Proposed Methods for Measuring Disproportionate Minority Contact (DMC)

Proposed Methods for Measuring Disproportionate Minority Contact (DMC)
as required by the Juvenile Justice and Delinquency Prevention Act §223(a)(22) Revised 2002 Dr. William Feyerherm Portland State University Portland, Oregon Dr. Jeffrey Butts The Urban Institute Washington, DC

Contents Why do we measure DMC?
Why do we need a new method of measuring DMC? What is the new method? What are the next steps in implementing the new method?

First, a Review of Key Terms

Minority? Racial and Ethnic Categories
(1) White A person having origins in any of the original people of Europe, the Middle East, or North Africa. (2) American Indian or Alaska Native A person having origins in any of the original peoples of North and South America (including Central America), and who maintains tribal affiliation or community attachment. (3) Asian A person having origins in any of the original peoples of the Far East, Southeast Asia, or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam. (4) Black or African American A person having origins in any of the black racial groups of Africa. (5) Hispanic or Latino A person of Cuban, Mexican, Puerto Rican, Cuban, South or Central American, or other Spanish culture or origin, regardless of race. (6) Native Hawaiian or Other Pacific Islander A person having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Island. Source:

Race and Ethnicity Categories in DMC
Should account for Hispanic/Latino status (e.g., “White” should include only non-Hispanic whites) Groups more detailed than the six major groups may be used if they can be aggregated back into the six major groups Any group accounting for 1% or more of the juvenile population (i.e., youth subject to juvenile court jurisdiction and juvenile justice handling) should be assessed independently Reports should describe the categories and allocation rules used – and be consistent throughout each report Data source:

Why Measure DMC?

DMC is a Question, Not an Answer
Measuring DMC is like taking vital signs in a hospital emergency room – it doesn’t identify illness or tell you how to treat it – it alerts you to potential problems and tells you where to focus your diagnostic efforts

OJJDP’s Goals for DMC Identify existence/extent of disproportionality – “between race” comparisons within jurisdictions and at specific decision points in the system Assess data about DMC to target detailed studies, identify points of needed intervention, and allocate resources for system interventions Intervene to reduce DMC – assist policymakers in choosing jurisdictions that should receive increased attention and intervention Evaluate how DMC responds to policy initiatives and system interventions Monitor trends in DMC within and across jurisdictions

The Meaning of DMC has Changed
No longer disproportionate “confinement” but Disproportionate”contact” with the juvenile justice system

Contact Initial encounter with law enforcement (i.e., arrest) and
Ongoing juvenile justice contacts, such as: Diversion Referral to juvenile court Hold in secure detention Issuance of petition Adjudication as delinquent Placement on probation Placement in secure facilities Transfer to adult court And other points where sufficient data exist: Pre-arrest diversion Aftercare Revocation of aftercare etc.

DMC Goals are not Changing
Disproportionate Minority Confinement The purpose of the DMC statute and regulation is to ensure equal and fair treatment for every youth in the juvenile justice system, regardless of race and ethnicity

DMC Goals are not Changing
Disproportionate Minority Contact new The purpose of the DMC statute and regulation is to ensure equal and fair treatment for every youth in the juvenile justice system, regardless of race and ethnicity

Old Method DRI “Disproportionate Representation Index”
Why a New Method? Old Method DRI “Disproportionate Representation Index” New Method RRI “Relative Rate Index”

% Old Method (DRI) Ratio Confined Youth Total Population
All Races & Ethnicities A Specific Ethnic Group Ratio % Confined Youth Total Population

# # Ratio New Method (RRI) Per-capita Rate Confined Youth
Ethnic Group 1 Ethnic Group 2 # Per-capita Rate Ratio # Confined Youth Total Population

The New Method (RRI): • Reduces Statistical Bias • Allows Fair and Accurate Comparisons • Handles Multiple Racial/Ethnic Groups

Consider two fictional places where Latinos are twice as likely as whites to be confined:
Latinos are twice as likely to be confined in both places, but the DRI distorts this equivalence Diversityville Homogenous Town Arrested Confined 1, White 1, Latino 10% 20% Confinement probability 67% Latino % of confined youth Divide 10% 50% Latino % of all youth only difference Equals DRI = 1.3 6.7

What other choice is there ?
What does this mean? The DRI is a biased estimator What other choice is there ? Using rates instead of proportions

Developing the RRI Creating Rates Examining Individual Decision Points

Youth Population (Ages 10-17)
Consider Two Hypothetical States Youth Population (Ages 10-17) State A: ,000 State B: ,000 Same youth population size, different demographic mix and different confinement population

State A State B 20,250 204,750 225,000 9% 91% 100% Black White Total
Youth Confined on July 1, 2003 20, ,750 225,000 Total Youth Population 9% 91% 100% Black White Total 60 200 260 23% 77% 100% Youth Confined on July 1, 2003 Black White Total 510 60 570 160,000 65,000 225,000 Total Youth Population 89% 11% 100% 71% 29% State B

State A State B 2.6 0.9 Black White Total 60 200 260 23% 77% 100%
Youth Confined on July 1, 2003 Total Youth Population DRI Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% State B Youth Confined on July 1, 2003 Total Youth Population Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% State B Youth Confined on July 1, 2003 Total Youth Population DRI Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% Using the old DRI method, it appears that DMC in State A is twice that of State B State B Youth Confined on July 1, 2003 Total Youth Population DRI Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

But, remember, the DRI is a biased estimator
It is affected by the relative size of minority youth populations

Rate Ratios are unaffected by either: 1
Rate Ratios are unaffected by either: 1. the relative proportion of minorities in the total youth population, or 2. the number of different population groups that must be compared

Let’s go back to State A and State B

Confined youth per 1,000 juveniles in the general population
State A Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% Confined youth per 1,000 juveniles in the general population State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

State A State B Black White Total 60 200 260 23% 77% 100%
Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% 3.0 State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% To calculate the “relative rate index” (or RRI), divide one rate over the other to create a ratio. State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% RRI = 3.0 State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100%

Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 60 200 260 23% 77% 100% 20, ,750 225,000 9% 91% 100% RRI = 3.0 Note the very different conclusions suggested by the DRI and the RRI State B Youth Confined on July 1, 2003 Total Youth Population DRI Rate Black White Total 510 60 570 89% 11% 100% 160,000 65,000 225,000 71% 29% 100% RRI = 3.6

Developing the RRI Creating Rates Examining Individual Decision Points

waived to criminal (adult) court
total population placed out of home probation other sanctions dismissed/released formally adjudicated The old method (DRI) compares percentages of minority youth at each stage of the juvenile justice system to the percentage of minorities in the population as a whole arrests petitioned placed out of home probation other sanctions dismissed/released not adjudicated court referrals placed out of home probation other sanctions dismissed/released not petitioned

total population placed out of home probation other sanctions dismissed/released formally adjudicated arrests petitioned placed out of home probation other sanctions dismissed/released not adjudicated court referrals placed out of home probation other sanctions dismissed/released not petitioned

total population placed out of home probation other sanctions dismissed/released formally adjudicated The new method (RRI) compares the percentage of minority youth at each stage of the juvenile justice system to the percentage of minorities at the previous stage arrests petitioned placed out of home probation other sanctions dismissed/released not adjudicated court referrals placed out of home probation other sanctions dismissed/released not petitioned

total population placed out of home probation other sanctions dismissed/released formally adjudicated arrests petitioned placed out of home probation other sanctions dismissed/released not adjudicated court referrals placed out of home probation other sanctions dismissed/released not petitioned

To make these comparisons possible, two compromises are likely:
The definitions of processing stages will probably vary slightly across jurisdictions Data will have to be for youth handled in one year (e.g., cases disposed in one year) even though they may have been arrested in different years Thus, calculations won’t be “true rates” – Not: “the charging rate among youth referred this year” But: “number of youth charged this year divided by the number of youth referred this year”

How Should States Calculate the RRI?
The same type of data used for the DRI can be used to calculate the RRI 1. Number of all youth… a) in the total population, and b) at various stages in the j.j. process 2. Number of youth of specific races and ethnicities… a) in the total population, and b) at various stages in the j.j. process

Let’s take a closer look.
This form contains real DMC data from Fresno, CA Let’s take a closer look.

The DRI requires states to calculate the proportion of various ethnic groups among the the general youth population. Then, the proportion of those youth in the committed population.

The ratio of these two proportions is the DRI.
In the case of Fresno, the DRI equals 2.5

A new data-entry tool for calculating the RRI could use the same data

The array of data required for the DRI
The same information could be entered in a spreadsheet and used to create the RRI and various tables or reports

through 12 / 2002 (Month / Year)
All case counts could be entered one time in one place AREA REPORTED Data Entry Section State : California County: FRESNO Reporting Period 1 /2002 (Month / Year) through 12 / (Month / Year) Total Youth White African-American Hispanic Asian Pacific Islanders Native American Other/ Mixed All Minorities 1. Population at risk (age 10_ through _17 ) 118,722 39,117 6,460 52,433 19,750 972 79,615 2. Juvenile Arrests 13,585 3,058 2,055 7,220 1,091 29 132 10,527 3. Refer to Juvenile Court 4. Cases Diverted 306 113 28 136 19 10 193 5. Cases Involving Secure Detention 2,314 401 354 1,300 243 8 1,913 6. Cases Petitioned 5,859 1,000 901 3,113 523 16 36 4,589 7. Cases Resulting in Delinquent Findings 4,058 555 894 2,195 384 15 3,503 8. Cases resulting in Probation Placement 2,501 585 362 1,330 201 13 1,916 9. Cases Resulting in Confinement in Secure Juvenile Correctional Facilities 1,629 284 241 908 189 3 4 1,345 10. Cases Transferred to Adult Court 24 7 22 Meets 1% rule? Yes No A prototype spreadsheet for entering DMC data

Case counts would be organized by multiple racial/ethnic groups and for multiple processing stages AREA REPORTED Data Entry Section State : California County: FRESNO Reporting Period 1 /2002 (Month / Year) through 12 / (Month / Year) Total Youth White African-American Hispanic Asian Pacific Islanders Native American Other/ Mixed All Minorities 1. Population at risk (age 10_ through _17 ) 118,722 39,117 6,460 52,433 19,750 972 79,615 2. Juvenile Arrests 13,585 3,058 2,055 7,220 1,091 29 132 10,527 3. Refer to Juvenile Court 4. Cases Diverted 306 113 28 136 19 10 193 5. Cases Involving Secure Detention 2,314 401 354 1,300 243 8 1,913 6. Cases Petitioned 5,859 1,000 901 3,113 523 16 36 4,589 7. Cases Resulting in Delinquent Findings 4,058 555 894 2,195 384 15 3,503 8. Cases resulting in Probation Placement 2,501 585 362 1,330 201 13 1,916 9. Cases Resulting in Confinement in Secure Juvenile Correctional Facilities 1,629 284 241 908 189 3 4 1,345 10. Cases Transferred to Adult Court 24 7 22 Meets 1% rule? Yes No

Note the addition of “Refer to Juvenile Court” as a stage AREA REPORTED Data Entry Section State : California County: FRESNO Reporting Period 1 /2002 (Month / Year) through 12 / (Month / Year) Total Youth White African-American Hispanic Asian Pacific Islanders Native American Other/ Mixed All Minorities 1. Population at risk (age 10_ through _17 ) 118,722 39,117 6,460 52,433 19,750 972 79,615 2. Juvenile Arrests 13,585 3,058 2,055 7,220 1,091 29 132 10,527 3. Refer to Juvenile Court 4. Cases Diverted 306 113 28 136 19 10 193 5. Cases Involving Secure Detention 2,314 401 354 1,300 243 8 1,913 6. Cases Petitioned 5,859 1,000 901 3,113 523 16 36 4,589 7. Cases Resulting in Delinquent Findings 4,058 555 894 2,195 384 15 3,503 8. Cases resulting in Probation Placement 2,501 585 362 1,330 201 13 1,916 9. Cases Resulting in Confinement in Secure Juvenile Correctional Facilities 1,629 284 241 908 189 3 4 1,345 10. Cases Transferred to Adult Court 24 7 22 Meets 1% rule? Yes No

The data entry tool would automatically identify which groups must be examined AREA REPORTED Data Entry Section State : California County: FRESNO Reporting Period 1 /2002 (Month / Year) through 12 / (Month / Year) Total Youth White African-American Hispanic Asian Pacific Islanders Native American Other/ Mixed All Minorities 1. Population at risk (age 10_ through _17 ) 118,722 39,117 6,460 52,433 19,750 972 79,615 2. Juvenile Arrests 13,585 3,058 2,055 7,220 1,091 29 132 10,527 3. Refer to Juvenile Court 4. Cases Diverted 306 113 28 136 19 10 193 5. Cases Involving Secure Detention 2,314 401 354 1,300 243 8 1,913 6. Cases Petitioned 5,859 1,000 901 3,113 523 16 36 4,589 7. Cases Resulting in Delinquent Findings 4,058 555 894 2,195 384 15 3,503 8. Cases resulting in Probation Placement 2,501 585 362 1,330 201 13 1,916 9. Cases Resulting in Confinement in Secure Juvenile Correctional Facilities 1,629 284 241 908 189 3 4 1,345 10. Cases Transferred to Adult Court 24 7 22 Meets 1% rule? Yes No

The Data Entry Tool Could Also Display Results Automatically
Comparisons across processing stages within race/ethnicity categories Comparisons across race/ethnicity categories within processing stages

For example, one automatic report could display the key results for each racial or ethnic group

The ratio of the two rates equals 4.07 --- or 318.11 divided by 78.18
Area: FRESNO, CA Population Group: AFRICAN-AMERICAN Reporting Period: 01/ /2002 Statistical Signif.? (p<. 05) Stages White Af-Amer RRI Number Rate Number Rate At risk (ages 10-17) Arrested Referred to Court Diverted Detained Petitioned/Charged Adjudicated/Found Delinq. Placed on Probation Securely Confined Transferred to Adult Ct. 39,117 3,058 -- 113 401 1,000 555 585 284 --- 78.18 -- 3.70 13.11 32.70 55.55 105.41 51.17 --- 6,480 2,005 -- 28 354 901 894 362 241 7 318.11 -- 1.36 17.23 43.84 99.22 65.23 26.96 0.78 4.07 -- 0.37 1.31 1.34 1.79 0.62 0.53 Yes -- --- White youth were arrested at a rate of per 1, because 3,058 divided by 39,117 times 1,000 equals 78.18 The ratio of the two rates equals or divided by 78.18 African-American youth were arrested at a rate of because 2,055 divided by 6,460 times 1,000 equals

Area: FRESNO, CA Population Group: AFRICAN-AMERICAN Reporting Period: 01/ /2002 Statistical Signif.? (p<. 05) Stages White Af-Amer RRI Number Rate Number Rate At risk (ages 10-17) Arrested Referred to Court Diverted Detained Petitioned/Charged Adjudicated/Found Delinq. Placed on Probation Securely Confined Transferred to Adult Ct. 39,117 3,058 -- 113 401 1,000 555 585 284 --- 78.18 -- 3.70 13.11 32.70 55.55 105.41 51.17 --- 6,480 2,005 -- 28 354 901 894 362 241 7 318.11 -- 1.36 17.23 43.84 99.22 65.23 26.96 0.78 4.07 -- 0.37 1.31 1.34 1.79 0.62 0.53 Yes -- --- Differences in the relative rate index could guide further investigations

Area: FRESNO, CA Population Group: AFRICAN-AMERICAN The black-to-white ratio of arrest rates is 4.07, meaning that African-American youth are 4 times more likely than whites to be arrested. Reporting Period: 01/ /2002 However, the black-to-white ratio of petitioning is just 1.34, meaning that, once arrested, African-Americans are only slightly more likely to be charged. Statistical Signif.? (p<. 05) Stages White Af-Amer RRI And, the black-to-white ratio of secure confinement is 0.53, which suggests that once adjudicated, African-American youth are half as likely to be confined. Number Rate Number Rate At risk (ages 10-17) Arrested Referred to Court Diverted Detained Petitioned/Charged Adjudicated/Found Delinq. Placed on Probation Securely Confined Transferred to Adult Ct. 39,117 3,058 -- 113 401 1,000 555 585 284 --- 78.18 -- 3.70 13.11 32.70 55.55 105.41 51.17 --- 6,480 2,005 -- 28 354 901 894 362 241 7 318.11 -- 1.36 17.23 43.84 99.22 65.23 26.96 0.78 4.07 -- 0.37 1.31 1.34 1.79 0.62 0.53 Yes -- ---

Area: FRESNO, CA Population Group: AFRICAN-AMERICAN Reporting Period: 01/ /2002 Statistical Signif.? (p<. 05) Stages White Af-Amer RRI Number Rate Number Rate At risk (ages 10-17) Arrested Referred to Court Diverted Detained Petitioned/Charged Adjudicated/Found Delinq. Placed on Probation Securely Confined Transferred to Adult Ct. 39,117 3,058 -- 113 401 1,000 555 585 284 --- 78.18 -- 3.70 13.11 32.70 55.55 105.41 51.17 --- 6,480 2,005 -- 28 354 901 894 362 241 7 318.11 -- 1.36 17.23 43.84 99.22 65.23 26.96 0.78 4.07 -- 0.37 1.31 1.34 1.79 0.62 0.53 Yes -- --- These cross-stage patterns in the RRI could lead to very different conclusions than an analysis based only on the DRI for confinement … Why?

Relative Rate Index Compared with White Juveniles
Other reports could portray decision-specific comparisons of the RRI by race and ethnicity Relative Rate Index Compared with White Juveniles African- American Hispanic Asian 2. Juvenile Arrests 4.07 1.76 0.71 3. Refer to Juvenile Court --- --- --- 4. Cases Diverted 0.37 0.51 0.47 5. Cases Involving Secure Detention 1.31 1.37 1.70 6. Cases Petitioned 1.34 1.32 1.47 7. Cases Resulting in Delinquent Findings 1.79 1.27 1.32 8. Cases resulting in Probation Placement 0.62 2.27 0.34 African-American youth are four times more likely to be arrested than white youth. Hispanic youth are twice as likely as white youth to be arrested. 9. Cases Resulting in Secure Juv Confinement 0.53 0.81 0.96 10. Cases Transferred to Adult Court --- --- --- What other differences do you see, and what could they mean?

The new method also allows stage-specific comparisons across jurisdictions

How should these comparisons be interpreted?
Comparing Two Jurisdictions Population Based Relative Rate of African- ARREST (compared to White) Amer. Hisp. Asian Alameda FRESNO Population Based Relative Rate of ADJUDICATION (compared to White) Alameda FRESNO Population Based Relative Rate of SECURE CONFINEMENT (compared to White) Alameda FRESNO How should these comparisons be interpreted? In this example, each rate is calculated by dividing the number of youth at that stage of the system by the number of youth in the general population (or, “population based rates).

Conclusions The RRI should be viewed as a set of “vital signs” for system monitoring and to guide analysis The shift to “Contact” allows a more encompassing examination of the juvenile justice system The shift to rates and the Relative Rate Index will eliminate many of the problems inherent in the DRI A single data entry tool could automatically calculate all necessary rates and ratios, tables, and reports

Next Steps -- In 2003 and 2004, OJJDP will:
Revise DMC Technical Assistance Manual Design and disseminate new a data-entry tool (spreadsheet first, then move to web-based) Train consultants and contractors on the use and interpretation of the RRI – make other T.A. available Convert previous data reports to RRI approach in order to establish comparative / historical base

Proposed Methods for Measuring Disproportionate Minority Contact (DMC)
For more information about this presentation: Dr. William Feyerherm Dr. Jeffrey Butts For more information about OJJDP’s DMC program: Dr. Heidi Hsia

Proposed Methods for Measuring Disproportionate Minority Contact (DMC)

Similar presentations

Presentation on theme: "Proposed Methods for Measuring Disproportionate Minority Contact (DMC)"— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

Proposed Methods for Measuring Disproportionate Minority Contact (DMC)

Similar presentations

Presentation on theme: "Proposed Methods for Measuring Disproportionate Minority Contact (DMC)"— Presentation transcript:

Similar presentations

About project

Feedback