Chapter Five
Findings and Recommendations
As policymakers and criminal justice administrators develop and implement policies that attempt to reduce crime and balance public safety with costs and proportional punishment for convicted offenders, they will need accurate information to make decisions and monitor outcomes. The program review committee analysis presented in this report can be used as the foundation for continued research into the state's offender population, crime rates, and sentencing patterns. The findings presented can provide policymakers and criminal justice administrators with information to review crime and sentencing policy, develop better models to predict inmate population and probation and parole caseloads, set budgetary priorities, and assess the classification and evaluation processes used by criminal justice agencies to manage the offender population. Ideally these data, in combination with other indicators developed over time, can assist in the evaluation of the efficacy of criminal sanctions and treatment and rehabilitation programs.
The program review committee study did not attempt to understand and explain all causes of recidivism, which typically are the result of complex societal issues. While the committee acknowledges offenders must certainly be accountable for their own behavior, repeat criminal activity is often rooted in problems within the family, school, or community and failures of rehabilitative and social service systems. It may be difficult for the criminal justice system, which does not have the primary responsibility for addressing most of these societal issues, to change well-established criminal behaviors of many serious and chronic offenders. However, the criminal justice system must still strive to protect public safety and rehabilitate offenders.
The recidivism rate data included here should be interpreted with caution. Recidivism is only one measure of the criminal justice system's performance. Further, responsibility for the rate of recidivism cannot be assigned to one agency within the criminal justice system. There are many examples of ways in which policy, resource allocation, or agency procedures affect the process or caseload of another criminal justice agency. For example sentencing laws, the speed of the court process, sentencing practices of judges, law enforcement activities, and probation and parole supervision procedures can increase or decrease admissions to prison. These factors are beyond the control of the Department of Correction, which manages the inmate population.
It is also important to note, this study relied on official records of criminal activity and, therefore, only measured offenses that were reported to or observed by the police and resulted in arrests. Many crimes go unreported or remain unsolved and, therefore, do not result in an arrest. The recidivism rate based on rearrest may underestimate repeat criminal activity by convicted felons. There is no way to accurately count the number of unreported crimes.
As stated, this study was the first step in measuring the rate of recidivism. While able to compile the data required to measure recidivism, ultimately many more questions were raised than answered by the analysis. This fact highlights the need to establish an on-going process to track and report on the rate of recidivism among Connecticut offenders. The following section sets forth the committee's recommendation to require the criminal justice system to provide the legislature with reliable and comprehensive recidivism data. In addition, a discussion of the key policy and budgetary issues for which the recidivism analysis may be useful is set forth.
Recidivism Reporting Requirement
Tracking the trends and patterns within the offender population is necessary to develop and implement effective and cost-efficient policies and programs that attempt to reduce crime and protect the public's safety. A key component of that information is an analysis of the new crimes committed by repeat offenders in the state. As this study showed more than half of the felony offenders committed new crimes after being discharged from prison or sentenced to probation. Therefore, tracking the rate of recidivism is crucial to a comprehensive understanding of crime. The program review committee found it is feasible to use existing automated criminal history data to calculate recidivism rates and to analyze the trends and patterns of reoffending among a large group of offenders.
Therefore, the Legislative Program Review and Investigations Committee recommended the Division of State Police, within the Department of Public Safety, begin to track and analyze the rates of rearrest, reconviction, and reincarceration of felony and misdemeanor offenders on a yearly basis. The division shall:
The program review committee recommended the Division of State Police is given the new responsibility of tracking the rate of recidivism for several reasons. First, the Division of State Police maintains all historical and current criminal record information. Its Bureau of Identification administers a repository of arrests, convictions, and sentencing information on all offenders arrested in Connecticut. To ensure the repository has current information, the bureau receives data from the courts and the Department of Correction on a regular basis.
The state police disseminate offender information -- usually in the form of a "rap sheet" -- to other law enforcement agencies, the state criminal justice system, the federal government, and other entities with the authority to request and receive such information (e.g., criminal background checks for licenses or employment). Under the planned OBTS structure, the state police will continue to be the "keepers" of automated criminal record information. (A summary of the offender-based tracking system and project development status are contained in Appendix B.)
Second, since 1977, the Division of State Police produced an annual report on statewide crime trends -- the Crime in Connecticut report. It is the only on-going analysis of crime data done by the state's criminal justice system. This report is published as part of the division's responsibility to provide information to the Federal Bureau of Investigation's Uniform Crime Reporting (UCR) program. The UCR program measures crime in the United States based on the number of violent and property offenses committed in each state. One objective of the UCR program is to produce reliable crime statistics for local law enforcement administration, operation, management, and research.
Third, the state police have 25 years of experience in conducting analysis and reporting on aggregate offender and crime data. It also has the staff, technical abilities, and equipment to process a large database. As previously stated, the division currently maintains the state repository of all criminal records.
Fourth, the state police have a good working relationship with the other executive and judicial branch criminal justice agencies as well as local law enforcement agencies. The state police have maintained the criminal offender data in accordance with all confidentiality requirements. There appears to be no issues surrounding the transfer of data to the division from other agencies.
Finally, the division supported the program review committee study and expressed an interest in continuing the research on recidivism among Connecticut offenders. In order to carry out this function, the division would need a modest increase in resources.
Policy Implications of Recidivism Data
Good quality recidivism data can be used by policymakers and criminal justice administrators to develop or evaluate various policy alternatives relating to budgeting priorities, crime and sentencing laws, administration of the criminal justice system, management of the offender population, and evaluation of state-funded programs and services. The policy implications of recidivism data in several critical areas identified through the program review committee research are discussed below. The committee, however, did not present specific recommendations.
Criminal justice administration. Consistently, only about one-quarter of convicted offenders are incarcerated. The vast majority plus most pre-trial defendants remain in the community, but not all are under active supervision. The program review committee found more than half of the convicted felony inmates and probationers who were released to the community repeatedly committed new crimes. Furthermore, offenders were rearrested on average two times during the three-year release threshold under review.
For the state's crime policy to be effective, it must address the cumulative impact of the thousands of offenders that return to or remain in their communities after conviction. The criminal behavior of these offenders must be curbed before any real reduction in the rate of recidivism can be achieved. While data show inmates in general have a high rate of recidivism, those in prison present less of an immediate concern in terms of repeat criminal activity than those in the community.
In general, there are two primary options to address recidivism. First, the state may incarcerate more convicted offenders and/or require inmates to serve longer periods of time in prison. Second, the state may redefine and reinvest in agencies that provide community-based supervision, treatment, and rehabilitation of sentenced offenders in an effort to reduce or prevent repeat criminal activity among the majority of offenders who are not incarcerated.
The recidivism data showed offenders who spent longer periods of time in prison had lower rates of rearrest. This option, however, is extremely expensive and may be contrary to other criminal justice goals such as making the punishment proportional to the severity of the crime.
It costs about $96 per day to house an inmate, and the Department of Correction's annual budget is currently over $500 million. In the short term, since most inmates are serving on average less than three years in prison, increased incarceration rates may merely postpone recidivism without reducing the total amount of repeat criminal activity over time.
The recidivism data also showed inmates and probationers who were under some form of community supervision (i.e., parole, probation, or DOC halfway house) after discharging from prison were less likely to be rearrested. Although the number of new crimes committed by the inmate and probationer cohort groups were high, overall the crimes were non-violent and consisted of less serious property and drug offenses and crimes such as disorderly conduct and motor vehicle infractions. The alternative option, therefore, is to use incarceration more selectively and cost effectively by investing in community-based supervision agencies and rehabilitation and service programs.
While this is a less costly model, it presents some risk to the public's safety and property because convicted offenders remain in or return to the community. It is typically viewed as the "soft on crime" approach and often lacks the political support necessary to receive appropriate funding to be implemented as intended. This model attempts to reduce the rate of recidivism by curbing the criminal behavior of those offenders in the community. The basic elements of this model are already in place in Connecticut.
The pressing policy implication of the program review committee's analysis is whether to incarcerate more offenders for longer periods of time and at great expense or re-examine and efficiently use limited prison resources by reinvesting in agencies and programs that attempt to control or improve offenders' behavior while they remain in the community. Policymakers must balance the need to reduce crime and protect public safety with limited financial resources and appropriate punishment for convicted offenders.
Sentencing laws. Sentencing law and policy is often developed to address the serious and violent offender in an effort to prevent future violence. Over the past five years, the legislature has enacted a series of "truth in sentencing" reforms aimed at increasing penalties and reducing the discrepancy between the court-imposed sentence and the actual time served by an inmate. For example, violent offenders are required to serve at least 85 percent of their court-imposed sentence to be eligible for parole, while all other inmates must serve at least 50 percent of their sentence. In addition, state law establishes mandatory minimum sentences for certain violent crimes. The current sentencing policies allow the criminal justice system to take a more conservative approach with all offenders including those who commit non-violent and less serious crimes.
As this study points out, some amount of repeat criminal activity is going to occur, but in general the new crimes are nonviolent. Violent offenders represented a small percentage of the total offender population. Violent offenders had the lowest rates of recidivism and were the least likely to be rearrested for another violent crime. Property offenders were rearrested at significantly higher rates and were more likely to re-commit another property offense.
The policy implication centers on the "tough on crime" debate11 and whether a more punitive approach should be taken against all offenders or be more specifically focused on a certain type of offender. As sentencing requirements are increased for the small percentage of serious and violent offenders, sanctions against all other types of offenders typically increase or toughen. In Connecticut this has resulted in an influx of many less serious offenders being incarcerated. The parole board's conservative approach has caused offenders to serve longer periods in prison than originally intended by state law, and the overloaded adult probation system cannot provide adequate supervision of offenders in the community. Together these factors have overwhelmed the criminal justice system.
Resource allocation. While there is no clear consensus on how to prevent recidivism, there is agreement that the economic, social, and political costs of crime are overwhelming. During 2000, the average daily offender population included almost 18,000 inmates in state prisons and over 60,000 convicted offenders in the community under probation or parole supervision or another community-based sanction. As previously stated, more than half were rearrested for new crimes. These numbers are alarming considering the direct and indirect costs offenders impose each time they are rearrested, prosecuted, evaluated and treated, and reincarcerated or sentenced to an alternative sanction.
Executive and judicial branch criminal justice agencies do not calculate the costs of the arrest process or disposition of a criminal court case. Nor is there national data on these costs.
The criminal justice system does calculate the average daily costs of incarceration and community supervision (e.g., parole and probation). In Connecticut, it costs $35,000 per year to incarcerate an inmate ($96 per day), about $4,000 per parolee per year ($11 per day), and $833 per probationer per year ($2 per day).
Since the cost information is incomplete, however, the program review committee was unable to determine the cost of repeat criminal activity. To calculate the costs of recidivism, the program review committee found a method to determine the cost of each step in the criminal justice process must be developed. Eventually, this information along with the recidivism data can be used to calculate the economic costs of repeat criminal activity.
Currently, what can be calculated are the annual expenditures for the state's criminal justice system. In fiscal year 2000, almost $912 million was appropriated to seven criminal justice agencies including the judicial branch, the Departments of Correction and Public Safety, the Board of Parole, state's attorneys and public defenders, and the Office of the Victims Advocate. (A summary of the state expenditures for criminal justice agencies is contained in Appendix C.)
Resource planning. Each year, less than one-quarter of convicted offenders are sent to prison. The majority of offenders receive a community-based supervision sanction such as probation. The recidivism data showed 22 percent of the inmate group and 11 percent of the probationer group were reincarcerated as a result of a reconviction for a new crime. Most received probation or another alternative sanction.
Each year since 1989, the Department of Correction has built new facilities or added new beds through expansion and renovation projects. It has spent over $1 billion to add about 9,000 prison beds. The addition of new prison beds is continuing. During the past two fiscal years, DOC received $35 million -- in addition to its $500 million annual operating budget -- to add 600 new prison beds and convert 500 existing dormitory beds into a Community Justice Program to assist inmates returning to the community.
Also, in fiscal year 2000, the DOC budget included $12 million to contract for 500 beds in two Virginia prisons. Due to overcrowding, the department has been sending Connecticut inmates out-of-state for the past two years.
The correction department and the Office of Policy and Management (OPM) do separate projections of the inmate population. The current methods for projecting growth of the inmate population have proved to be inaccurate. Both agencies track the trend by calculating growth based on past increases. They do not calculate other factors that impact the inmate population such as trends in the population within their crime-prone years, arrests and prosecution rates, sentence lengths and time-served percentages, or other influences such as policy changes, increases or decreases in funding or resources, and the capacity of community-based supervision programs. Furthermore, they have not analyzed the rates of recidivism, which have a significant impact on the growth of the prison population.
Recidivism data can provide information not only on the potential number of repeat offenders who will be reincarcerated, but on the types of offenders and their supervision, rehabilitation, and service needs. These data can be used to more accurately determine the number of new beds needed as well as the security level of the facility and the type of programming space required. Over time, these data can provide a more precise analysis of the increases and decreases in the prison population.
During the past five years, budgetary shortfalls and continued growth in the number of offenders under community supervision (e.g., probation and parole) have resulted in larger caseloads, but fewer services and programs. The inequities in state expenditures for prison and community supervision staff and programs have stalled the development, operation, and effectiveness of community corrections agencies. An unintended result of a disproportionate share of the total budget being allocated to prison services is a high rate of recidivism among inmates and probationers under community supervision.
As importantly, recidivism data can be used to project the growth and future needs of the vast majority of the offender population that are in the community. Any effort to reduce recidivism must focus on those offenders living in the community who pose the most immediate risk to public safety.
Program and service planning. While there is conflicting research about the effectiveness of rehabilitation, treatment, and service programs for offenders, the research suggests programs can have a modest impact on reducing recidivism if they combine a variety of components such as education, work training, and counseling and are intensive and tailored to offender subgroups such as women, sex offenders, or the mentally ill.
The program review committee found prison and community-based programs did not reduce the rates of rearrest among the inmate and probationer samples. In fact, offenders who participated in certain programs (e.g., administrative segregation and disciplinary, mental health, alternative incarceration, and 12-step substance abuse) had a higher rate of recidivism than those who did not participate. The only programs that significantly reduced the likelihood of rearrest were the DOC prison industries program for inmates and the judicial branch's day incarceration center for probationers.
The analysis provided some indication certain groups of offenders may warrant special attention in the development of rehabilitation and service programs. For example, younger offenders, those in a minority group, offenders with serious substance abuse problems, and those with extensive criminal histories require programs tailored to their specific needs.
Despite inconclusive evidence programs are effective, the criminal justice system relies heavily on these services, especially those based in the community. The courts, correction department, and parole board continue to place thousands of offenders each day in residential and nonresidential programs.
With a limited understanding of program effectiveness, Connecticut is inefficiently spending a significant amount of money. The policy implication centers on determining the cost benefit and efficacy of rehabilitation, treatment, and service programs for offenders. The recidivism data can be used to more wisely and effectively allocate the limited resources, and better identify the appropriate offenders for each program.
The data may also be useful in developing outcome measures, benchmarks, and targets to evaluate the effects of the rehabilitative, treatment, and supervision programs on reducing recidivism. However, as used in this study, the data alone are insufficient to conclude whether a specific program is effective in reducing recidivism. To enable more definitive conclusion, the program review committee found the criminal justice system must conduct periodic evaluations to compare the rates of rearrest, reconviction, and reincarceration of offenders participating in programs with those of similar groups who do not participate in programs.
In addition, the data may be used to determine the most effective sanction for dealing with misbehavior that is not necessarily criminal, but is a technical violation of probation or parole. The criminal justice system can evaluate whether the practice of reincarcerating offenders with technical violations reduces future criminal behavior, or if it is more effective to place them in alternative, less costly, community-based sanctions.
Inadequacies of offender records. Program participation data for the inmate and probationer samples are not automated. The program review committee staff conducted a review of Department of Correction, Board of Parole, and judicial branch probation case files. The committee staff found information in the files to be missing, inconsistent, often times inaccurate, and insufficient to fully identify the programs offenders had participated in and completed. The lack of data impedes the evaluation of the impact these programs have on the rate of recidivism and contributes to the difficulties in holding service providers accountable.
To provide complete and reliable information to evaluate the efficacy of prison and community-based programs and to effectively allocate limited resources, the program review committee found the criminal justice agencies need to improve their record keeping and case management practices. Agencies should be able to immediately and accurately identify offenders who have obtained GEDs, participated in a work training program, received substance abuse or mental health treatment, or who have participated in various programs that may reduce recidivism.
The program review committee acknowledges improvements in data collection have been thwarted by budget cuts and a lack of staff. It is understandable certain administrative functions such as record keeping become less of a priority as probation and parole supervision caseload requirements and the inmate population increase. The daily management of offenders obviously takes precedent.
Policymakers need to be aware the system will continue to be unable to provide certain information as long as the data systems are inadequate. Until an investment is made in developing quality information, policy and budgetary decisions will continue to be based on estimates, anecdotes, and imprecise analysis.
10 The offender-based tracking system (OBTS) is a statewide, multi-agency, automated tracking system, which is scheduled to come on-line in 2003. The OBTS will allow for immediate access to an offender's current and historical criminal information, including a current report on his or her status and custody.
11 Refer to the Legislative Program Review and Investigations Committee report on Factors Impacting Prison Overcrowding (December 2000).