Chapter 2: Charlottesville/Albemarle County, Virginia

BACKGROUND

Charlottesville and Albemarle County are located at the foothills of the Blue Ridge Mountains, in the Commonwealth of Virginia. Charlottesville is an independent city adjacent to, but separate from, Albemarle County. As of 2015, the city proper had a population of 46,597 (2016 U.S. Census Bureau) and the county’s population was 105,703 (2016 U.S. Census Bureau). The area is home to approximately 20,000 students attending the University of Virginia. The county’s population is 82.2% white, 9.2% black or African American, 4.9% Asian, and 5.8% Hispanic or Latino (2016 U.S. Census Bureau); the city’s population is 70.2% white, 19.2% black or African American, 7.1% Asian, and 5.0% Hispanic or Latino (2016 U.S. Census Bureau estimate). The median household income is $68,449 in the county (2016 U.S. Census Bureau) and $49,775 in the city (2016 U.S. Census Bureau). The population living in poverty is greater in the city (20.7%; 2016 U.S. Census Bureau) than in the county (9.5%; 2016 U.S. Census Bureau).

The Charlottesville/Albemarle EBDM Policy Team is unique in that its work encompasses two jurisdictions (a city and a county) that routinely share resources and have a long history of collaboration. The local and state probation offices have worked together for many years, particularly around the implementation of evidence-based practices (EBP); both were selected as pilot sites by the Virginia Department of Criminal Justice Services (DCJS) and the Virginia Department of Corrections (VADOC). For Charlottesville/Albemarle, the EBDM initiative provided an opportunity to systematically and collaboratively expand EBP across the justice system decision points; its selection as a pilot site was based in large part on the high level of commitment to EBP and justice system improvements by both city and county officials.

Profile of Charlottesville/Albemarle County’s Justice System	2015
Jail Rated Capacity	329
Jail Bookings	4,097
Jail Average Daily Population	455
Felony Court Filings – Charlottesville General District Court	589
Misdemeanor Court Filings – Charlottesville General District Court	1,487
Felony Court Filings – Albemarle County General District Court	514
Misdemeanor Court Filings – Albemarle County General District Court	1,813
State Probation Admissions – District 9	535
Local Probation Admissions – OAR/Local Probation	1,105

CHARLOTTESVILLE/ALBEMARLE COUNTY’S EBDM MISSION AND TEAM

The Charlottesville/Albemarle County EBDM Policy Team developed the following mission statement: “The agencies in the Charlottesville/Albemarle criminal justice system seek to achieve justice and make communities safer by working closely together, applying the best-known research to policies and practices, listening to those affected by crime, and recognizing that every interaction can lead to improved outcomes.”

Due to the policy team’s size (i.e., having representatives from two localities for each discipline group), a steering committee was formed to guide the team’s activities. The steering committee, with representation from both the city and county, is responsible for managing day-to-day planning and implementation activities, and for bringing work products to the full policy team for consensus and final approval. The policy team includes:

a general district court judge;
the chief magistrate;
city and county Commonwealth attorneys;
a county public defender;
city and county police chiefs;
city and county sheriffs;
city and county victim witness coordinators;
the chief probation officer (state level);
the director of pretrial services and community corrections (local level);
the regional jail superintendent;
the community services board director; and
the Thomas Jefferson Area criminal justice planner.

The criminal justice planner, employed by the OAR, convenes the EBDM policy team on a monthly basis, provides the policy team with analysis of local criminal justice data, and coordinates the work of a number of work groups. The CCJB provides support to and oversight of the EBDM policy team and its steering committee.

CHARLOTTESVILLE/ALBEMARLE COUNTY’S CHANGE TARGETS

In addition to the change efforts highlighted in this case study, the Charlottesville/Albemarle County EBDM Policy Team also:

conducted a review of their domestic violence policies and practices and worked to align them with the research-based “Blueprint Model” for preventing domestic violence homicides;

examined court processes for potential streamlining;

conducted a gap analysis and fidelity review to determine the degree of alignment between available services and EBP; and

expanded its partnership with the University of Virginia to assess and improve automated data collection processes.

Charlottesville/Albemarle joined the EBDM initiative in 2010. Guided by the vision statement “Working together for a safer community, one person at a time,” the EBDM policy team developed a system map, logic model, scorecard, and harm reduction goals. They also identified change targets, developed an implementation plan to address each target, and created a plan to build awareness about their work among their local colleagues. The team developed three work groups that focused on 1) arrest, plea, and trial, 2) sentencing, violations, and supervision, and 3) institutional and community interventions.

This case study offers a summary of the team’s Phase II/III efforts related to:

use of actuarial tools to identify recidivism risk and criminogenic needs and enhance jail programming; and
efficient and effective responses to violations of probation.

For more information on these change targets, see the Charlottesville/Albemarle County Phase III implementation plan.

What they did: ImplementED an actuarial assessment

Despite significant crowding in the regional jail (as of 2011, the jail was at 165% of its rated capacity), Charlottesville/Albemarle County officials agreed that expansion was not cost-effective, nor was it an effective way to reduce recidivism rates—a priority for local justice officials. As part of the EBDM work, NIC supported an assessment of the regional jail’s programs. While much of the programming offered was determined to be evidence-based and consistent with programming provided by local and state probation, some improvements were recommended, such as guiding inmates to programs matching their criminogenic needs. To address this, the team agreed to implement an evidence-based screening tool at jail intake (Northpointe’s COMPAS).

Beginning in October 2013, the jail began using the COMPAS screener for all inmates held long enough to classify (i.e., at least 24 hours). A full assessment—which includes dynamic risk factors—is completed for those assessed as medium or high risk on the screening tool who also have a length of stay sufficient to at least begin programming. Inmates with short stays typically begin programming while incarcerated and then continue the same programming when released to community supervision or transferred to the DOC; those with longer sentences have an opportunity to complete programming while incarcerated.

RESULTS

Data on inmate risk and needs, as assessed by the COMPAS, were collected and analyzed to inform policy decisions about inmate programming. From 2013 to 2016, 5,480 screenings of classified inmates were conducted, and 1,081 full risk/needs assessments were completed for those assessed as medium or high risk. Regarding risk, it was determined that more than half (51%) of the inmates held long enough to classify were low risk (defined as having an 8–11% likelihood of rearrest over a two-year period following release). This finding was instructive to the team from a jail programming point of view but also important given the extensive discussions and planning around applying the risk principle to key decision points, including the strategic use of jail resources. Future data analyses will focus on understanding the average length of stay by risk level and exploring the circumstances under which low risk offenders are sentenced to jail.

Analysis of these data also provided information about the criminogenic needs of the medium and high risk population and supported strategic expansion of the programming available within the jail. With support from the U.S Department of Justice, Bureau of Justice Assistance, under a Justice Reinvestment Initiative-Local (JRI) grant, the jail put into place a number of additional evidence-based curricula (Moral Recognition Therapy, the Matrix Model for substance abuse, and Parenting Inside and Out) and certified staff in the delivery of these services.
What they did: developed structured responses to violations
Charlottesville/Albemarle County Probation Decision Making Process and Administrative Response Matrix

The goals of the violation process are to hold offenders accountable while encouraging and reinforcing prosocial behavior by:
- responding to all noncompliant behavior, thereby reinforcing that antisocial behavior has consequences; and
- responding to behavior in the manner most likely to result in behavior change.
The goals of the incentive process are to promote and reinforce prosocial behavior by:
- responding to all prosocial behavior, thereby reinforcing future prosocial behavior; and
- responding to behavior in the manner most likely to result in behavior change.
In order to maximize the effectiveness of the behavior management process, responses to prosocial and noncompliant behavior will be:
- swift;
- certain;
- proportional to an offender’s risk and needs, the severity of the violation behavior, and in consideration of the other pertinent factors;
- transparent;
- consistent with evidence-based practices;
- economically sound; and
- evaluated over time for effectiveness.
During the EBDM planning process, probation violations rose to the surface as a priority for the Charlottesville/Albemarle EBDM Policy Team, given their significant impact on court processing time, jail bed days, and probation officer workload. In 2011, as part of their work under the EBDM initiative, an assessment of both state and local probation violation practices was conducted. The assessment revealed that both policies and practices lacked adherence to empirical research around swiftness, consistency, and proportionality. A subsequent analysis under the auspices of the team’s JRI work revealed important empirical information: According to a dataset of persons released from the jail between January 1, 2008 and December 31, 2011, 10% of the jail population represented probation violators, and their average length of stay was 118 days. Violations were determined to be among the top three drivers of the jail population.

In an effort to align responses to noncompliant behavior with social science research, and to reduce the impact on court processes and lengthy jail/prison stays due to technical violations, a subcommittee of OAR and District 9 Probation and Parole representatives was tasked with revising policies and practices around responses to probationer behavior. The process included defining the goal of the behavior management system (“to promote behavioral changes by empowering probation officers to implement consistent, swift, and responsive practices when addressing behavior”) and creating a policy framework to guide the work. An instruction guide was developed for use by officers which explains the goals, policies, and steps of the new probation decision making process. The guide contains a list of both noncompliant (technical and non-technical) and prosocial behaviors and guidelines for staff regarding identification of an appropriate response(s), taking into consideration offender risk, needs, and violation severity.

With support from JRI, the team automated their structured response matrix and developed a web-based case management information system. The new system, called the Administrative Response Matrix (ARM), enables probation staff to similarly apply objective criteria to noncompliant and prosocial behavior. The software takes into account risk of reoffense and violation severity and provides officers a menu of responses proportional to the behavior from which to choose.

RESULTS

The new process for behavioral responses was piloted from April through July 2014 and was followed by full-scale implementation immediately thereafter. Data collected from FY 2013 through FY 2016 (i.e., approximately two years prior to implementation and two years following) on the misdemeanor probation population supervised by OAR show modest increases in successful closings (success rates rose from 73.2% in FY 2013/14 to 77.6% in FY 2015/16) and decreases in unsuccessful closings (rates of unsuccessful closings dropped from 26.8% to 22.4%). Of these unsuccessful terminations, fewer probationers were revoked for technical violations following implementation of the matrix (20.3% prior to implementation vs. 13.6% following implementation). In another analysis, the average length of stay for probation violators dropped significantly (by 61%). However, this analysis constituted a relatively small number over a short period of time and therefore should be viewed cautiously. There was also a modest increase in the percentage of unsuccessful terminations due to new misdemeanor and felony convictions; however, these closings made up a very small percentage of all closings (2.4% in FY 2013/14 and 4.7% in FY 2015/16).

Data on revocations of state probation cases out of the Albemarle and Charlottesville Circuit Courts (largely felonies) show a modest (6.5%) decrease in revocations (from 413 in FY 2013/14 to 386 in FY 2015/16). In FY 2015/16, approximately 50% of revocations were for technical violations only, 13% were for new convictions, and 22% were for both new convictions and technical violations. Detailed data on successful and unsuccessful case closures is not available given the complexities in tracking this data by DOC.

OAR and District 9 also collect fidelity measures to determine the extent to which officers are using the ARM as intended. For example, in FY 2015 the average length of time between the officer’s knowledge of a violation and their response was 4 days for OAR officers and 7 days for District 9 officers. Overrides of the recommended response were made in about 3% of cases by OAR and in just over 9% of cases by District 9. Charlottesville/Albemarle County officials continue to collect and monitor data to see if these trends are sustained over time.

More recent data (FY 2016) suggest that some of the modest gains in successful closings and reduced jail bed use may have faded, returning to levels such as those prior to the implementation of ARM. However, these findings are considered preliminary and must be placed within the context of broader policy changes. For example, the FY 2016 data may be attributed, at least in part, to a significant DOC policy change in FY 2015/16 that resulted in a doubling of DOC inmates serving their sentences locally (sentences of two years or less), including felony probation violators. In addition, the FY 2016 probation violator population in the jail had a higher risk profile than the baseline (pre-ARM) population as a result of the transition of approximately 300 low risk probationers from active supervision to a state DOC-controlled administrative caseload during the time of ARM implementation.

MOVING FORWARD

Building on their EBDM efforts, the Charlottesville/Albemarle Policy Team sought funding to support ongoing data collection and evaluation. They were awarded a grant from the National Criminal Justice Association as part of the Justice Information Sharing Initiative to build and pilot an integrated database to share information across their criminal justice agencies. The database captures data about incidents, offenders, and offenses across the entire criminal justice system, while complying with national justice information-sharing standards and maintaining individual stakeholders’ legacy data systems.

The Charlottesville/Albemarle County EBDM Policy Team continues with their work of actively implementing the EBDM principles and serves as a model for communities across Virginia. For instance, as a result of their work on ARM, other localities in the Commonwealth have expressed interest in replicating the ARM. In addition, Charlottesville/Albemarle County’s EBDM policy team members have provided multiple presentations on EBDM to state officials and other local stakeholders and at professional association meetings throughout the Commonwealth. Policy team members met with state-level officials to encourage Virginia’s participation in Phase IV and provided valuable assistance in the development of the Commonwealth’s Phase IV, V, and VI applications. Some of its representatives also serve as capacity builders to support expansion of EBDM to other localities in Virginia. Two Charlottesville/Albemarle policy team members serve on the Virginia State EBDM Policy Team and have taken a leadership role in working on two of the state team’s change targets.