REINVENTING COMMUNITY SUPERVISION: Georgia Parole's Results-Driven Supervision.

The Reinventing Probation Council, a group of 12 distinguished professionals in corrections and correctional research, published a monograph in 1999 examining the state of probation and parole in America. According to Corbett et al., the council's assessment was clear in statements such as "probation isn't working," "totally under-researched" and "all too often, probation departments have employed practices that simply do not do the job."

That same year, having come to the same conclusion, the Georgia State Board of Pardons and Paroles embarked on a new path for parole supervision. For at least 30 years, Georgia had experienced a three-year offender return-to-prison rate of 40 percent. Moreover, revoked probationers and parolees were becoming an increasing portion of new prison admissions.

Georgia adopted a new model for supervision based on proven theories supported by solid research. On this foundation of theory and research, parole officials adopted policies and implemented practices that focus on results as the measure of success. This new model, Results-Driven Supervision (RDS), was proposed and developed by a group of the agency's field managers who used as one of their guides, Results-Driven Management: Implementing Performance-Based Measures in Community Corrections, published by the American Probation and Parole Association.

The Georgia parole board has sole authority for making discretionary release decisions affecting more than 17,000 offenders admitted to prison each year in a system of 45,000 state inmates. Through 60 field offices, the agency is directly responsible for the reintegration of 21,000 adult offenders under parole supervision.

Integrating Theory, Research, Policy, Practice and Feedback

RDS is a wholesale redesign of traditional supervision integrated with theory, research, public policy, practice and feedback. Carefully examining the building blocks of any successful program, a linear process was employed that first analyzed theories of crime causation. Three theories were selected to build the model: social learning, deterrence and routine opportunities. Next, a thorough examination of correctional research was completed. The validity and successful application of each theory was documented. Field managers and line officers explored the activities that make up the real-world application of the theories.

Building on research and practice, public policy was developed detailing expected, measurable outcomes, which, today are reported to the state Legislature in the agency's budget document. Next, RDS practices were implemented throughout the state. Finally, a mechanism was created to provide continuous feedback to help officers and managers focus on the supervision activities that are most likely to change behavior. This feedback loop ensures that the outcomes are meeting or exceeding expectations.

The RDS model requires parolees to be assessed for criminogenic traits, which research has demonstrated can be altered through appropriate intervention(s). RDS targets those offenders, matching them with the intervention programs most highly correlated with reductions in recidivism. The goal is to divert offenders from criminal lifestyles and increase the likelihood that they will become self-sufficient, productive members of society. Intervention efforts are broadly classified into four categories, or "tracks": substance abuse, education, cognitive traits and employment. Depending on the identified needs, an offender could be placed on one or all four tracks for intervention.

Local Development/Delivery of Programs

All the agency's field offices are required to identify or develop local intervention programming. This often requires building partnerships with both public and private service providers. In many geographic areas where intervention programs are not available, parole officers are trained to deliver the intervention programming. An example of this is cognitive skills training. In addition to their regular responsibilities, trained parole officers facilitate semiweekly cognitive skills classes. Many offices provide space for Alcoholics Anonymous and Narcotics Anonymous meetings, which are facilitated by outside volunteers. Other field offices have partnered with local school systems to provide certified instructors for in-house general educational development classes.

Data-Driven Processes And Outcomes

Parole officers are responsible for familiarizing parolees with supervision requirements, ensuring completion of needs assessments and implementation of intervention plans. Perhaps the most significant change RDS has brought to parole supervision in Georgia is in how parole officers measure their own productivity, from a contact-driven surveillance model to a results/outcome-focused treatment and surveillance combination. Rather than having to make an arbitrary number of contacts, the parole officer is required to focus on the activities that support the three theories underlying RDS.

Instead of counting the number of times a parolee was seen by an officer or how many reports were submitted in a particular month, reports to RDS focus on whether offenders with identified criminogenic needs are enrolled in appropriate programs and if they are attending those programs. RDS also is concerned with offenders being employed. Those who do not have jobs receive immediate attention.

Using deterrence theory, which maintains that the time between a violation and sanction is most important, RDS requires the swift application of sanctions for noncompliance. As a result, RDS focuses not only on testing offenders who have histories of drug use, but also on the immediate application of sanctions when drug use and other violations are detected. RDS also focuses on ensuring that parolees spend the maximum amount of time under some measure of capable guardianship, such as working, attending programs or participating in some other pro-social activity.

In addition to its focus on offender activities, other measures of RDS include interim outcomes that lead to law-abiding citizenship, such as the number of parolees who are employed or who complete counseling programs; have negative drug screen results; and are arrested for new offenses, as well as measures of stability, such as the number of days employed. The final measure of RDS' effectiveness is the impact it has on the parolee rate of return to prison.

Data Collection And Analysis

A statewide, networked computer application called FLOID (Field Log of Interaction Data) was a critical tool necessary for the operation of the RDS model. An entirely new system for documenting and processing field activity was required to manage offenders' participation in programs, to swiftly address noncompliance and to provide managers with timely and accurate information on all supervision processes. Data had to be available to parole officers and managers almost immediately and continuously, rather than in the once-a-month, hand-collated paper reports summarizing numbers of contacts used at the time.

The agency's computer services staff, parole officers, managers and researchers worked as a team to develop FLOID, which is a series of data entry screens built on a Lotus Notes platform. The screens allow parole officers to document every interaction with parolees. The system is designed so that information is entered in data form by employing a predominantly "point and click" method to accommodate the maximum amount of collation, reporting feedback and analysis. The result is data that allow for almost continuous monitoring and reevaluation of offender participation in programs and compliance with conditions of parole.

The basic offender record is created in FLOID automatically from data in the parole board's clemency and the Department of Corrections databases. This includes demographic data such as inmate name, race, sex, date of birth and sentencing information. FLOID entry screens used by parole officers include a number of check boxes and drop-down lists as well as a section in each interaction for open text comments. Intervention tracks and three objectives under each track are listed, which are established at a statewide level. Local parole managers design the strategies that will be employed to meet each objective.

Although a statewide initiative, RDS recognizes the importance of local control over parolee supervision. Officers in Atlanta know the resources available for supervising in the city since they best understand the community. Similarly, officers in Savannah know their resources best. Statewide agency objectives are clearly stated, but field managers are responsible at the local level for devising the methods for accomplishing the objectives. Along with local strategies, each parole office submits a list of the vendors and specific programs that have been approved for referrals. Vendors are added to the FLOID system to facilitate later evaluation of program effectiveness.

Each parole officer carries a laptop computer into the field, replacing cumbersome three-ring binders, full of useful information that was virtually worthless in paper form. The Lotus Notes computer program enables officers to carry a local "replica" of their caseloads on their laptops, allowing them to look up or add new information to individual cases at any time, whether in the field or in the office. New interactions can be entered as they occur. Later, when the officers dial in from home or the office to the server in Atlanta, they automatically transfer or "replicate" information to the server, and information on the server, including any new cases assigned to officers, is copied to their laptops.

While officers are in the field, managers or any other agency employees can review any case. While connected to the server, officers can see FLOID information on any parolee under supervision anywhere in the state. If a parolee reports to the office while his or her officer is unavailable, another staff member can see all the notes and enter new interactions, which are "stamped" with the time, date and name of the employee who made the entry. (For a more technical discussion of the electronic case management system, see The Journal of Offender Monitoring, Winter, 1999.)

Officers are required to replicate any new interactions to the statewide server every day. Each night, the statewide version of FLOID updates a database in Oracle. Taking advantage of the huge amount of information in data form, the agency created a reporting tool called Thelmalou. Agency programmers have created and entered into Thelmalou hundreds of reports to track processes and measure results and outcomes. Using an Internet browser, parole officers and all levels of management can access reports via Thelmalou to help them manage their work and determine their success at meeting goals.

Most reports can be run at four levels: an individual officer's caseload, office, region and statewide summaries. The data in reports are no more than two days old. Reports include: the parolee employment rate; drug testing information; program participation; unemployed parolees and their length of unemployment; parolees who have committed violations and the length of time between violations and sanctions; and parolees with violations and no sanctions.

Determining What Works

For the first time, the agency is determining which interventions are effective and which offenders are most responsive to different treatment approaches. Just as important, the agency is finding out which interventions are ineffective to avoid spending time on activities that produce no measurable results or that provide unacceptable outcomes. RDS is a dynamic rather than static model of supervision. The more information and data fed into the model, the better the model is able to refine effective supervision techniques and intervention activities.

The greatest obstacle the parole board's managers have had to overcome from the full implementation of RDS has been harnessing and redirecting organizational inertia. Moving philosophically and organizationally away from a traditional surveillance approach that was used for more than 50 years has been, and continues to be, a monumental and courageous undertaking.

Data integrity, accuracy and completeness make computer training critical. The initial version of FLOID included two open text fields for entering money order numbers and fee payments. The first reports summarizing fee amounts found that 10-digit money order numbers had been inadvertently entered into several amount fields, resulting in collection totals in the billions of dollars instead of amounts in the ten-thousands. Subsequent updates to FLOID include edit checks that limit what can be entered with pick lists in drop-down boxes or require certain fields to be filled in before the screen can be closed.

FLOID also has been a tremendous help and teaching tool for parole officers. The screen that documents the officers' initial interviews with parolees lists all the items officers must review with parolees, including special conditions of parole such as those applying to substance abusers and other special cases. In the rush of the day, officers can miss some of the initial steps.

Prior to implementing RDS, the board's field managers had been provided with little in the way of statistical reports. Under RDS, myriad data are available and creating statistical reports is a fairly streamlined process. The agency's field managers had to actually see some reports to fully understand what was possible. Reports they thought they needed were found not to be that useful. The process that has emerged for developing the best reports includes discussion with managers about what is important, coupled with analysis of results-driven activities.

National RDS Recognition

The Georgia parole board has been pleased with reactions both within the state and nationally from public officials to its RDS model. In 1999, the agency was selected by the Council of State Governments for its Innovations in State Government Award. At the American Probation and Parole Association's 2000 Training Conference, the Georgia parole board received the Presidents Award for RDS. Finally, also during the summer of 2000, the board's Field Division was reaccredited by the American Correctional Association's Commission on Accreditation for Corrections after meeting 99.5 percent of the reaccreditation standards.

The Georgia parole board believes that RDS embodies the vision of the Reinventing Probation Council. It attempts to provide a full range of sanctions and services that offer the best hope for successfully reintegrating offenders into society. Parole officers activities are directed toward programs that change offender behavior, maintaining the maximum amount of capable guardianship and swiftly addressing any violations.

Primary prevention services have been developed from the bottom up, initiated at the line officer and line supervisor levels and eventually approved by senior management. Those closest to offenders, the parole officers, are empowered to set the standard for supervision in Georgia. By relying heavily on local resources, communities are empowered to participate in solutions.

Finally, RDS is entirely outcome-dependent at a time when the public is demanding accountability for its tax dollars. The Georgia parole board believes that combining results-based supervision strategies, built on proven theory, with the latest information technology, represents the best model for the new millennium.

James Bralley is director of the Field Division and John Prevost is director of the Ofice of Criminal Justice Research for the Georgia State Board of Pardons and Paroles.

REFERENCES

Corbett, Ronald P. Jr. et al. 1999. "Broken windows" probation: The next step in fighting crime. New York: Center for Civic Innovation at the Manhattan Institute.