Analyzing Data To Identify the Extent of the Problem

OJJDP has developed a standard equation for assessing the relationship between the proportions of minorities in the juvenile justice system and in the overall juvenile population. This equation, shown in table 2, can be used to calculate rates of overrepresentation at any point in the juvenile justice system, up to and including secure detention or confinement. The actual mathematical calculation is relatively straightforward. The experiences of the pilot States, coupled with lessons learned from the literature, suggest, however, that applying the calculation within a real world context is complicated by several factors. These factors include(1) acquiring accurate quantitative data about juveniles within the justice system and the broader community; (2) conducting systematic data analysis; (3) interpreting the data within the local social and political context; and (4) engaging stakeholders in the data analysis process.

The importance of two of these factors—conducting systematic analyses and interpreting data within the local social and political context—was a clear lesson from the pilot States. Systematically identifying the extent of minority overrepresentation and the associated decisions concerning juvenile justice system processing provides the basis for identifying the factors contributing to DMC and designing appropriate interventions. Lessons learned from this initiative, together with other findings from the DMC literature, provide guiding principles for assessing the extent and analyzing the causes of disproportionate minority representation. One of the lessons from the pilot State experiences is that the problem of overrepresentation is reflected neither in a single decision nor in the presence or absence of a single attribute in a juvenile justice system. No single feature distinguished communities with high levels of overrepresentation from those with low levels. Even among communities with high levels of overrepresentation, no single decision point or decisionmaker was identified. Thus, understanding overrepresentation is a matter of understanding how a specific juvenile justice system operates, with all its interdependent parts, to result in more minority juveniles entering and penetrating further into the system. This knowledge is gained through the acquisition of accurate quantitative data and the application of a systemic analysis model. These principles and associated processes are described briefly below.

Acquire Accurate Quantitative Data

Accurate quantitative data about the number and characteristics of juveniles in the justice system are often difficult to obtain. Historically, information about juveniles in the justice system was agency specific, largely inaccessible, and lacking in standardization. The data typically used to assess DMC came from the records of juvenile justice and other relevant State agencies and, in some cases, from census records. As State and local juvenile justice systems moved to automated data collection systems, computerized records were frequently incomplete, missing and/or inaccurate, and rarely comparable across jurisdictions. The move by State and local juvenile justice agencies to develop more adequate information systems ironically creates a potential problem for analyzing DMC across local systems. The more reporting systems “aggregate up”—from local, to county, to State—the more likely that evidence of racial disparity may be lost or hidden (Pope and Feyerherm, 1993).

Conduct Systematic Data Analysis

The experiences of the pilot States, reinforced by significant research, suggest that a DMC assessment must employ a systemic approach. In doing so, overrepresentation is defined broadly as a situation in which minority juveniles have unequal probabilities (as compared with nonminority juveniles) of negative processing decisions within the juvenile justice system. As discussed previously, focusing only on the end result—confinement—instead of on the overall process that creates overrepresentation limits understanding of the DMC problem, the contributing factors, and the need for a full range of coordinated interventions.

National research efforts have found that the process of quantifying disproportionality must examine multiple decision points in juvenile processing because race/ethnicity effects at any one point may be canceled out or enhanced at a subsequent point. For a more complete picture of how minority status does or does not influence confinement and other juvenile justice processing decisions, multiple decision points must be examined (Pope and Feyerherm, 1993).

To fully disentangle the effects of race/ethnicity on juvenile justice system decisionmaking, an analytic model developed by Feyerherm and colleagues was refined during the OJJDP DMC initiative. This generic model depicts the major decision points common to virtually all juvenile justice systems, including:

• Decision to arrest a juvenile, who then appears in juvenile court for intake processing.

• Decision at intake either to dispense or to process further.

• Decision to remove the juvenile from the current living arrangements during processing (e.g., detention or shelter home care).

• Decision to file a formal petition of delinquency, engage in other formal action such as a citation or fine, or seek informal resolution such as a warning or a remand of a juvenile to his or her parents without going to court.

• Decision to resolve the case by informal probation, formal probation, or custody transfer. Informal probation includes diversion programs and other mandated activities for nonadjudicated youth. Formal probation is for adjudicated youth who are not assigned to detention or confinement but have a court-appointed probation officer and court-defined responsibilities (e.g., frequency and type of contact, frequency of urinalysis, etc.).

Decision point charts specific to a particular juvenile justice system can be used in combination with the analysis model to specify all possible decision combinations and summarize the operation of the overall system. This approach helps to identify the decisions, or combinations of decisions, potentially influenced by race/ethnicity. Through probability analysis, the effects of race/ethnicity at each decision point, or combination of points, in the juvenile justice system can be assessed for minority overrepresentation.³

This analysis model can also be used to identify the probability of moving from one point in the juvenile justice system to another. For example, if there is an intake referral and the juvenile is detained (e.g., removed from current living arrangement), the probability that a petition will be filed can be calculated. A comparison of these probabilities for majority and minority juveniles may identify combinations of decisions that are likely to impact minority juveniles differently. In addition, the model provides the opportunity to identify the extent to which the overall system appears to operate differently for majority and minority juveniles by calculating the “accumulation” of small race/ethnic-related discrepancies.

Several of the pilot States applied this model to the assessment of DMC within their juvenile justice systems to support their analysis of disproportionate minority representation. This approach also helped to assess the availability of appropriate data for the analysis because the model requires specific juvenile justice data for each juvenile justice processing decision.

Interpret the Data Within the Local Context

Researchers have long recognized that findings from quantitative data analysis are meaningful only if they accurately reflect the context of the information. Context is revealed through the collection of qualitative information about the juvenile justice system and the community. The importance of interpreting the data within the local social and political context was reinforced by the pilot State experiences. For example, one site’s quantitative data indicated that minority juveniles were not attending their diversion programs and, ultimately, were confined. An examination of client case records revealed, however, that many juveniles were referred to diversion programs far from their homes and lacked transportation to and from the programs. An analysis of the client case records showed that these juveniles were noncompliant because of physical barriers to program completion. This example illustrates how qualitative information can support meaningful interpretation of the quantitative data analysis.

Accurately identifying the scope and intensity of DMC requires an understanding of social and political forces that may account for, or artificially inflate or suppress, the appearance of disproportionality. For example, in one pilot State, the DMC incidence increased from 1990 to 1994. An investigation of the contributing factors revealed that this time period coincided with the “war on drugs” and the resultant intensification of daily law enforcement activities within urban areas. A closer examination of the county data revealed that the largest increases in disproportionate arrest rates were in those counties with large urban populations and that increases in arrests were associated with drug-related offenses. Although the explanation of increased minority confinement does not minimize its importance, the social and political events that precipitated the increase help to put it in perspective and identify opportunities for DMC interventions.

Engage Stakeholders in the Data Analysis Process

Disproportionate minority confinement encompasses wide-ranging issues involving many stakeholders: policymakers, juvenile justice system professionals, the broad community of service providers, minority juveniles, and their families. Among the DMC pilot States, the lead agencies sought full participation of all key stakeholders to assist in data collection and to support the analysis and interpretation of results. Stakeholder participation increased access to data under their control and their confidence in the quality of data provided. This, in turn, appeared to increase stakeholders’ consensus as to the meaning of the data and their commitment to the findings from the data analysis. Creating a common understanding of DMC issues also promoted stakeholder agreement about factors contributing to DMC. By jointly addressing DMC issues, key stakeholders learned to work together and build a foundation for future collaboration on DMC, the contributing factors, and the need for interventions.

Disproportionate Minority Confinement: Lessons Learned From Five States	OJJDP Bulletin • December 1998