Co Parole Board Audit 2008
Download original document:
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
REPORT OF THE STATE AUDITOR Discretionary Parole State Board of Parole Performance Audit November 2008 LEGISLATIVE AUDIT COMMITTEE 2008 MEMBERS Representative James Kerr Chair Representative Dianne Primavera Vice-Chair Senator Jim Isgar Representative Rosemary Marshall Representative Frank McNulty Senator David Schultheis Senator Jack Taylor Senator Lois Tochtrop Office of the State Auditor Staff Sally Symanski State Auditor Cindi Stetson Deputy State Auditor Jennifer Harmon Julie Chickillo Kate Shiroff Trey Standley Legislative Auditors STATE OF COLORADO OFFICE OF THE STATE AUDITOR 303.869.2800 FAX 303.869.3060 Sally Symanski, CPA State Auditor Legislative Services Building 200 East 14th Avenue Denver, Colorado 80203-2211 November 24, 2008 Members of the Legislative Audit Committee: This report contains the results of a performance audit of the discretionary parole process administered by the State Board of Parole. The audit was conducted pursuant to Section 2-3-103, C.R.S., which authorizes the State Auditor to conduct audits of all departments, institutions, and agencies of state government. The report presents our findings, conclusions, and recommendations, and the responses of the State Board of Parole, Department of Corrections, and Division of Criminal Justice at the Department of Public Safety. This page intentionally left blank. 1 Discretionary Parole Authority, Purpose, and Scope This performance audit was conducted in response to a legislative request for a review of statistics and reports related to mandatory and discretionary parole releases. Specifically, data reported by the Department of Corrections (Department) in December 2007 indicated that discretionary parole releases had increased by 80 percent from Fiscal Year 2006 to 2007. The Office of the State Auditor was asked to determine the reasons for the significant increase in discretionary parole releases, analyze changes in discretionary parole figures, and evaluate whether consistent and effective systems exist to capture and compare data on discretionary and mandatory parole releases. The audit evaluated significant trends in parole and the accuracy, reliability, and usefulness of parole data currently available and reported by the Department and State Board of Parole (Board). Additionally, our audit evaluated the Board’s practices for reviewing its parole decision-making process and the outcomes of its decisions. This audit was conducted under the authority of Section 2-3-103, C.R.S., which authorizes the State Auditor to conduct audits of all departments, institutions, and agencies of state government. The audit was conducted in accordance with generally accepted government auditing standards and performed from July through September 2008. As part of our audit we interviewed staff at the Department, Board, and the Division of Criminal Justice (Division) at the Department of Public Safety, attended Board hearings, surveyed parole boards in other states, and reviewed research on best practices for assessing and evaluating parole board decisions. We acknowledge the assistance and cooperation provided by the Department, Board, and Division. Overall we found that no single source of parole data exists to provide decisionmakers with the information needed to develop and evaluate parole policies in the State. Data maintained by the Board need to be enhanced to provide more comprehensive information to measure outcomes and support parole decisions. Data maintained by the Department need to be recorded accurately to ensure consistent reporting and analysis over time. Differences in the data reported by the Board and the Department need to be explained so that the different perspectives of the Board and Department are understood and erroneous conclusions do not result. We address these issues later in this report. 2 Discretionary Parole Performance Audit - November 2008 Despite the problems we found with existing parole data, we were able to further analyze the data and report certain basic parole trends that describe Colorado’s parole practices and which may facilitate decision-making, as follows: • The total number of paroles granted (mandatory and discretionary) by the Board in Colorado increased by about 34 percent from Fiscal Year 2004 through Fiscal Year 2008. Mandatory paroles increased by 30 percent, and discretionary paroles increased by 41 percent during this period. • The one-year increase in the number of discretionary paroles granted by the Board was about 35 percent, from 2,000 in Fiscal Year 2006 to 2,700 in Fiscal Year 2007. This contrasts with the 80 percent increase in discretionary parole releases reported by the Department and referenced earlier in this report. • The majority of paroles in Colorado are mandatory, and the Board does not have the discretion (or statutory authority) to deny them. In Fiscal Year 2008 the Board granted a total of 7,100 paroles. More than 60 percent (4,300) of these were mandatory paroles and almost 40 percent (2,800) were discretionary paroles. The percentage of mandatory and discretionary paroles has been fairly stable from Fiscal Year 2004 through Fiscal Year 2008. • More than three-quarters of discretionary parole requests are denied by the Board. For example, in Fiscal Year 2008 the Board denied 15,000, or 84 percent, of the 17,800 requests for discretionary parole. In the following section we provide additional detail regarding each of these points, as well as other descriptive information and statistics about mandatory and discretionary parole. In the second section of the report we present our findings related to improving the accuracy and usefulness of parole data. Description of the Parole System In the criminal justice system, parole generally refers to the supervised release of an inmate from incarceration before the completion of his or her sentence. Technically, offenders are still serving their sentence while on parole. One purpose of parole is to provide offenders a period of transition back into the community while still under the supervision of the correctional system. It is important to note that parole differs from probation in that probation is a sentencing option imposed by a judge that does not place an offender in a correctional facility. According to Colorado’s current sentencing laws [Section 18-1.3-401(1)(a)(V)(A) and (B), C.R.S.], all offenders who are convicted after July 1, 1993, and sentenced Report of The Colorado State Auditor 3 to a term of incarceration, other than offenders who have been convicted of a Class 1 felony or who are on death row, must serve both a prison sentence and a required period of parole prior to discharge from the correctional system. In other words, an offender’s sentence includes both periods of incarceration and parole. As a result, the majority (97 percent) of offenders released in Colorado as of Fiscal Year 2007 were subject to a law requiring parole after release from prison. The remaining 3 percent of offenders were sentenced prior to 1993 and therefore, were not subject to a required period of parole. In Colorado there are two categories of parole: mandatory and discretionary. Generally, mandatory parole refers to the release of an offender on the latest possible date under the offender’s sentence that he or she can be paroled. Discretionary parole refers to the early release of an offender (i.e., anytime between his or her parole eligibility date and mandatory parole date). Most offenders are eligible for discretionary parole once they have served one-half of their sentence less earned time received. An offender receives “earned time” for acceptable behavior while incarcerated. In general, the earliest an offender can be eligible for parole is after serving 37.5 percent of his or her sentence. Certain violent offenders, defined in statute, must serve 75 percent of their sentence, less earned time, before becoming eligible for parole. Despite the possibility of discretionary parole, 65 percent of offenders are not released to parole until their mandatory parole date. That is, most offenders are incarcerated for the maximum period allowable. State Board of Parole The State Board of Parole is a Type 1 agency within the Department of Corrections and is charged with conducting parole hearings for every offender applying for parole. As a Type 1 agency, the Board operates independently of the Department. The Department has no authority over Board decisions, policies, or procedures. The Board consists of a full-time Chair as well as six full-time members, all of whom are salaried and appointed by the Governor to serve three-year terms. By statute [Section 17-2-201(1)(a), C.R.S.] all Board members must have knowledge of parole, rehabilitation, correctional administration, the criminal justice system, and victims’ issues. Two members must have law enforcement backgrounds, one member must have been a parole or probation officer, and four members are considered citizen members with no specific background required. The Board Chair is the administrative head of the Board and is responsible for establishing procedures for Board hearings and policies for the conduct of Board members. Most parole hearings are conducted by a single Board member, either at the correctional facility where the offender is located or through a teleconference. At the hearing, a case manager provides the Board member with information on the parole applicant, such as criminal history, substance abuse history and treatment, behavior 4 Discretionary Parole Performance Audit - November 2008 while incarcerated, and vocational abilities. The Board member then conducts an interview with the offender and gives the offender an opportunity to explain his or her progress toward rehabilitation and parole plans following release. As part of the hearing, the Board member must allow the victim or the victim’s representative—and may also allow the offender’s family—to make statements. Although there is no time limit, the Board Chair estimates that each hearing takes an average of 20 minutes. In Fiscal Year 2008, the Board conducted a total of about 17,800 parole (mandatory and discretionary) application hearings. While individual caseloads vary, some Board members report conducting as many as 40 hearings during a day. Parole Decisions By statute [Section 17-22.5-404, C.R.S.], Board members must consider various mitigating and aggravating factors in making their parole decisions. These factors include criminal history, behavior while incarcerated, and statistical risk assessment scores. Although statute mandates these factors be considered, it does not specify the weight that must be placed on each factor or limit the Board from considering additional factors. Each Board member works autonomously and has discretion, with second signature approval by another Board member, to grant or deny discretionary parole. The Board documents its overall decision to grant or deny parole, but does not fully document the rationale for its decisions. Parole decisions may take any of the following forms: • Deny Parole. If the Board member denies parole, statute requires in most cases that the Board member set a date for a new hearing within one year. • Grant Discretionary Parole. The offender is granted parole on a date set by the Board member prior to the offender’s mandatory release date. • Grant Discretionary Parole with Conditions. The offender is granted a discretionary parole subject to conditions, such as completion of a substance abuse program or vocational training prior to release. If the offender does not meet these conditions, he or she will not be paroled. • Mandatory Parole Release. mandatory parole date. • Refer to Full Board. If a Board member recommends granting discretionary parole to a violent offender, Board policy requires that the case must be referred to the full Board for its review and approval before the offender can be released. Board members, at their discretion, may also refer non-violent The offender is paroled on his or her Report of The Colorado State Auditor 5 offenders for a full Board decision. At the full Board hearing, at least four members of the Board must approve the decision. In granting parole, the Board member also sets conditions an offender must follow upon release to parole, such as regular drug and alcohol testing and avoiding contact with the victim. With the exception of full Board hearings, as mentioned previously, a Board member’s decision receives final approval through review and signature by a second Board member. The Board does not maintain data on the number of times Board member’s decisions are not approved by a second signature; however, Board members indicated it is rare that another Board member fails to approve the decision. After being paroled, if an offender does not follow the conditions of the parole or is arrested for a new offense, the Board may revoke the parole. A revocation hearing is conducted by a single Board member, and if revocation is necessary, the member will revoke parole for a set period of time. Statutes [Sections 17-2-103(11) and 1722.5-403(9), C.R.S.] provide the maximum time limits for parole revocation and require that certain offenders must periodically receive new parole hearings during the revocation period. The Board is entirely funded through general funds. During Fiscal Year 2008 the Board was appropriated and expended about $1.6 million in general funds. Additionally, during this same fiscal year the Board was appropriated 17.5 FTE, which includes the Board members and administrative staff. The Board Chair is authorized to contract with qualified individuals to serve as hearing officers. These individuals conduct hearings in the same manner as the Board members but may only conduct hearings for offenders convicted of less serious Classes 4 through 6 felonies. According to the Board Chair, the use of contract hearing officers is limited to periods when the caseload exceeds the amount Board members can reasonably accomplish within a required timeframe. Other Agencies Involved with the Parole System Statutes, including Section 17-22.5-404(6), C.R.S., require multiple entities, such as the Board, Department of Corrections, and the Division of Criminal Justice to jointly administer the parole system and parole guidelines. While the Board has sole authority to hear and decide applications for parole, the Department and Division have important roles in administering and evaluating the parole system and providing Board members, policymakers, and other state agencies with information necessary to make informed decisions about parole policy and the correctional system. 6 Discretionary Parole Performance Audit - November 2008 Department of Corrections The Department manages 22 state and oversees 6 private correctional facilities across the State. As of September 2008, it supervised an offender population of over 34,000 individuals, including incarcerated adult offenders, offenders in the youthful offender system, and offenders who have been released on parole. The Department is responsible for administering each phase of the parole process. Once an offender is admitted to a correctional facility, the Department’s Time and Release Unit tracks the offender’s sentence to determine a parole eligibility date and notifies the Board when the offender is eligible for a hearing. At the parole hearing, Department case managers prepare and present information about the offender. If an offender is granted parole, the Department is responsible for releasing the offender in accordance with the Board’s decision and ensuring that any conditions of release are met. After an offender is released to parole, the Department’s Division of Adult Parole is responsible for assigning a parole officer and supervising the parolee. The Department relies on its central database, the Department of Corrections Information System (DCIS), to monitor and manage the offender population and provide information for parole hearings. This database is the central repository for all information about offenders and is designed to provide current, “point in time” information. Department staff use the database for a variety of purposes, including determining parole eligibility, monitoring release dates, and preparing case files for parole hearings. Additionally, the Department uses data in DCIS to regularly prepare and report statistical information on the correctional system and prison population, including parole. This information is made available to the general public on the Department’s website and is widely distributed to and used by agencies such as the Division, Office of Legislative Council, and General Assembly. The Division of Criminal Justice Statute [Section 24-33.5-503, C.R.S.] requires the Department of Public Safety’s Division of Criminal Justice to work “in cooperation with other agencies, to collect and disseminate information concerning crime and criminal justice for the purpose of assisting the general assembly and of enhancing the quality of criminal justice at all levels of government in this state.” The Division administers a statistical analysis center that conducts research and presents numerous reports on a variety of subjects, including effective recidivism reduction programs, population projections, and standards for the supervision of sex offenders. The Division also creates statistical risk assessment tools to assist the Board in making parole decisions and evaluating the risk posed by sex offenders. Report of The Colorado State Auditor 7 Parole Trends We were asked to report on changes in parole trends, particularly with respect to discretionary parole. In general, over the past five Fiscal Years (2004 through 2008), the majority (65 percent) of paroles granted were mandatory; that is, the Board was required to release the offender because the offender was at the end of his or her prison sentence. During this period an average of 35 percent of paroles were discretionary. For discretionary paroles, the Board releases offenders before their mandatory parole dates based on the Board's assessment of the offenders' readiness for parole. The following table shows mandatory and discretionary parole statistics for Fiscal Years 2004 through 2008. State Board of Parole Type of Parole Granted to Offenders1 Fiscal Years 2004 Through 2008 Fiscal Year Total Paroled 2004Offenders Percent Offenders Percent Offenders Percent Offenders Percent Offenders Percent 2008 Paroled of Total Paroled of Total Paroled of Total Paroled of Total Paroled of Total 2004 Type of Parole 2005 2006 2007 2008 Percent of Total Paroled Discretionary 2,000 38% 1,600 28% 2,000 32% 2,700 39% 2,800 39% 11,100 35% Mandatory 3,300 62% 4,100 72% 4,300 68% 4,300 61% 4,300 61% 20,300 65% Total Paroles 5,300 100% 5,700 100% 6,300 100% 7,000 100% 7,100 100% 31,400 100% Source: Office of the State Auditor’s analysis of State Board of Parole data. 1 Does not include re-paroles (offenders who have been released again on parole after having their prior parole revoked). The table shows the number of both mandatory and discretionary paroles has increased over the five-year period; however, as noted previously, the percentage of mandatory paroles compared with discretionary paroles has remained relatively constant over this time period. As discussed earlier, the Board hears a large volume of discretionary parole requests annually, and the Board denies most of these requests. The following table shows the number and percentage of discretionary parole applications granted or denied for the past five years. 8 Discretionary Parole Performance Audit - November 2008 State Board of Parole Offenders Granted or Denied Discretionary Parole1 Fiscal Years 2004 Through 2008 Fiscal Year Total Denied/ Percent Percent Percent Percent Percent Percent Granted of Total Denied/ 2004of of of of of Granted 2008 Offenders Total Offenders Total Offenders Total Offenders Total Offenders Total 2004 Discretionary Parole 2005 2006 2007 2008 Denied 13,700 87% 14,100 90% 14,100 88% 15,300 85% 15,000 84% 72,200 87% Granted 2,000 13% 1,600 10% 2,000 12% 2,700 15% 2,800 16% 11,100 13% Total Hearings 15,700 100% 15,700 100% 16,100 100% 18,000 100% 17,800 100% 83,300 100% Source: Office of the State Auditor’s analysis of State Board of Parole data. 1 Does not include re-paroles (offenders who have been released again on parole after having their prior parole revoked). The table shows that, on average, the Board denied about 87 percent of discretionary parole requests between Fiscal Years 2004 and 2008. The percentage of denied requests decreased from 87 to 84 percent, while the percentage of granted requests increased from 13 to 16 percent during the same period. External Factors There are a number of external factors that influence parole statistics at both state and national levels. Although no single factor can fully explain parole trends, we identified some significant changes that have occurred in the criminal justice system over the past several years that impact changes in the number of mandatory and discretionary paroles. These factors can affect the number of offenders eligible for parole at any given point in time: State Prison Population. The number of mandatory and discretionary paroles granted is partially driven by the State’s overall prison population, which has increased substantially in recent years. According to the Department, from June 2004 to September 2008 the inmate population rose from about 19,500 to about 23,100 (18 percent). Parole releases may not necessarily parallel prison population increases; however, an overall increase in the prison population will eventually increase the number of offenders paroled. State Sentencing Laws. A reduction in sentencing length can temporarily increase the number of offenders eligible for parole, and an increase in sentence length may have the opposite effect. For example, Senate Bill 03-318 [Section 18-18-404, C.R.S.] reduced sentences for the use of certain drugs from felony Classes 4 and 5 Report of The Colorado State Auditor 9 to a felony Class 6. This reduction in felony class decreased the sentences for these drug crimes from four years to two years. Thus, offenders sentenced under the new law are eligible for parole earlier than under previous sentencing laws, and between Fiscal Years 2003 and 2007 the total number of Class 6 offenders paroled increased by more than 100 percent. Conversely, House Bill 04-1189, increased the amount of time violent offenders must be incarcerated before becoming eligible for parole. Specifically, certain violent offenders must now serve 75 percent, as opposed to 50 percent, of their sentence before becoming eligible for parole. Since offenders are incarcerated for an average of 29 months, it may take several years for the impact of this sentencing law to affect parole rates. State Inmate Programs. The availability of inmate programs, such as substance abuse treatment and education, may also affect parole decisions. Several Board members indicated that the completion of substance abuse programs is often a significant consideration in their evaluation of an offender’s application for parole and that they are likely to view an offender as a better candidate for parole when the offender has completed such treatment. The Department reports that in Fiscal Year 2007 about 82 percent of inmates sentenced to prison had some substance abuse problems, and 17 percent were severe. According to the Department, it has not always been able to provide offenders with the necessary treatment services due to budget constraints. Parole Trends in Other States As part of our analysis of parole trends, we analyzed national parole research conduced by the U.S. Department of Justice. Nationwide, in recent years more states have adopted mandatory parole to ensure all offenders receive a period of parole supervision following their release from prison. As a result, fewer offenders have been released on discretionary parole, and more offenders have been released on mandatory parole. Specifically, between 1980 and 2000 the percentage of offenders released on mandatory parole more than doubled, while the percentage of those released on discretionary parole fell by almost half. Following this trend, Colorado first adopted mandatory parole in 1979 and experienced an increase in mandatory parole releases. Additionally, we found a wide variation among the states in the amount of discretion vested in parole boards and equivalent agencies. According to U.S. Department of Justice statistics, as of Calendar Year 2000, 16 states, including Kansas, Arizona, and California, had either abolished discretionary parole or limited its use to a small fraction of offenders. In these states, the point at which offenders are eligible for parole is determined by statute; that is, the parole board generally has no discretion over when to release an offender on parole. The remaining 34 states, including Colorado, give their parole boards full discretion to release most offenders on 10 Discretionary Parole Performance Audit - November 2008 discretionary parole, although some states limit the board’s discretion for certain violent offenders. Parole Board Decision Outcomes A 2008 report published by the U.S. Department of Justice’s National Institute of Corrections, Comprehensive Framework for Paroling Authorities, emphasizes the importance of two practices with respect to parole decisions: (1) measuring the outcome of parole decisions to increase accountability and ensure decision integrity, and (2) using evidence-based models to support decision-making. Specifically, the report states that “paroling authorities and supervision agencies must routinely measure intermediate and long-term outcomes . . . .[P]roviding performance feedback increases accountability and ensures integrity with the mission.” It also states that “evidence-based models ensure that all board members are aligned with the mission and goals of the agency and that paroling decisions are based on factors that can be evaluated to determine if they are consistent with the strategies that current research indicates support successful transition and reentry.” The trend toward using evidence-based models for decision-making in the correctional system is evident within many states, including Colorado. For example, the General Assembly established the Colorado Commission on Criminal and Juvenile Justice through House Bill 07-1358 for the sole purpose of engaging in “an evidence-based analysis of the criminal justice system in Colorado . . . .” Moreover, statute [Section 17-2-201(4)(a), C.R.S.] requires that the Board grant discretionary parole to eligible offenders only when “there is a strong and reasonable probability that the person will not thereafter violate the law and that release of such person from institutional custody is compatible with the welfare of society.” Although the Board members’ discretion is central to the parole decision-making process in Colorado, there is also need to ensure that there is a rational and objective basis for those decisions. We reviewed parole decision data collected by the Board and found the Board lacks data to demonstrate assurance that its parole decisions comply with the statutory requirement that eligible offenders only be released when there is a strong possibility the inmate will not reoffend. We found problems in two areas. First, we found that although the Board considers a number of aggravating and mitigating factors during its parole hearings and compiles and reports basic data on the types and numbers of paroles granted, it does not routinely collect data on the outcomes of its decisions, such as whether or not the offender commits another crime. Second, we found that the Board cannot demonstrate that it makes full use of evidence-based research to support its decision-making. Because the Board does not collect and review data related to its decisions including the outcomes of its decisions, or make full use of Report of The Colorado State Auditor 11 evidence-based research, it cannot ensure that offender transitions from incarceration to the community are likely to be successful. Statute [Section 17-22.5-404, C.R.S.] specifically outlines a process for the Parole Board to report parole decisions to the Division of Criminal Justice so that the Division can analyze the decisions and provide feedback to the Board on outcomes. The purpose of this requirement is to identify the factors the Board is using to support its decisions and to determine whether those factors have a research basis. Prior to 1996, the Board and Division followed the statutory process for reporting and analyzing Board decisions. However, we found that as of July 1996, this process was discontinued and neither entity has conducted the procedures or analysis required by statute. It is important that the Board implement and document an evidence-based process for making parole decisions. By keeping high-risk offenders incarcerated and granting low-risk offenders parole when appropriate, the Board can better ensure public safety is adequately protected and resources are efficiently used. The Board and Division need to reinstitute the statutory process to support and improve the Board’s decision-making practices, as described below. First, the Board needs to document its rationale for granting or denying parole and periodically provide these data to the Division for evaluation. As the research arm of the criminal justice system, the Division has statistical evaluation tools that can determine whether the Board’s rationale for granting parole is supported by evidence-based research and whether each of the offenders paroled by the Board successfully reintegrated back into the community. The Division should provide its analysis of parole outcomes and rationale to the Board periodically and the Board should review the analysis and make changes to its decision-making practices as necessary. Second, the Board and Division need to clearly define the terms and measures used in evaluating parole practices. For example, recidivism rates are a frequently used measure of offenders’ success or failure once released to parole. However, the Board and Division do not define this term consistently. For example, the Board generally defines recidivism as an offender’s return to prison for either violating conditions of parole or committing a new crime. In contrast, the Division defines recidivism as an offender’s arrest or charge for a new crime, even if the offender is not convicted and sentenced to a period of incarceration. This issue needs to be resolved to ensure reliable analysis and interpretation of data over time. Third, the Board and Division need to work together to ensure the Board members know how to interpret and use the Division’s analysis of the Board’s decisionmaking rationale. As part of this process, the Board should receive training from the Division on the Colorado Actuarial Risk Assessment Scale (CARAS), which is a tool the Division created for the Board’s use in determining whether an offender has a 12 Discretionary Parole Performance Audit - November 2008 high or low risk of being arrested or committing a new crime after release from prison. Although Board members currently review the CARAS risk rating for each offender, the Board and Division have not worked together to ensure that members are properly trained and understand how the CARAS was developed and can be relied upon when making parole decisions. Fourth, the Board and Division need to evaluate the resources necessary to collect and analyze data related to Board decisions and work with the General Assembly as necessary. This process should include automating processes to improve efficiencies for collecting and reporting parole data. Currently the Board collects data through manual, paper processes. According to the Colorado Government Efficiency and Management Performance Review conducted by the Governor’s Office in May 2007, the Board’s processes could be streamlined through the use of technology. The study recommended purchasing laptops for the Board members and providing software to allow the Board to electronically access data on offenders; generating and modifying hearing forms, including converting “all typewritten and handwritten Parole Board action forms to an electronic format;” and eliminating handwritten paperwork. The General Assembly authorized about $54,000 to the Board for this project in the Fiscal Year 2009 Long Bill. The Board should work with the Department of Corrections and Division to ensure that any new automated system includes data fields and reporting capabilities that allow the Board to capture and transmit data on the rationale for its parole decisions and facilitate analysis of parole outcomes. Finally, the Board and Division should formalize their agreements in writing for collecting and reporting parole data. The agreements should identify the specific responsibilities of each agency in carrying out processes for reporting and analyzing parole decisions, as defined in statute. The agreements, along with the steps previously discussed, are important for providing information to policymakers and helping them make informed decisions about parole. Recently the Governor submitted his Fiscal Year 2010 budget request which includes additional funding for a Crime Reduction and Recidivism Package aimed at improving public safety through evidence-based, cost-effective reforms. The request also includes $7.2 million in state general funds for additional substance abuse treatment services and expanded vocational and academic programs during incarceration. This significant investment of additional funds for recidivism and treatment programs should be accompanied by systematic analysis and assessment of results. By improving its efforts to measure and assess the impact of its parole decisions, the Board can help policymakers determine whether the additional investment in treatment programs is reducing recidivism and improving transitions for offenders. Report of The Colorado State Auditor 13 Recommendation No. 1: The State Board of Parole should work with the Division of Criminal Justice to implement a process for capturing and analyzing the basis for and outcomes of its parole decisions by: a. Tracking data related to the Board’s rationale for granting or denying parole and providing these data to the Division for its review and analysis. The Board should review results of the Division’s analysis and make the necessary changes to its decision-making practices. b. Resolving different practices for reporting and defining terms and measures, such as recidivism, when evaluating parole practices. c. Ensuring the Board receives adequate feedback and training to know how the results of the Division’s analysis, along with the Colorado Actuarial Risk Assessment Scale, can be used to improve the Board’s decision-making practices. d. Determining the resources needed to collect and analyze data related to Board decisions and working with the Department of Corrections and General Assembly as necessary. This should include automating paper processes to improve efficiencies for collecting and reporting parole data. e. Identifying the specific responsibilities of the Board and Division in carrying out processes for reporting and analyzing parole decisions as defined in statute, and formalizing these responsibilities in writing. State Board of Parole Response: Agree. a. Implementation date: July 2009. The Board has entered into preliminary discussion with the Division of Criminal Justice to accomplish this task. The Board has upgraded to a MSACCESS 2000 database for data collection. The Board will work with the Division to identify a process to transmit the necessary data for review and analysis. b. Implementation date: Ongoing. The Board will work with the Division and Department of Corrections to resolve different practices for defining terms. 14 Discretionary Parole Performance Audit - November 2008 c. Implementation date: Ongoing. The Board will schedule time on a regular basis to work with the Division to know how to use the results of the Division’s analysis, including the Colorado Actuarial Risk Assessment Scale. d. Implementation date: Summer 2009. The Business Technology component of the Department of Corrections is working with the Board and the Division to address this issue. This project is centered on an automation of the parole hearing process. Laptop computers needed for the project have been purchased and will be issued to Board members on November 25, 2008. Board members are scheduled for a training session on using the laptops the same day. The first phase of the implementation of the project should begin soon. It will include testing the connection capability of the laptops at the many sites throughout the state where parole hearings are held. e. Implementation date: March 2009. The Board will work with the Division to develop a written agreement by March 1, 2009. Division of Criminal Justice Response: Agree. a. Implementation date: July 2009. Once data are available electronically, we will request quarterly downloads of the Colorado Actuarial Risk Assessment Scale, the Parole Board Action Form, and other information that may be useful to the Parole Board. Upon receipt we will analyze the data and report the information in a quarterly memorandum to Parole Board members. b. Implementation date: Ongoing. We agree to define and use multiple measures of recidivism, such as rearrest, new court filing, new conviction, and for those convicted, their placement (e.g., on probation, in community corrections, or in prison). We will work with the Department of Corrections to obtain return-to-prison outcome data. Using multiple measures of recidivism provides stakeholders with the most comprehensive information. c. Implementation date: Ongoing. We agree to work with the Parole Board to discuss the results of our analysis of the Parole Board’s data and how this information can be used to improve decision-making. Report of The Colorado State Auditor 15 d. Implementation date: Fall 2009. The Division currently has 0.1 FTE to perform the tasks mandated in statute. We agree to request additional resources from the General Assembly through a decision item for the Fiscal Year 2011 budget cycle to ensure our ability to comply with this recommendation and the Division’s mandate. e. Implementation date: March 2009. We will work with the Parole Board to develop a written agreement by March 1, 2009. Parole Statistics Statistical data on offenders within the State’s correctional system are necessary to effectively manage the offender population, make informed policy decisions, evaluate outcomes, and measure the achievement of program objectives. Therefore, it is important that state agencies responsible for collecting, compiling, and reporting parole data do so in a way that provides policymakers and others with reliable, complete, and meaningful information. As stated previously, our audit found that currently no single source of parole data exists that can reliably be used to evaluate and inform parole policy. Specifically, we found that parole statistics differ depending upon the agency collecting and reporting the data, and explanations about these differences are lacking. As a result, there are risks that the data could be misinterpreted. Parole statistics, particularly those originating with the Department, are widely distributed to and relied upon by a variety of public and private sector entities and individuals including federal and state policymakers, judges, district attorneys, community advocacy groups, the Colorado Commission on Criminal and Juvenile Justice, and other state government entities. Because of the importance of parole data for decision-making, we reviewed the available data and identified areas for improvement as described below. First, we identified discrepancies between the Department’s and Board’s data that could lead to inaccurate conclusions and misstatements. The Department counts and reports paroles as the number of offenders who are actually released from a state correctional facility to parole, while the Board’s statistics report paroles as the number of decisions granting parole. The figures will differ for several reasons. For example, when the Board grants an offender parole it may set a release date that is several months after the parole hearing. The Board’s data will show that parole was granted, but the Department’s data will not show the parole release until the inmate is actually released from incarceration. In addition, even when the Board grants a parole request, if an offender does not meet conditions set by the Board, he or she will not be released. In this situation, the Board’s data will indicate that a parole was 16 Discretionary Parole Performance Audit - November 2008 granted, but the Department’s data will not show that a parole release occurred. Another difference in the data reported by the Board and the Department is related to re-paroles. The term “re-parole” refers to offenders who have been released again on parole after having their prior parole revoked. The Department includes reparoles in its count of releases, while the Board does not include them in its decision data. Second, we found that the Department’s data overstate the number of discretionary parole releases due to a change in the Board’s release policy. Specifically, Department data indicate that discretionary parole releases increased by 80 percent from about 2,800 in Fiscal Year 2006 to more than 5,050 in 2007. These figures are significantly higher than the number of paroles granted by the Board during the same time period. The reason for this discrepancy is that the Department inappropriately included early mandatory releases and re-paroles in its count of discretionary parole releases. In December 2005 the Board implemented a policy change allowing offenders whose mandatory parole dates fell on a Friday, weekend day, or holiday to be released a few days early to alleviate departmental transportation problems. When this change occurred, the Department recorded these early mandatory releases as discretionary releases, because technically, as defined by statute [Section 17-22.5403, C.R.S.], the Board’s approval of parole prior to an offender’s mandatory parole date is a discretionary parole, even if the offender is released only a few days earlier than his or her mandatory parole date. According to staff, the Department needed to create a separate field in the database to more appropriately account for these early releases; however, budget limitations restricted staff ability to do so in a timely manner. Budget constraints do not exempt agencies from providing true and accurate information, particularly when that information is widely disseminated and used by policymakers and others to make critical decisions about offenders and parole, including forecasting prison populations, developing new correctional system programs, and evaluating recidivism. The Department needs to improve the usefulness and comprehensiveness of Colorado’s parole statistics so policymakers can rely on the data to make informed decisions. To accomplish this, we believe the Department and Board should work together to identify and develop a cost-effective means of providing basic, consistent data on parole in Colorado. This could include a single report that encompasses data currently collected by the Department and the Board with clear definitions and descriptions of the differences in the data. Alternatively, this could include separate reports, with data from the Department and Board clearly explaining what the data does or does not include. Regardless of the source, the Department and Board should resolve different practices for defining terms to reduce erroneous conclusions on the numbers of paroles granted and offenders released. If the Department believes it is important to provide information on re-paroles or early mandatory releases, then it should clearly distinguish these Report of The Colorado State Auditor 17 from other parole types. Once the Department and Board determine their respective roles in reporting data on parole, they should formalize the agreement in writing. Recommendation No. 2: The State Board of Parole and Department of Corrections should work together to ensure that accurate and meaningful data are collected and reported on parole decisions by the Board and parole releases by the Department. This reporting should include information that is useful for policymakers to identify trends, understand changes, and make appropriate decisions. Further, the Board and Department should ensure mutual understanding of their duties related to the reporting of parole decisions by formalizing the process in a memorandum of agreement. State Board of Parole Response: Agree. Implementation date: January, 2009. The State Board of Parole concurs with the response of the Department of Corrections. Reporting needs of the Board and the Department differ, but it is recognized that any conflict in the statistical reporting should be explained. The Parole Board will include a caveat to address this issue in any future publication including the Board’s annual report. Department of Corrections Response: Agree. Implementation date: January 2009. The Department of Corrections is in agreement with this recommendation and will cooperate with the Parole Board to clarify why certain data reported by the Board may appear to be in conflict with the Department’s data. The Department will report data and trends on parole releases but will discontinue reporting Board decisions. By January 2009, the Department will meet with the Board to discuss language that can be added to each agency’s respective statistical reports to distinguish and report difference in terms such as decisions versus releases, statistical methods, and data interpretations, particularly as they relate to trends, changes, and decision-making. The Department will work with the Board to document each agency’s reporting responsibilities in a memorandum of agreement. Additionally, as of December 2008, the Department will begin separate tracking of “weekend releases” to identify offenders who are released a few days prior to their mandatory parole date. This page intentionally left blank. The electronic version of this report is available on the website of the Office of the State Auditor www.state.co.us/auditor A bound report may be obtained by calling the Office of the State Auditor 303.869.2800 Please refer to the Report Control Number below when requesting this report. Report Control Number 1975