. 93-415, 42 U.S.C. 5601 et seq.) to require states that received formula funds from the Office of Juvenile Justice and Delinquency Prevention (OJJDP) to ascertain the proportion of minority youth detained in secure detention facilities, secure correctional facilities, and lockups compared with the general population and, if the number of minority youth was disproportionate, to develop and implement plans to reduce the disproportionate representation (Section 223(a)(23)). In 1992, the JJDPA was amended. Disproportionate minority confinement was made a core requirement, and 25 percent of a state’s for-
mula funds could be withheld if states did not comply. In 2002, Congress again modified the disproportionate minority confinement requirement and mandated states to implement juvenile delinquency prevention efforts and system improvement efforts designed to reduce, without establishing or requiring numerical standards or quotas, the disproportionate number of juvenile members of minority groups who come into contact with the juvenile justice system (P.L. 107-273, Sec. 12209). Thus, the disproportionate minority contact (DMC) core requirement was broadened from “confinement” to “contact,” and states were required to implement strategies aimed at reducing disproportionality (Office of Juvenile Justice and Delinquency Prevention, 2009a). See Chapter 10 for a detailed description of OJJDP’s DMC activities.
Public and scholarly discussions about race/ethnic inequities and the role they play in the genesis of antisocial and criminal behavior and in shaping societal responses have a very long history (Hawkins and Kempf- Leonard, 2005, p. 3). Given the long-standing discussions over race/ethnicity in the United States more generally (National Research Council, 2001a), it is not surprising that discussions oriented around race/ethnicity1 and crime are among the most contentious of all (Sampson and Wilson, 1995; Kennedy, 2001; Peterson and Krivo, 2009).
Despite a research and policy focus on this matter for more than two decades, remarkably little progress has been made on reducing the disparities themselves or in reaching scholarly consensus on the root source of these disparities (National Research Council and Institute of Medicine, 2001). Volumes of data documenting disparities have been collected, but comparatively little progress has been made in addressing the problem (Kempf-Leonard, 2007; Piquero, 2008a; Bishop and Leiber, 2012). Thus, one assessment (Bell and Ridolfi, 2008, p. 15) observed with considerable irony:
There’s been a lot of motion but little movement in the last two decades. This inherited culture of the lowest common denominator in disparities
1 Throughout this chapter and throughout the report, we have chosen to link race/ethnicity together because their definitions are often overlapping. The Office of Management and Budget recognizes a minimum of five racial categories: white, black (or African American), American Indian or Alaskan Native, Asian, and Native Hawaiian or other Pacific Islander. It also recognizes at least two ethnicities: Hispanic or Latino and non-Hispanic or Latino. People who identify themselves as Hispanic, Latino, or Spanish can be of any race. These racial/ethnic categories were also included in the 2010 decennial census. But an analysis of census data had this to say about the racial groupings: “The race categories included in the census questionnaire generally reflect a social definition of race recognized in this country and are not an attempt to define race biologically, anthropologically, or genetically. In addition, it is recognized that the categories of the race question include race and national origin or sociocultural groups” (Humes, Jones, and Ramirez, 2011, p. 2).
reduction has resulted in a class of decision makers who could have significant impact on racial and ethnic disparities, but are unmotivated to do so. Instead, they make-up a multi-million dollar cottage industry whose primary activity is to restate the problem of disparities, in essence, endlessly adoring the question of what to do about DMC, but never reaching an answer.
Several reasons can be identified as a means of understanding the lack of movement on this issue, including, but not limited to, lack of motivation, lack of cross-system collaboration, inadequate resources, and the extreme difficulties of disentangling the many complex, multilevel and interrelated factors that contribute to this problem (Kempf-Leonard, 2007; Bell and Ridolfi, 2008; Bell et al., 2009; Nellis and Richardson, 2010; Parsons-Pollard, 2011). Some observers have suggested that lack of progress may be related to the deeper continuing problem of racial injustice in American society. The current period has been characterized as a time of “laissez-faire racism,” in which a “more covert, sophisticated, cultured-centered and subtle racist ideology, qualitatively less extreme and more socially permeable than Jim Crow racism,” is influencing American culture and politics (Bobo, 2011, p. 15). Whatever the reason, a discomfort in discussing race and racial inequities noted by the National Academies a decade ago does not appear to have changed significantly (National Research Council and Institute of Medicine, 2001, p. viii).
In effect, racial disproportionality (and race generally) has become the elephant in the room: most people concede that racial disparities pose a huge problem but are reluctant to candidly discuss their underlying causes and possible remedies.
Several thorough reviews of the literature on racial/ethnic disparities in the juvenile justice system have been published (National Research Council and Institute of Medicine, 2001; Pope et al., 2002; Leiber, 2003; Bishop, 2005; Hawkins and Kempf-Leonard, 2005; Piquero, 2008a; Bishop and Leiber, 2012). Instead of presenting another detailed review, this chapter briefly summarizes the problem, reviews the two main frameworks that have been used to understand and explain the problem (differential offending and differential selection), and then addresses a variety of factors that may contribute to both offending and the juvenile system’s response to it.
The conceptual and definitional challenges associated with racial/ethnic differences in general (National Research Council, 2001a) are evident in the context of juvenile and criminal justice. The terms oft-associated with DMC are “disproportionate representation” (or disparity) and “discrimination”
(or bias).2 On one hand, disproportionate (minority) representation, or disproportionality, occurs when a minority group (historically the research has centered on black youth) comprises a far greater percentage of persons in the juvenile justice system than their numbers in the general population would predict. According to Bishop (2005, pp. 24-25), disparity is used to denote between-group differences in outcomes, irrespective of their origins. (Disparity might stem from differences in offending, from laws or policies that differentially impact minority youth, or from racism in the juvenile justice system.) If defined in this neutral way, the committee regards “disproportionate representation” and “disparity” as interchangeable terms. On the other hand, discrimination refers to “situations in which evidence suggests that extralegal or illegitimate factors are the cause of disparate justice system outcomes” (National Research Council, 2001, pp. 230-231; for other variants, see Walker et al., 2000, pp. 14-18).
Definitions take one only so far, however, and there are important distinctions to consider. For example, disparity, particularly large and persistent disparity, is often interpreted as indicative of unfair or illegitimate processes at work. It is critical analytically to stress that not all statistical disproportion is an immediate indicator of bias or discrimination. However, particularly in the domain of juvenile justice and when matters of race/ethnicity are concerned, persistent disparity should be taken as a strong signal that some underlying problematic circumstance and process are operating, whether or not direct race bias is the cause. Taking this concept one step further, when there is evidence that racial disparities are systematic and intentional, then they can be considered racial inequities (Chapin Hall Center for Children, 2009).3
MINORITY YOUTH INVOLVEMENT IN THE JUVENILE JUSTICE SYSTEM
Researchers typically draw on three possible sources of data to gauge the extent of minority4 youth involvement in crime and delinquency: official
2 The term “disproportionate minority contact” is used to describe the disproportionate number of minority youth at various stages of processing in the juvenile justice system (Office of Juvenile Justice and Delinquency Prevention, 2009a, 2009b). Throughout the report, we use “racial disparities” to refer to racial/ethnic disparities more generally and use DMC when it is common usage, for example, associated with OJJDP’s core requirement or in a program initiative by the government or other organization, such as the MacArthur Foundation’s Model for Change DMC Action Network.
3 A very helpful graphic presentation of the relationship of disproportionality, disparities, and factors leading to disparity, can be found in Chapin Hall Center for Children (2009, p. 32).
4 The term “minority” is not being used as a proxy for black or African American but is used when the term applies to minorities more broadly. The term “black or African American” is used when the statement applies specifically to that racial group.
statistics on arrests, criminal victimization surveys of the population, and self-report surveys and questionnaires administered to youth. Each potential source of data has limitations.
We begin with a consideration of official statistics on juvenile arrests based on the Federal Bureau of Investigation’s Uniform Crime Reports (UCR). (See Chapter 3 for a discussion of juvenile crime arrest data.) Table 8-1 reports official arrest results for people under age 18 by race for the year 2009, the most recent period for which these data were available to the committee. These results show disproportionate black arrests in most categories of offenses. The overrepresentation of black youth is greatest for violent crimes, particularly for homicide and manslaughter and for robbery. For homicide and manslaughter, black youth represent 58 percent of those arrested in 2009, although only 16 percent of youth under age 18 are in this age category. Similarly, blacks constitute 67 percent of those arrested for robbery.
Disproportionate arrests remain the pattern for black youth in most of the property crime offenses, although the extent of overrepresentation relative to their share of the total youth population is smaller. Thus, black youth constituted 37 percent of burglary arrests and 43 percent of motor vehicle thefts though only 16 percent of all youth. These percentages are half the extent of overrepresentation seen in some of the violent crime data.
Two further points are worthy of note. The one category in which black youth are underrepresented relative to their share of all youth is that of alcohol violations (6 percent of arrests). This is also the one type of offense for which white youth tend to be overrepresented. In addition, the degree of black overrepresentation is at its lowest in the category of drug abuse violations, in which blacks make up roughly 26 percent of youth arrests.
These data consistently show that there are important differences by race in rates of arrest—especially across offense type, with black youth arrested for violent index crimes at much higher rates than whites (Bishop, 2005; Bales and Piquero, 2012). These disparities tend to be smaller (but tend to persist) for property crime rates, with white rates being higher, on average, for other offenses, such as vandalism and offenses involving alcohol. The UCR does not produce data for offending rates across ethnic groups so, as a result, there is no official national arrest information relating to Hispanics—thus similar comparisons cannot be made between Hispanics and other racial/ethnic groups. Turning to the postarrest official data, blacks have higher rates than whites for ensuing juvenile and criminal justice decision stages, such as being referred to court, detained, formally
|White||Black||American Indian or Alaskan Native||Asian or Pacific Islander||Total|
|Population Under 18||57,563,627 (77.2)||12,045,688 (16.2)||1,081,363 (1.5)||3,857,537 (5.2)||74,548,218 (100)|
|Total Arrests||993,428 (65.9)||472,929 (31.3)||18,766 (1.2)||23,427 (1.6)||1,508,550 (100)|
|Murder and nonnegligent manslaughter||380 (40.4)||546 (58.0)||8 (0.9)||7 (0.7)||941|
|Forcible rape||1,501 (63.4)||818 (34.5)||19 (0.8)||30 (1.3)||2,368|
|Robbery||7,854 (31.1)||16,968 (67.3)||112 (0.4)||292 (1.2)||25,226|
|Aggravated assault||21,790 (55.4)||16,694 (42.4)||394 (1.0)||463 (1.2)||39,341|
|Burglary||36,073 (60.9)||22,082 (37.3)||511 (0.9)||571 (1.0)||59,237|
|Larceny/theft||164,701 (65.0)||80,670 (31.8)||3,148 (1.2)||4,948 (2.0)||253,467|
|Motor vehicle theft||8,454 (54.0)||6,765 (43.2)||234 (1.5)||213 (1.4)||15,664|
|Arson||3,222 (76.7)||865 (20.6)||56 (1.3)||60 (1.4)||4,203|
|Alcohol violations||98,113 (89.6)||6,946 (6.3)||3,105 (2.8)||1,343 (1.2)||109,507|
|Drug abuse violations||97,232 (72.4)||34,295 (25.6)||1,212 (0.9)||1,468 (1.1)||134,207|
|Weapons offenses||16,190 (60.7)||9,938 (37.3)||210 (0.8)||328 (1.2)||26,666|
NOTE: Percentages in parentheses.
SOURCE: Federal Bureau of Investigation (2010).
charged, adjudicated delinquent, and placed out of the home (Bishop, 2005).
A second source of data is the Relative Rate Index (RRI), which was developed by OJJDP in order to measure disparity at each decision point in the system: arrest, court referral, diversion, detention, petitions/charge filing, transfer to adult court, delinquency findings, probation, and secure confinement.5 Table 8-2 breaks down these processing stages by race. RRI data can be easily calculated on the basis of readily available data maintained by some states. Feyerherm (2011) recently examined RRI data from OJJDP’s DMC website that included information from 1,043 jurisdictions (47 states and 996 substate jurisdictions, mainly counties). Based on these data, one is able to ascertain patterns among Hispanic youth and compare them to black and white youth. For example, RRI data suggest that Hispanic youth experience greater contact with the juvenile justice system than do white youth and that the extent of these differences (disparities) is not as great as those experienced in general by black youth (Feyerherm, 2011, p. 46).
These official records generate useful information, but they also suffer from some notable limitations (see Chapter 3). For example, official data and associated record-keeping systems are complex and not wholly integrated or infallible. For example, processing data may not be integrated with data from other child-serving systems with which the youth may have had contact or from which he or she may have been referred. Moreover, official records are contingent on the justice system responding to some action or call for service. Thus, official records do not include a large amount of criminal behavior that goes undetected and does not come to the attention of the formal justice system. Also as indicated above, the UCR data collection system treats race/ethnicity as two distinct characteristics and does not provide a means for identifying non-Hispanic and Hispanic members of different racial groups (Feyerherm, 2011, p. 46). This not only leads to difficulty in comparing arrest trends but also obfuscates the RRI because “arrest numbers cannot easily be traced into the juvenile justice system to follow the cumulative impacts of arrest, referral, detention, etc.” (Feyerherm, 2011, p. 47).
An additional problem with the RRI calculations is that they do not come with any sort of statistical significance measure; thus, there is no way to measure whether an RRI of 1.0 is statistically significant—much
5 Specifically, the RRI consists of three components: (1) a system map describing the major contact points or stages at which a juvenile may have additional contact or penetration into the justice system, (2) a method for computing rates of activity (by race/ethnicity) at each of the stages, and (3) a method to compare the rates of contact for different demographic groups at each of those stages (Feyerherm, Snyder, and Villarruel, 2009; Feyerherm, 2011, p. 37).
|White||Black||American Indian or Alaskan Native||Asian or Pacific Islander||Total|
|Population Ages 10-17||25,251,300 (77.4)||5,437,700 (16.5)||455,700 (1.4)||1,549,000 (4.7)||32,963,700 (100)|
|Arrested||1,246,900 (66.7)||574,600 (30.8)||21,900 (1.2)||26,200 (1.3)||1,868,600|
|Arrest rate per 100,000 in population||4,886||10,567||4,806||1,637|
|Relative risk black:white = 2.1:1|
|Referred to Courts||1,043,600 (63.1)||563,500 (34.1)||23,500 (1.4)||22,700 (1.4)||1,653,300|
|Referral rate per 100,000 in population||4,089||10,363||5,157||1,465|
|Relative risk black:white = 2.5:1|
|Detained||194,100 (55.8)||143,300 (41.2)||5,300 (1.5)||5,100 (1.5)||347,800|
|Detention rate per 100,000 in population||761||2,635||1,163||329|
|Relative risk black:white = 3.5:1|
|Formally Charged||554,800 (60.0)||342,000 (37.0)||14,400 (1.6)||13,200 (1.4)||924,400|
|Petition rate per 100,000 in population||2,174||6,289||3,160||852|
|Relative risk black:white = 2.9:1|
|Adjudicated Delinquent||350,900 (62.2)||194,900 (34.6)||10,100 (1.8)||8,000 (1.4)||563,900|
|Adjudication rate per 100,000 in population||1,375||3,584||2,216||516|
|Relative risk black:white = 2.6:1|
|Placed Out of Home||91,000 (57.7)||61,500 (39.0)||3,200 (2.0)||2,000 (1.3)||157,700|
|Placement rate per 100,000 in population||357||1,131||702||129|
|Relative risk black:white = 3.2:1|
NOTE: Percentages in parentheses.
SOURCE: Puzzanchera and Adams (2011a).
less whether an RRI of 1.38 is significantly different from an RRI of 2.53. As a result, these sorts of official statistics provide limited leverage on the larger question of disproportionate minority youth contact with the juvenile justice system.
Self-Report and Victimization Data
Other sources of racial/ethnic disparities emerge from data on offending patterns. Lauritsen’s (2005) review of this line of work was based on victim reports from the National Crime Victimization Survey (NCVS) and a series of self-report surveys that gathered individual-level reports of offending. The analysis showed that “the most commonly occurring crimes exhibited few group differences, while more rare and serious crimes of violence showed generally higher levels of black and Latino involvement” (Lauritsen, 2005, p. 99). Thus, the salient message from Lauritsen’s review is that data on youth violence are comparable across reporting sources because the same general patterns have emerged for the most serious but least common offenses (Lauritsen, 2005, p. 100). At the same time, an important difference emerged in relation to drug abuse violations. Lauritsen (2005, p. 96) reports that black youth are disproportionately involved in such offenses as measured via official records, whereas self-report data indicate that white youth report higher levels of drug abuse violations.
Similar to the Lauritsen study but using both UCR and self-report data sets, Piquero and Brame (2008) found little evidence of racial/ethnic differences in either self-reported offending (either in the frequency of offending or in the variety of offending) or officially based arrests leading to a court referral in the year preceding study enrollment.
Both victim and self-report data suffer from problems similar to those that plague official records. For example, the race/ethnicity of the offender may not be known in victim and self-report data. Furthermore, victim survey data are limited to the main race categories of black, white, and other. Self-report data suffer from both over- and underreporting, and these tendencies may vary across racial/ethnic groups. They are often collected from high school or general population samples, a practice that tends to limit reports of serious violence. Finally, there have been few comparisons of self-reports across racial/ethnic groups (Huizinga et al., 2007; Piquero and Brame, 2008), few data collection efforts focused on Hispanics (Maldonado-Molina et al., 2009), and even fewer studies examining the relationship of immigration status to offending (Lee, Martinez, and Rosenfeld, 2001; Nielsen, Lee, and Martinez, 2005; Bersani, 2012).
Research on the factors that might affect DMC at the police contact and court referral levels also has employed both official and self-report data with a common set of delinquency measures across data sources
(on violence, property, weapons, and drug offenses). Huizinga and colleagues (2007) used data from the three delinquency studies in Pittsburgh, Pennsylvania; Rochester, New York; and Seattle, Washington, to examine DMC and the factors that might affect it at the police contact and court referral levels.
First, in all three cities, African American youth had the highest rate of contact/referral, and it was significantly greater than for white youth. Hispanics in Rochester had a significantly higher rate than whites; in Seattle, Asian American youth had a slightly higher rate of contact/referral compared with whites. These results were replicated in overall crime figures. Second, when the researchers examined race/ethnic differences in self-reported offending, they found that minority youth did exhibit higher self-reported offending than whites, but the differences were not so pronounced as they were with the official record data. In general, minority– white differences in the official record comparisons were roughly double what they were for the self-reported offending estimates. Thus, differences in self-reported offending were not able to completely eliminate the effects of race/ethnicity on official criminal records (Huizinga et al., 2007, p. 32). Third, Huizinga and colleagues examined the effect of race/ethnicity on contact/referral in the juvenile justice system after controlling for self-reported offending. Results from this analysis indicated that, across virtually all comparisons, although controlling for self-reported offending was itself significantly associated with official contact, it did not eliminate (nor very much reduce) any direct effect for race/ethnicity.
In sum, these results show that self-reported offending does not explain the differential rates of juvenile justice system contact by race/ethnicity.6 When a risk factor composite (e.g., socioeconomic status, family structure, academic performance) was added to assess whether inclusion of this additional measure altered the significant race/ethnicity effect on official record representation, once again, with one exception (Pittsburgh), the results held: although both self-reported offending and the risk factor composite were significantly associated with disproportionate involvement as measured by official records, controlling for the risk factor composite did not affect the still-significant effect for race/ethnicity on official records (Huizinga et al. (2007).
Similarly, Bersani (2012) used self-report data from the National Longitudinal Survey of Youth 1997 (NLSY97) and official crime reports to
6 Only a few other studies have examined self-reported delinquency and subsequent juvenile justice processing (Huizinga and Elliott, 1987, in the National Youth Survey; Fergusson, Horwood, and Swain-Campbell, 2003, in Australia; and Piquero and Brame, 2008, in the Research on Pathways to Desistance study). Although these studies contain longitudinal data, the methodological approaches thus far have not made explicit use of the longitudinal data in order to examine the racial disparity question in a developmental manner.
conduct trajectory analyses that examined immigrant offending histories from early adolescence to young adulthood. Her findings showed that first-generation immigrants had lower rates of criminal involvement compared to native-born persons. In fact, violence and drug crimes were virtually nonexistent among first-generation immigrants while second-generation immigrants evinced offending patterns similar to native-born persons. These findings are consistent with those of other studies using other data sources that report a crime-suppression effect of immigrant concentration on crime rates even in areas marked by concentrated disadvantage (Lee et al., 2001; Nielsen et al., 2005; Sampson et al., 2005).
Reviews of DMC Research
A number of assessments over the years make it clear that minority youth are disproportionately represented in the system. Several recent careful reviews, in particular, have found that “race matters” beyond the characteristics of an offense. One recent major assessment that took stock of 72 quantitative studies of DMC had three major results (Cohen et al., 2011). First, it found that the vast majority of studies (82 percent) found some race effect that disadvantaged minority youth relative to white youth. Second, the evidence for race effects was greatest at earlier stages of the process, particularly at the stages of arrest, referral to court, and placement in secure detention. Third, although black youth are most likely to be disadvantaged, this is not uniformly the case and similar patterns tend to emerge for Hispanic youth as well.
Their review covered studies conducted in 2002-2010 on the official processing of minority youth at nine different decision points in the juvenile justice system (arrest, court referral, delinquency findings, detention, diversion, petition/charge filings, probation, secure confinement, and transfer to adult court). (Note: some decision points have been more intensively studied than others; i.e., arrest has been less thoroughly studied than the secure confinement decision and white-black disparities have been studied more often than others.) The analysis shows that the majority of reviewed studies indicated some race effects in the processing of minority youth, with the majority of those studies reporting mixed results (for some minority youth or at some processing points but not others). Black males were more likely to receive harsh treatment than females or whites, and minority youth, on average, were more likely to receive harsh treatment for certain but not all offenses. At the same time, the analysis also indicates a lower race effect in formal court processing, adjudication, and postadjudication.
In nearly all juvenile justice systems youth of color also remain in the system longer than white youth. From 2002 to 2004, although black youth accounted for approximately 17 percent of the youth population,
they represented 28 percent of juvenile arrests, 37 percent of the detained population, 38 percent of those in secure placement, and 58 percent of youth committed to state adult prison (National Council on Crime and Delinquency, 2007, p. 3; The Sentencing Project, 2010, p. 1). Furthermore, 2008 case processing data for delinquency offenses from the Office of Juvenile Justice and Delinquency Prevention’s National Disproportionate Minority Contact Databook (Puzzanchera and Adams, 2011a) indicate that black youth have much higher rates of arrests than their white counterparts, as well as higher rates of being detained, having petitions filed, and being placed, but lower rates of being diverted and referred to probation (see Table 8-2).7 The pattern of differences for American Indian and Asian American youth compared with whites is not so straightforward. Both American Indian and Asian American youth have a higher rate of disproportionate contact at the case referral stage and the detention stage than whites. Asian youth have higher rates of processing than black youth in the referral, petition, and adjudication stages as well higher rates of transfer to adult court. Both groups are diverted at a lower rate than either white or black youth (see Table 8-2).
In sum, with few exceptions, data consistently show that youth of color have been overrepresented at every stage of the juvenile justice system, that race/ethnicity are associated with court outcomes, and that racial/ethnic differences increase and become more pronounced with further penetration into the system through the various decision points (Rodriguez, 2010).8 When one includes the compound and cumulative character of racial/ethnic involvement throughout (and through progressive stages of) the juvenile justice system, it is no surprise that the issue has been subject to much discussion and, in turn, received persistent attention.
The remaining important question is why minorities are overrepre-sented in the juvenile and criminal justice systems. We begin with the two main perspectives (differential offending and differential selection by the justice system), which have often been viewed—incorrectly in the committee’s view—as competing, rather than complementary, explanations for the disparity (Piquero, 2008a; Bishop and Leiber, 2012). We then expand our
7 In a different analysis of 2005 data from the National Juvenile Court Data Archive that include ethnicity data for about two-thirds of the nation’s Latino population, Latino youth are 4 percent more likely than white youth to be petitioned; 16 percent more likely than white youth to be adjudicated delinquent; 28 percent more likely than white youth to be detained; 41 percent more likely than white youth to receive out-of-home placement; 43 percent more likely to be admitted to adult prison (Arya et al., 2009).
8 The Rodriguez study appears to be at odds with the Cohen et al. (2011) review of 72 studies cited earlier. Although they are addressing similar issues, the Rodriguez study and others like it focus on a single site and study youth through various juvenile justice stages from beginning to end.
discussion to other explanations that either do not fit neatly into either of those two perspectives or may have relevance for both.
EXPLAINING RACIAL DISPARITIES
Accounts of DMC typically fall into one of two broad camps. Some scholars emphasize differential offending as the root source of disproportionate minority involvement in the juvenile justice system and of the system’s differential response. This approach points, in effect, to real, underlying differences between white and minority youth in the actual extent of engaging in (or the severity of) law-breaking behaviors. Other researchers point to differential selection by the justice system (by the police in enforcement and by prosecutors, intake officers, judges, and other justice system officials thereafter) as the primary source of racial disparities. As discussed below, findings of differential selection have sometimes been interpreted as demonstrating systematic and often institutional bias, but differential enforcement and justice system processing are not necessarily or always attributable to bias or discrimination.
As referenced by Lauritsen (2005), there are more similarities than differences among youth across races with respect to offending patterns in self-reported data, with the exception of participation in serious violence. As noted, minority youth (especially black youth)9 tend to offend more with respect to serious person crimes, and they have also been found to persist in crime into early adulthood at a higher rate than whites (Elliott, 1994; Haynie, Weiss, and Piquero, 2008). This finding is important because research shows that serious violence is more likely to be reported to the police, more likely to result in the offender’s apprehension, and more likely to trigger severe juvenile and criminal justice sanctions (Piquero, 2008a, p. 64). And although research shows that much of the minority overrepresentation in secure confinement and prisons can be attributed to differences among racial groups in arrests for crimes that are most likely to lead to confinement, this same research also shows that it is unlikely that behavioral differences account for all minority overrepresentation (Blumstein, 1982, 1993; Crutchfield, Bridges, and Pitchford, 1994; Sorensen, Hope, and Stemen, 2003).
9 As previously noted, most disparity research is limited to comparisons between whites and blacks, largely because of the lack of data for Hispanics, Asian Americans, and American Indians in both self-reported and especially official records. The intersection of race and gender is even less frequently studied despite the rapid growth of black girls in the juvenile justice system (Sherman, 2012, p. 1617).
Although space precludes a detailed investigation and review of theoretical accounts of racial/ethnic differences in (serious) offending (Hawkins and Kempf-Leonard, 2005), these differences have been attributed to several risk factors that span the individual, familial, and neighborhood levels. (See Chapter 6 for an explanation of risk factors and risk markers.)10 In general, these can be considered as “contexts for risk” (National Research Council and Institute of Medicine, 2001) so as to not be confused with another set of system-based factors that could also be implicated in disproportionality.
Minorities, especially blacks are more likely than whites to live in economically disadvantaged communities (Sampson and Wilson, 1995). Such communities have distressed education, child welfare, and public health systems (Sharkey and Sampson, 2010; Ryan, Chiu, and Williams, 2011). They also tend to have many social structural conditions that contribute to delinquency, crime, and violence, such as poverty, disorder, residential segregation, and neighborhood disadvantage (Wilson, 1987). These effects tend to compound and accumulate in mainly minority communities so that poor, inner-city residents find it to difficult to move out of this urban core and escape to more affluent neighborhoods that come with improved opportunities for education and employment.11 The ramifications of these minority-centered contexts of risk include poor health care (and subsequent health)12 and substance abuse problems and disparities (Piquero, Moffitt, and Lawton, 2005), low-performing schools, absence of recreation programs or other organized activities for youth (Bishop and Leiber, 2012), disadvantaged familial and community-level socialization and controls (Sampson, Morenoff, and Raudenbush, 2005), and greater exposure to violence and other negative experiences (Crouch et al., 2000). The totality of these risk factors is such that minority youth are born into and raised in severely compromised familial, community, and educational environments that set the stage for a range of adverse behaviors and outcomes, including problems in school, relationships, and engaging in prosocial behavior.
Investigating this phenomenon, Fite and colleagues (2009) noted that differences observed in offending across race/ethnicity (and in subsequent
10 In this chapter, we are using “risk factors” instead of “risk markers” because of its usage by the writers we are citing.
11 Massey and Denton (1993) argue that racial segregation is the principal organizational feature of American society that is responsible for the creation of the urban underclass.
12 For example, based on available Canadian data, youth with fetal alcohol spectrum dis order, an umbrella term that covers the range of outcomes associated with all levels of prenatal alcohol exposure, are 19 times more likely to be incarcerated than are youth without the disorder in a given year (Popova et al., 2011). A similar study has not been done on minority youth in the United States, but, given the high rates of heavy alcohol consumption among African Americans and Native Americans (Galvan and Caetano, 2003), one can infer that minority youth would be at great risk for the disorder.
juvenile and criminal justice experience) could be traced to the fact that minority (especially black) youth display and experience more risk factors for offending and risk, such as poor health care and compromised education systems. They examined the effect of exposure to early risk factors on arrest rates and found that the risk factors themselves were predictive of a juvenile arrest. In fact, the risk factors accounted for 60 percent of the total effect between race and general arrest (Fite, Wynn, and Pardini, 2009, p. 921). Exposure to concentrated disadvantage can also have detrimental and long-lasting consequences even after a youth leaves a severely disadvantaged neighborhood (Sampson, Sharkey, and Raudenbush, 2008).
The differential selection hypothesis asserts that a combination of differential enforcement (differing police presence, patrolling, and profiling in minority and nonminority neighborhoods) and differential processing by the juvenile justice system (differing dispositions and placements in the courts and correctional systems) leads to more minority youth being arrested, convicted, and subsequently confined than white youth (Piquero, 2008a, p. 65). This hypothesis may be especially pertinent to victimless crimes, such as drug use and sales and public order crimes, in which more discretion is available to formal social control agents, especially police, and virtually all interactions (especially among police and juveniles) are made out of the public eye (Piquero, 2008a, p. 65). Thus, the differential selection hypothesis would anticipate that minority youth emerge in official records at a disproportionate rate because of differential police, court, and correctional decisions.
To illustrate the differential selection hypothesis at the police level, consider a policy decision to differentially assign police to particular neighborhoods with higher reports of crime, especially serious and violent crimes. Because such neighborhoods often tend to be overrepresented in impoverished, minority locations, this places minority offenders at an increased risk of detection and potential arrest as a result of their encounters with the police. Increased police presence also creates greater opportunities for discretion to be exercised in street encounters and, as a result, for arrest decisions to vary across race/ethnicity.
As this example suggests, conventional enforcement practices or patterns of judicial administration can lead to racial/ethnic disparities even if they are not intended. Thus, it would be a mistake to regard differential selection by the juvenile justice system as equivalent to proof of bias. Bias or even intentional discrimination may well be operating, but disparities can also arise from otherwise legitimate justice system processes.
Black youth who live in segregated communities tend to have more contacts with police than white youth (Brunson and Weitzer, 2009; Crutchfield, Bridges, and Pitchford, 2009). They are more likely to go to schools with police presence, more likely to be suspended or expelled from school (Skiba et al., 2002; Fabelo et al., 2011; Skiba et al., 2011), and more likely to have contact with officers as a result of disciplinary action. Children engaging in the same behavior in schools or in neighborhoods without a police presence or who live where there are occasional patrols will have less contact (Crutchfield, Bridges, and Pitchford, 2009).
Many studies focus on institutional policy and practice around selective enforcement. Some focus on the role of drugs in minority communities (e.g., open-air drug markets, the passage of certain drug laws and punishment) as well as the controversial subject of racial profiling. With respect to the race– drugs relationship, Tonry (1995), for example, claimed that the passage of the crack cocaine sentencing laws was virtually known to differentially target minority—especially black youth—in urban communities because the sale and use patterns of crack cocaine (i.e., inner-city, open-air markets, violence-ridden streets) are largely race based. Thus, because the passage of the crack cocaine sentencing laws were made, in part, as a response to the violence that was permeating many inner cities in the mid- to late 1980s, and because the police had to selectively target certain communities and drug markets, an obvious by-product was that minority youth would be exceedingly more likely to fall under formal social control. Analyses of racial disparity in drug arrests in Seattle by Beckett and colleagues (2005, p. 419) centered on “the racialization of imagery surrounding drugs in general and crack cocaine in particular” as the driving force shaping police perceptions and practices, as well as disparities in drug possession in Seattle.
Turning to the potential effect of racial profiling on racial disparities, there is a large body of research that has examined a wide range of data on traffic stops, driving patterns, and public perceptions associated with racial/ethnic profiling by the police (Rice and White, 2010). Because space constraints preclude a detailed overview of this body of work, a few such studies are highlighted.
Fagan and colleagues have produced a comprehensive body of research on “order-maintenance policing” and its effect on racial profiling in New York City. In one recent report, Fagan and colleagues (2010) examined data on police street stops between 1998 and 2006 and focused on the rates of stops in New York City neighborhoods with the highest concentration of black residents. Their analyses showed that street stops were disproportionately concentrated in the city’s poorest areas, that the most recent increases in stops were concentrated in predominantly minority neighborhoods, that
minority residents were more likely to be disproportionately subjected to law enforcement contact based on the neighborhoods in which they lived rather than the crime problems in those areas, and that black citizens not only had an elevated risk of police contact compared with non-Hispanic whites and Hispanics, but also that the standards used to justify stops in their neighborhoods appeared to be lower than those in neighborhoods with larger white populations (Fagan et al., 2010, p. 311).
In short, there is a sizable literature indicating that minority youth are more likely than white youth to be stopped, arrested, and subsequently referred to court by police (Bishop and Leiber, 2012, p. 461). Although isolation of a single factor for this is beyond the reach of any study, it is fair to conclude that a range of factors—including differential deployment and police surveillance (Smith, 1986; Krivo and Peterson, 1996; Warren et al., 2006);13 differential police suspicion (Alpert, MacDonald, and Dunham, 2005) and use of cognitive shortcuts and unconscious stereotypes in minority neighborhoods and on minority youth (Kennedy, 1997; Smith and Alpert, 2002); and juvenile demeanor (“Black and Hispanic youth tend to be [or are perceived to be] less cooperative, more gang-involved, and more threatening”) (Bishop and Leiber, 2012, p. 461)—are implicated in differential policing handling of minority juvenile offenders (Piliavin and Briar, 1964).14
Race, police contact, and minority youth’s behavior are also intertwined in complicated ways. When contacts with police occur early, the likelihood that a black youth will have future contacts with police is increased. For example, early contacts with police (by eighth grade) have been shown to increase the risk for arrest by high school by fivefold, even when accounting for all other environmental domains, including self-report criminal behavior (Crutchfield, Bridges, and Pitchford, 2009). These contacts with police also shape a youth’s perception of and compliance with legal authorities (Fagan and Piquero, 2007). Lee and colleagues (2010) found that the stronger the sense of racial identification as a minority group, the higher the perceived discrimination by police. Race also affected perceptions of global police prejudice, procedural justice, and police legitimacy. (For a fuller discussion of youth’s perceptions, see Chapter 7.) Youth who considered police contacts overly aggressive and confrontational tended to avoid police at all costs and were likely to perceive themselves as having been badly treated (Weitzer and Brunson, 2009). As a consequence, black youth have very
13 For example, significant racial disparities in the implementation of marijuana law enforcement were observed in New York City during 2004-2008 (Geller and Fagan, 2010).
14 At the same time, however, some additional evidence from traffic stops exploring citizens’ demeanor and race shows that black and Hispanic motorists are not more likely than whites to be arrested during traffic stops when other legal and extra legal factors are considered (Engel, Klahm, and Tillyer, 2010).
troubled relationships with police compared with white youth (Brunson and Weitzer, 2009).
Differential Processing by the Justice System
A voluminous literature examines the decisions at each stage of the juvenile justice system in order to examine how minority and white youth are treated and the extent to which they receive similar or different outcomes (Bishop, 2005).15 In the earlier cited review, Cohen and colleagues (2011) concluded that some race effects exist in the processing of some minority youth, in some locations, at some time periods, and for certain offenses; that minority youth are more likely to receive harsh treatment for certain but not all offenses; and that racial disparities can be documented for certain stages but not others. Here, a few studies are highlighted in order to show how this research has been conducted.
In a classic piece, Bridges and Steen (1998) examined the tone and value of word choices that were used to describe black and white juvenile offenders by probation officers. Officers attributed offenses by black juveniles more to negative attitudinal and personality traits. They attributed traits and offenses by whites more to the social environment. These authors also found that these differences contributed significantly both to the officers’ differing assessments of the risk of reoffending and to their recommendations about sentencing, even after controlling for case and offender characteristics (Piquero, 2008a, p. 66).
In a related study in Washington, Bechtold and colleagues (2011, p. 5) examined juvenile probation at three sites with a mixture of black, Hispanic, and white youth by exploring whether judges set different conditions of probation and ordered different services for youth of different racial/ethnic groups and whether probation officers treated them differently according to their race/ethnicity. Results were mixed, but in general the authors reported no consistent pattern of discrimination. Specifically, all youth regardless of race/ethnicity received very similar conditions of probation, were cited for similar violations at similar rates, and received similar responses.
Graham and Lowery (2004) conducted two experiments in Los Angeles involving police officers and juvenile probation officers in order to examine unconscious racial stereotypes of decision makers in the juvenile justice system. Specifically, the sample was subliminally exposed to words related to the category black—such as ghetto, homeboy, and dreadlocks—or to
15 It is important to point out that, although there is a body of research on the influence of race in police juvenile contacts, research on police arrest decisions is limited in comparison to what is known about the other stages of juvenile justice processing (Bishop and Leiber, 2012).
words neutral with respect to race. At the same time, the officers read two scenarios about a hypothetical adolescent who allegedly committed either a property (shoplifting) or violent (assault) crime. In addition to answering questions about conscious attitudes about race, the officers rated the offender on a number of individual characteristics and made judgments about culpability, expected recidivism, and deserved punishment. Findings showed that, compared with officers in the neutral condition, officers in the racial prime condition reported more negative trait ratings, greater culpability, and more expected recidivism and also endorsed harsher punishment. Significantly, the race primes had the same effect regardless of the officers’ own race and consciously held attitudes about blacks. The findings held even among those who reported that they were tolerant and not biased toward nonwhites (Piquero, 2008a, pp. 66-67).
Using data from Black Hawk County, Iowa, Leiber (2009) examined the factors associated with pre- and postadjudication secure detention and subsequent decision making. His analysis showed that legal factors were strongly related in the expected direction to each type of secure detention and subsequent decisions but that race effects were also apparent for some juvenile justice decisions but not others. Moreover, his findings also revealed that race effects did not always result in more severe sanctions for minority youth.
In a sample of more than 23,000 Arizona youthful offenders, Rodriguez (2010) examined the cumulative effect of race/ethnicity via detention on various juvenile court outcomes. Her results showed that black, Hispanic, and American Indian youth were treated more severely in juvenile court outcomes than white juvenile offenders, both at the front-end court processes (diversion and detention) as well as the back-end processes and outcomes (out-of-home placement). The findings revealed that detention produces indirect racial/ethnic effects in subsequent stages of processing and that “youth who were detained pre-adjudication were more likely to have petitions filed, less likely to have petitions dismissed, and more likely to be removed from the home at disposition” (Rodriguez, 2010, pp. 391-392).
In examining differential involvement, it is also important to take account of the structural context of juvenile court administration in understanding racial/ethnic disparities in judicial processing. Some evidence suggests that urban courts tend to be more formal and bureaucratic and have greater access to detention facilities than rural courts and that these characteristics are associated with harsher sentences. Because they disproportionately reside in urban counties, black youth are at increased risk of being processed, detained, and punished than white youth in rural localities who have committed similar offenses. Thus, it is also possible that location, race, and punitiveness are intertwined (Feld, 1991; Sampson and Laub, 1993; Bray, Sample, and Kempf-Leonard, 2005; Rodriguez, 2010).
Other Explanations for Differential Involvement
As reflected in the discussion thus far, it remains difficult to apportion documented disparities to either differential offending or differential selection and to ascribe the role that bias plays. The research to date has not attempted to do that, given that the impact of either perspective is confounded by the effects of underlying social and cultural factors. In addition, researchers have identified other factors that may contribute to differential juvenile justice outcomes that may not fit neatly into either of those perspectives or may be distinct from both. These include jurisdictional differences in the treatment of youth, such as case processing (Kempf- Leonard, 2007); organizational issues throughout the juvenile justice system, including resources and agency roles (Bishop, Leiber, and Johnson, 2010); “justice-by-geography,” that is, local institutional culture (Feld, 1991; Bray, Sample, and Kempf-Leonard, 2005); legislative decisions (Tonry, 1995); and administrative policies, such as zero-tolerance policies in schools that propel minorities into the system (Verdugo, 2002; Hirschfield, 2008). Several of these additional factors are discussed below.
Code of the Street
Anderson’s (1999) code-of-the-street thesis, which contends that minority youth—especially black youth—form and espouse an attitude that is organized around informal rules governing street behavior and response to personal affronts. These attitudes form mainly in response to the economic disadvantage, social isolation, and racial discrimination encountered by black youth in the most disadvantaged urban communities. Adoption of these codes, which center on the issue of respect (i.e., being treated right or granted the deference one deserves), is deemed a virtual necessity for respect and survival in the most disadvantaged, distressed, and impoverished minority—especially black—communities. But hanging out with people who adopt the code of the street or being in places where such people are known to congregate may increase the risk of greater involvement with the police (Crutchfield, Bridges, and Pitchford, 2009) as well as actual crime. A study by Stewart and Simons (2010) using data on 800 black adolescents ages 10-15 in Georgia and Iowa showed that a youth’s expressed street code attitudes significantly predicted violence two years later, so that youth who internalized and lived by the code were the most likely to be involved in subsequent violence. It is important to note as well that scholars have identified a similar respect-based code of the streets orientation among Hispanics (Bourgois, 2003).
Juvenile Justice System Feeders
A less investigated explanation for racial/ethnic differences in juvenile justice involvement concerns feeder systems and agencies that funnel youth into the juvenile justice system, including the school disciplinary system (Fabelo et al., 2011), the mental health system (Feld, 1998b; Teplin et al., 2002), and the child welfare system (Bowser and Jones, 2004; Herz et al., 2010). See Chapter 3 for a discussion of how these systems act as feeders for the juvenile justice system.
For example, a longitudinal study of almost a million adolescents in Texas schools found that 1 in 5 African American students had involvement with the juvenile justice system compared to 1 in 6 Hispanic students and 1 in 10 white students. The study controlled for 83 variables and found that African American students were more likely than students of other races to be disciplined and to receive a harsher punishment (Fabelo et al., 2011, p. 40). For example, African American youth were almost twice as likely as Hispanic students and three times as likely as white students to be placed on out-of-school suspension for the first violation (Fabelo et al., 2011, p. 42). This suggests that discretionary action by school officials is contributing to the higher rate of involvement with the juvenile justice system, with the Texas data showing that multiple discretionary disciplinary actions were more common among African American and Hispanic students than white students (Fabelo et al., 2011).
As with youth referred to the justice system by schools, race is an important predictor of whether youth cross over from the child welfare system to the juvenile justice system (Herz and Ryan, 2008a; Herz, Ryan, and Bilchik, 2010). This is not surprising, since African American youth are overrepresented in foster care at a rate of more than twice their proportion in the U.S. child population. African American youth in the child welfare system are up to two times more likely than white adolescents to experience at least one arrest (Ryan and Testa, 2005) and to be disproportionately represented in the arrest and detention population (Herz, Ryan, and Bilchik, 2010). As a result, previous child welfare contact is highly correlated with the overrepresentation of African American youth in the juvenile justice system (see also Chapter 3). Girls in the child welfare system are also more likely to be detained by the juvenile justice system than nonfoster care girls. Girls’ histories of multiple foster home placements, child protection system policies that penalize girls for running away, and inadequate communication across the juvenile justice and child protection systems contribute to these disparities (Sherman, 2012).
Youth held in juvenile detention centers and other residential facilities exhibit high rates of mental disorder (see Chapter 3). Evidence suggests, however, that there are few differences between youth from different racial/
ethnic backgrounds on levels of symptoms at screening (Vincent et al., 2008) but African American youth may show greater levels of need than white youth on broader measures of mental health needs (Rawal et al., 2004). White youth are more likely to be ordered to services (Pumariega et al., 1998) or designated as severely mentally ill and referred for services than African American youth (Herz, 2001; Lopez-Williams et al., 2006; Maschi et al., 2008; Dalton et al., 2009), making them more eligible for smaller, more specialized treatment programs (Bishop, 2005). How much this differential in service involvement is attributable to juvenile justice system involvement is unclear (Garland et al., 2005).
Gender differences are more pronounced. Rates of symptoms at screening appear to be higher for girls than for boys (McCabe et al., 2002; Cauffman, 2004; Osterlind, Koller, and Morris, 2007; Vincent et al., 2008), and girls appear to have higher rates of prior maltreatment and family history of mental illness (McCabe et al., 2002). They are also more likely to be ordered to receive mental health services than boys (Yan and Dannerbeck, 2011).
As a general rule, studies documenting racial/ethnic differences suggest that blacks and other minorities experience a disproportionate amount of contact with these agencies than white youth. We turn again to these systems in our discussion of strategies for addressing racial/ethnic disparities.
Negative Stereotypes and Media Imagery of Minority Youth
Although negative stereotyping appears to have declined over the past two decades, negative images still remain quite commonplace (Bobo, 2011). Negative stereotypes and media imagery of minority youth may play a role in the differential treatment they receive from police and other actors in the juvenile justice system. For example, Bishop and Leiber (2012) suggest that although there is little evidence that police are overtly biased, they often do not have adequate information on which to base a decision to engage or arrest a youth and may be influenced by more subtle forms of bias arising from their perceptions of places and people.
Television crime reports contribute to negative stereotypes of minority youth. Iyengar’s (2010) analysis of local news shows demonstrated a systematic overemphasis on violent crime and associated crime with the actions of racial minorities. Bjornstrom and colleagues (2010) showed that ethnic and racial portrayals in television news reports were influenced by the context of the story itself (the race of the victim and the race of the perpetrator as well as the social structural context). Their study also showed that victimization in minority communities was routinely minimized.
Social Structure and Culture
A macrosociological explanation of disparities looks to racial inequality and concentrations of “underclass” poverty that influence levels of offending, enforcement practices, and formal court processes. Building on earlier research that examined the structural variations in court administration (community-level variations, budget, personnel, availability of facilities, rates of referral) and their impact on court processes (Feld, 1991), research by Sampson and Laub (1993) examined such community attributes as underclass poverty, racial inequality, wealth, court referral rates, mobility, urbanism, youth density, and criminal justice resources on court processes.
Sampson and Wilson (1995) theorized that black-white disparities resulted from racially segregated neighborhoods in which members of minority groups were differentially exposed to key violence-inducing and violence-protecting social mechanisms. Wilson (2009) later expanded on this idea by acknowledging the prevalence of powerful structural factors impacting blacks, such as discriminatory laws, policies, hiring, housing, and education and the interplay of structural factors and the stereotypical attitudes and assumptions of various ethnic and racial groups, including social science researchers. To arrive at a fuller understanding of the causes of racial inequalities, he suggests that it is necessary to go beyond the independent contributions of social structure and culture and to focus on how they interact to shape different group outcomes that embody racial inequality, a view strongly endorsed by other researchers (Kempf-Leonard, 2007; Bishop and Leiber, 2012).
The body of relevant evidence on racial/ethnic effects in the juvenile justice system demonstrates differential involvement of minorities in serious offending as well as differential selection and processing by the justice system. However, the race/ethnicity effects have been found to be both direct and indirect—operating both because of and through other factors. Moreover, the disparities are not uniform throughout the juvenile justice process (tending to be more common in the front-end processes, which afford much more discretion than back-end processes), and disparities seem to accumulate as youth are processed into the system (although studies of these trends are limited and may be hampered by various selection artifacts) (Engen, Steen, and Bridges, 2002; Leiber and Johnson, 2008). Other structural and contextual factors also may influence how minorities come to be disproportionately involved in the juvenile justice system, and these additional factors also need to be considered in designing possible strategies for reducing disproportionality.
More than a decade ago, the report Juvenile Crime, Juvenile Justice (National Research Council and Institute of Medicine, 2001, p. 229) concluded that the debate between the “behavior [differential offending] versus justice [differential selection]” positions has led to a “conceptual and methodological impasse.” Following this, Piquero (2008b) concluded that future research should move beyond the “which matters more” debate and instead seek to understand how both hypotheses can explain the overrepresentation of minorities in the system and then to identify steps that can be taken to lower any disparate effect and treatment. Few steps have been taken in this regard. To be sure, there has been much attention devoted to racial/ethnic disparities over the past decades, yet the empirical research has primarily focused on assessing the effect of differential offending and differential enforcement (and to a lesser extent differential processing) in an isolated manner. There has been little effort to use statistical methods to quantitatively partition the various identifiable factors (differential participation in the crimes that lead to involvement with the juvenile justice system, socio-economic/poverty effects, police patrol patterns in high-crime areas, family composition, etc.) that, in combination, produce racial/ethnic disparities.
The committee recognizes the challenges that must be overcome to quantify the various contributions to racial disproportionality, given the difficulty of assembling the necessary data, designing the study, and interpreting the findings. However, some progress seems possible by focusing separately on the sequential stages of the juvenile justice system. It is likely that some factors are more influential at spawning racial/ethnic differences at initial stages of the system (i.e., police decision to patrol and/or stop youth) compared with other stages of the system (i.e., prosecutor’s decision to charge and/or judge’s decision to institutionalize youth). Although this would entail a complex research effort, it should be undertaken for the purpose of helping to identify specific, actionable policy recommendations at each stage of the juvenile justice system. See Chapter 11 for a fuller presentation of research needs.
That said, the possibility of further research on the causes of racial/ethnic disparities should not delay policy actions aiming to reduce them. Many initiatives have been undertaken in recent years, and some promising strategies appear to have emerged. The next section highlights some of these efforts.
INTERVENTIONS AND PROMISING REFORM STRATEGIES
Several intervention efforts and policy reform initiatives and strategies have been developed to reduce DMC in the juvenile justice system. Little has been systematically documented about these strategies and their effectiveness—and even less has been published in the traditional academic
literature. Evaluations of these strategies have typically been undertaken by agencies involved in the implementation of reform strategies rather than by independent researchers.
The committee agrees with the general conclusion that there is “little objective evidence that interventions designed to reduce DMC actually do so” (Poulin, Orchowsky, and Iwama, 2011, p. 118). However, two research studies are worth noting. An evaluation of community-based delinquency prevention programs designed in part to address racial/ethnic disparities found that programs were successful in reducing recidivism, that recidivism was lowest among the high-attendance group, but that program effects on school outcomes were negligible (Welsh, Jenkins, and Harris, 1999). A second study assessed how legal and extralegal factors changed in predicting outcomes at two decision-making stages (intake and judicial disposition) about 10 years before and 10 years after the DMC mandate. Their findings regarding the impact of the DMC initiative were equivocal. Specifically, they found direct effects for race at intake, but such effects were less pronounced than at judicial disposition largely because of the “wide latitude for discretion at the front-end of the system” (Leiber, Bishop, and Chamlin, 2011, p. 26).
Research on differential juvenile offending, differential processing, and the broader structural context that impacts both suggests possible strategies worthy of exploration.
Addressing DMC at the Front End
Focusing on Arrest and Detention
Given the evidence that race is strongly associated (both directly and indirectly) with decisions made at the front end of the system (Engen et al., 2002; Bishop and Leiber, 2012), strategies targeted at reducing the likelihood of arrest and detention, particularly from sources of referral to the juvenile justice system, offer a promising approach to reduce racial disparities.
Juvenile Detention Alternatives Initiative. Funded by the Annie E. Casey Foundation, the Juvenile Detention Alternatives Initiative was designed to reduce reliance on secure detention by promoting changes to policies, practices, and programs. (See Chapter 9 for a fuller description.) The initiative has been credited with assisting in the closure of detention units or entire facilities as well as leading to reductions in Latino youth detained in Santa Cruz due to the opening of an evening reporting center (Office of Juvenile Justice and Delinquency Prevention, 2009a; Annie E. Casey Foundation at http://www.aecf.org/initiatives/jdai).
The W. Haywood Burns Institute. The Burns Institute works with community stakeholders and local agencies in a data-driven, consensus-based approach to change policies, procedures, and practices that result in the detention of low-offending youth of color and poor youth. As part of its technical assistance function, Burns reports some successes in developing DMC-reduction policies to reduce the number of youth who were held in secure detention and to develop alternatives to detention that have been shown to be related to a significant decrease in detention among black youth (Bell and Ridolfi, 2008; Bell et al., 2009; Poulin, Orchowsky, and Iwama, 2011, p. 106).
Models for Change. The MacArthur Foundation Models for Change Initiative was launched in 2004 in Pennsylvania and expanded to several other states, including Illinois, Louisiana, and Washington, and in 2007 the foundation established a county-level Action Network to address specific DMC initiatives in eight states. (For a more detailed description of the Models for Change initiative, see Chapter 9.) Anecdotal evidence suggests that several jurisdictions have initiated efforts to collect and analyze data on race/ethnicity across key decision points and have taken steps to use data to inform policy and practice. A Model for Change site (Philadelphia) has developed a minority youth–law enforcement training curriculum that was a joint project of the district attorney and the police department; in Berks County, Pennsylvania, the DMC Action Network enhanced Spanish-language capability and cultural competence, developed workforce opportunities, and showed some signs of reducing minority detentions through improved assessment screening and diversion (Armour and Hammond, 2009, p. 6). Griffin (2008) reports that the DMC Action Network in Peoria, Illinois, found that many arrests of black youth were for aggravated battery and that once alternative conflict strategies were started, arrests for black youth dropped significantly.
Working with the Child Welfare and School Systems
As noted, the child welfare and school systems are contributors to the overrepresentation of minority youth in the juvenile justice system. Researchers supported by child welfare organizations, such as the Child Welfare League and Georgetown’s Center for Juvenile Justice Reform, have been working for more than a decade to identify “crossover youth” and to develop an integrated, multisystem approach to program development and delivery of services (Wiig and Tuell, 2011). For those youth appropriately referred to the juvenile justice system, identifying appropriate services and placements for them at entry would aim to limit their deeper penetration into the system.
The differential treatment of minority students for disciplinary infractions is the object of close scrutiny by both the U.S. Department of Education and the Department of Justice. Publicly available data representing 85 percent of the nation’s students are being used to determine disparate discipline rates for suspensions and expulsions as well as arrests and referral to law enforcement.16 A recent rollout of the expanded Department of Education civil rights database and the Texas study showing the high degree of discretion being exercised by school administrators in suspension and expulsion decisions (Fabelo et al., 2011) have resulted in widespread media coverage and a collaborative project between the Justice and Education departments to address the “school to prison pipeline.” Among the goals of the initiative are to promote collaborative research and data endeavors, including evaluations of alternative disciplinary policies and interventions and to encourage positive discipline options and awareness of evidence-based and promising policies and practices among each state’s judicial and education leadership (U.S. Department of Justice, 2011).
An innovative legislative approach to reducing racial/ethnic disparities has been tried in Iowa and Connecticut. Iowa became the first state to require “minority impact statements” for proposed legislation related to crimes, sentencing, parole, and probation and for grants awarded by state agencies, and Connecticut requires racial/ethnic impact statements for bills and amendments that could, if passed, increase or decrease the pretrial or sentenced population of state correctional facilities (Armour and Hammond, 2009, p. 6). Although these legislative efforts have yet to be empirically evaluated for reducing DMC, they represent the kind of innovations that are needed in addressing a serious but admittedly complicated problem. The minority impact statement challenges all participating agencies to inventory their policies and practices to heighten awareness of contributing factors and provide a tool for monitoring progress.
Characteristics of Promising Strategies
Soler and Garry (2009) have highlighted some traits that are characteristic of promising strategies to address disparities. First, these efforts need to have community support, originate at the community level, and include community stakeholders. Second, strategies need to rely on data from several sources to paint a complete picture of the nature and extent of the problem. Third, strategies need to be transparent about both successes and
setbacks. Fourth, all interested parties need to be committed to long-term investment in lowering DMC that relies on evidence-based practices and follow-through with sustainable initiatives. It should also be added that a set of realistic expectations should be put in place so as to manage what stakeholders hope will happen and what is actually likely to happen in the short and long terms. Furthermore, DMC-related programs should have strong process evaluations in place prior to outcome evaluations being conducted on program effects because poorly implemented programs are likely to evince ineffective results and conclusions (Piquero, 1998).
Based on experiences in reducing disparities in the child welfare system and for crossover youth who enter the juvenile justice system, five general strategies have been identified (Chapin Hall Center for Children, 2009):
- Increase transparency—by building management information systems that collect race/ethnicity information.
- Reengineer structures and procedures—by reviewing processes and procedures routinely to determine whether they contribute to disparities.
- Change organizational culture—by influencing attitudes of agency staff and identifying the subtle ways attitudes can affect policy and practice.
- Mobilize political leadership—by building awareness and consensus among them.
- Partner with developing community and family resources to build political will.
The committee endorses these strategic suggestions and thinks they should be pursued.
Several National Academies reports have described with concern the differential handling of minorities by the justice system: Juvenile Crime, Juvenile Justice (National Research Council and Institute of Medicine, 2001); Fairness and Effectiveness in Policing: The Evidence (National Research Council, 2004a); and Informing America’s Policy on Illegal Drugs: What We Don’t Know Keeps Hurting Us (National Research Council, 2004b). Two reports have undertaken a broad review of the status of racial relations and racial trends (National Research Council, 1989, 2001a), and each contains thought-provoking chapters on racial trends in the administration of justice. Each aims for better understanding of the role that race and specifically racial disparities play in American culture and institutions.
Each reflects the complexity of the overrepresentation of minorities and the lack of easy solutions.
We know that racial/ethnic disparities are not reducible to either differential offending or differential selection. Many other factors affect disproportionality of minority youth in the juvenile justice system, including the troubling entrenched patterns of poverty, segregation, gaps in educational achievement, and residential instability. DMC exists in the broader context of a “racialized society” in which many public policies, institutional practices, and cultural representations operate to produce and maintain racial inequities.
The literature reflects continuing uncertainty about the relative contribution of differential offending, differential enforcement and processing, and structural inequalities to these disparities. However, the current body of research suggests that poverty, social disadvantage, neighborhood disorganization, constricted opportunities, and other structural inequalities—which are strongly correlated with race/ethnicity—contribute to both differential offending and differential selection, especially at the front end of juvenile justice decision making. Because bias (whether conscious or unconscious) also plays a role, albeit of unknown magnitude, juvenile justice officials should embrace activities designed to increase awareness of these unconscious biases and to counteract them, as well as to detect and respond effectively to overt instances of discrimination. Although the juvenile justice system itself cannot alter the underlying structural causes of racial/ethnic disparities in juvenile justice, many conventional practices in enforcement and administration magnify these underlying disparities, and these contributors are within the reach of justice system policy makers.
Based on the current knowledge base and the context in which DMC occurs, the committee identifies four reform strategies for moving the DMC agenda forward. We think, given the importance and persistence of the problem, that the existing data are sufficient to warrant serious consideration of these strategies.
First, reform efforts to reduce racial/ethnic disparities should pay special attention to the arrest and detention stages at the front end of the system. Reducing discretion by police and court officers through the use of written guidelines and risk assessment instruments; eliminating detention for youth who do not pose a danger; providing mental health, substance abuse, and other services up front so that youth can avoid penetrating deeper into the system; and providing alternatives to detention and alternatives to prosecution should all be part of an improved response to youth who are at the entry threshold of the juvenile justice system.
Second, a comprehensive reform strategy should encompass review of school disciplinary practices and elimination of those that are punitive and discretionary and are likely to result in a referral to the juvenile justice
system. As indicated earlier, schools are the source of numerous minority youth who are caught up in the discretionary disciplinary practices of schools and are referred often to law enforcement for nonserious offenses. More research is needed to understand the pipeline process and the role that various actors play (school resource officers, school management) in these referrals. Similarly, policies and practices involving youth who have ties to the mental health and child welfare systems need to be carefully assessed to ensure that the reasons for their handling are legitimate and their subsequent processing by the juvenile justice system is appropriate and nondiscriminatory.
Third, any reform strategy should focus on eliminating formal and informal agency policies and practices that are shown to disproportionately disadvantage minority youth. To do so will require the identification of key decision points and decision-making criteria that appear in practice to fall disproportionately on minority youth and perhaps to reflect implicit bias. It will also require the availability of proper legal representation for all minority youth and, for Hispanic youth and their families, translators.
Fourth, reform efforts are needed to increase the accountability of national, state, and local governments for reducing racial/ethnic disparities. At the local level, political leaders need to take responsibility for identifying the extent of disproportionality in their communities. At the state level, cabinet-level leadership on juvenile justice administration should monitor efforts to address these disparities and to provide the necessary resources to enable the necessary data to be collected and reported. As mentioned earlier, state legislatures should consider statutes that would give heightened urgency and visibility to this problem, including establishing oversight bodies. Even though state policy makers do not control all the levers that must be engaged to address the problem, they do have the power to command attention. Part of the long-term solution is for state juvenile justice leaders to keep this issue at the forefront of the reform agenda. Finally, reform strategies at the national level, specifically those involving the OJJDP, the lead agency on this issue, are described in Chapter 10.