ALAMEDA COUNTY AB 109 EVALUATION CLIENT OVERVIEW AND

ALAMEDA COUNTY AB 109 EVALUATION: CLIENT OVERVIEW AND OUTCOMES REPORT – PRELIMINARY FINDINGS May 6, 2020 Ardavan Davaran, Ph. D. Nicole Liner-Jigamian, MPH MSW

Agenda 2 Project Overview Methodology Jail and Probation Populations and Service Trends Probation Violations and Recidivism Outcomes

3 Project Overview

Key Components of Report 4 Population Overviews � Jail Population � Probation Population Service Overview � Number of Individuals Receiving AB 109 -Funded Services, by Service Type Recidivism Outcomes � Revocations � New Convictions Service Receipt and New Convictions

5 Methodology

Methodology 6 Data Sources Behavioral Health Department Contracted Service Providers District Attorney’s Office Analysis Housing and Community Development Information Technology Department Probation Department Public Defender’s Office Sheriff’s Office Descriptive Statistics to measure probation population type (e. g. PRCS, county realigned, felony, and mandatory supervision), demographic and criminal history information, sentencing outcomes, jail populations, service receipt, and recidivism and revocation outcomes. Logistics Regression & Predicted Probabilities to compare the probability of recidivism for individuals who did and did not receive services. Survival Analysis to examine the time to recidivism and compare the risk of recidivating for individuals who did and did not receive services.

7 Jail and Probation Populations and Service Trends

Jail Population 8 Jail Bookings, by Charge Severity Key Findings The average daily jail population has declined since October 2011. A greater proportion of bookings are for misdemeanor than felony offenses, and the average length of stay has declined over time. 8 963 10 199 5 189 6 399 FY 11/ 1 2 Misdemeanor Average Daily Jail Population by FQ Average Length of Stay in Days, by Charge Severity 3 317 50 9 FY Felony 11/ 12 51 11 45 10 39 10 32 9 39 36 10 10 31 9 FY F F F F 11/ Y 12/ Y 13/ Y 14/ Y 15/ Y 16/ Y 17/ Y 18/ 12 13 14 15 16 17 18 19 Misdemeanor Felony

9 Pre-Sentencing and In-Custody Services Early Intervention Court Key Findings The County continues to invest in programs and resources to support individuals prior to and at sentencing. *Service data for FY 19/20 is through December 2019. Number of Individuals Served Annually by OMHT AB 109 -Funded Social Workers 299 218 86 80 FY 14/15 132 141 FY 15/16 54 58 FY 16/17 FY 17/18 182 29 FY 18/19 PRCS & 1170(h) Sentenced Between April 2016 and April 2020: • 154 individuals participated in the program • 70 successfully graduated and had their cases dropped • 18 were removed from the program to face traditional sentencing • 66 are currently enrolled 168 30 FY 19/20* Between FYs 2015/2016 and 2018/2019: • 255 individualized case plans were submitted to the court and 85% were accepted In fiscal year 2018/2019: • 41 individualized case plans were submitted and 90% were accepted AC Justice Restoration Project • 154 individuals participated in the program (program started September 2017)

Probation Populations 10 Key Findings Since October 2011, the number of individuals under Probation supervision has declined, from almost 10, 000 individuals at the end of 2011 to just over 8, 000 at the end of 2018. Average Daily Number of Individuals on Probation by FQ 5 179 4 665 4 552 86 2 2 909 631 31 FY F 11/ Y 12/ 12 13 CR During that same period, the number of new probation cases declined by almost half, from almost 900 to just over 450. FY 15/ Felony 16 PRCS FY 18/ 19 MS Number of New Probation Supervision Case Starts by FQ 349 319 214 234 1 145 84 0 FY FY F 11/ Y 12/ 12 13 CR Felony 15/ 16 MS FY PRCS 18/ 19

AB 109 -Funded Services 11 Key Findings Available Services Beginning in 2015, a greater variety of AB 109 -funded services were available to individuals under Probation supervision. � In Alameda County, AB 109 -funded services tend to target higher risk populations, as evidenced by the large proportion of individuals with PRCS cases who receive services. Percentage of Supervised Population Receiving One or More Services Between 2015 and 2019 13% PRCS (n = 1, 472) County Realigned (n = 4, 159) 17% 7% Felony (n = 2, 876) Mandatory Supervision (n = 30) Realignment Housing Project (RHP) � 38% Housing Abode & East Oakland Community Program (EOCP) Bay Area Community Services (BACS) and Building Opportunities for Self-Sufficiency (BOSS) Men of Valor Academy (MOVA) Shelter Employment Transition Day Reporting Center (T/DRC) Mentoring and Reengagement: For Us By Us (FUBU) Education: Five Keys Charter School

Housing Services 12 Key Findings Across the study period, RHP Abode and EOCP have served the most individuals because they have been the main AB 109 funded housing service providers. The number of individuals served by MOVA Shelter declined between 2016 and 2019. *Service data for FY 19/20 is through November/December 2019. Number of RHP BACS and BOSS Clients Served by FY 47 Number of RHP Abode and ECOP Clients Served by FY 442 270 26 342 405 502 317 122 FY FY 12/13 13/14 14/15 15/16 16/17 17/18 18/19 19/20* Number of MOVA Shelter Clients Served by FY 109 100 71 55 58 22 FY 15/16 FY 18/19 FY 19/20* FY 16/17 FY 17/18 FY 18/19 FY 19/20*

Employment, T/DRC, FUBU, and Education Services 13 Number of FUBU Clients Served by FY Key Findings Since coming online, AB 109 funded programs and services have consistently served individuals on probation. *Service data for FY 19/20 is through November/December 2019. 243 433 479 223 125 FY 18/19 FY 19/20* 66 FY 17/18 Number of T/DRC Clients Served by FY 226 198 182 195 FY 15/16 FY 16/17 FY 17/18 FY 18/19 FY 14/15 FY 15/16 FY 16/17 103 204 FY 17/18 FY 18/19 FY 19/20* Number of Education Clients Served by FY 51 FY 14/15 160 52 Number of Employment Clients Enrolled by FY 467 138 FY 19/20* FY 16/17 FY 17/18 74 FY 18/19 59 FY 19/20*

14 Probation Violations and Recidivism Outcomes

Overview 15 Cohort 1 Cohort 2 Started Probation October 1, 2011 – December 31, 2014 Started Probation January 1, 2015 – December 31, 2018 Population Size 9, 485 Population Size 8, 537

Demographic Profile 16 Key Findings Across demographic and criminal justice characteristics, the two cohorts are quite similar. 19% of people who started probation between 2015 and 2018 received services, compared to only 4% who started between October 2011 and December 2014.

Revocation Outcomes 17 Key Findings Jail Bookings by Violation Type 41% of individuals who started probation between 2015 and 2018 were revoked, compared to 28% who started between October 2011 and December 2014. 5 666 2 375 42% 27% 42% 26% 41% 28% 6 356 5 229 682 919 736 816 FY FY 11/12 12/13 13/14 14/15 15/16 16/17 17/18 18/19 Probation & MS Three-Year Revocation Rates, by Population Type 38% 5 650 3 792 246 Jail bookings for probation violations have increased over time, however the average length of stay in jail has decreased. 40% 3 431 4 183 Parole & PRCS Average Days in Jail by Violation Type 51 56 47 35 37 18% 27 28 30 30 29 36 24 31 0% PRCS County Realigned Cohort 1 (N=8, 537) Felony MS Overall Cohort 2, 3 Year Period (N=4, 684) FY FY 11/12 12/13 13/14 14/15 15/16 16/17 17/18 18/19 Probation & MS Parole & PRCS

Recidivism Outcomes 18 Key Findings One, Two, and Three-Year Recidivism Rates 38% of individuals who started probation between 2015 and 2018 recidivated, compared to 40% who started between October 2011 and December 2014 Between 2015 and 2018, oneyear recidivism rates dropped from 21% to 14%. One-Year Recidivism Rates Between 2015 and 2018

19 Recidivism Outcomes Continued Key Findings Individuals with PRCS cases had higher recidivism rates than other probation types, as did individuals with more extensive and serious criminal histories and those who were younger. Men and women had similar recidivism rates.

Service Receipt and Recidivism 20 Descriptive Key Findings Overall, recidivism rates are similar for those who did and did not receive services. 21% of women who received services recidivated compared to 41% of men. Recidivism Outcomes for Individuals who Received AB 109 -Funded Services

Service Receipt and Recidivism 21 Inferential Analysis Key Findings All else equal, those who received services were 29% less likely to recidivate than those who did not. The predicted probability of recidivism for those receiving services was 32%, compared to 40% for those who did not receive services. All else equal, individuals who received any type of service recidivated at a slower rate than those who did not receive services, and overall had a ~23% lower risk of recidivating than individuals who did not receive services. Predicted Probability of Recidivating 40% 32% Received Services Risk of Recidivating Did Not Receive Services

Contact Us 22 David Onek, JD Project Director donek@resourcedevelopmen t. net Ardavan Davaran, Ph. D Project Manager adavaran@resourcedevelopmen t. net Nicole Liner-Jigamian, MSW MPH Research Associate nliner@resourcedevelopment. n et