Chapter Four

Recidivism Among Connecticut Felons

A main purpose of the program review committee study was to provide for the first time a comprehensive picture of recidivism rates among Connecticut felons. As previously stated, the analysis focused on five questions -- highlighted below -- to accomplish this task. Answers to each, based on the committee's research are also summarized below. Then the complete data analysis of the two cohort groups -- inmates and probationers -- is presented.

To what extent were Connecticut felons arrested for new criminal activity, convicted of those offenses, and sentenced to either imprisonment or other supervision sanctions?

• Within three years of discharge from prison, most (70 percent) felony inmates were rearrested at least once for a new crime.
• Almost half of the discharged inmates were reconvicted of a new crime.
• About one-quarter of the inmate group was reincarcerated and an additional third sentenced to another sanction as a result of a new crime.
• Within three years of being sentenced to probation for a felony conviction, more than half (58 percent) of probationers were rearrested at least once for a new crime.
• One-third of the probationers were reconvicted of a new crime.
• Very few probationers were sent to prison as a result of a new crime (11 percent), but 21 percent were sentenced to another community supervision sanction.

How did recidivism rates differ among released inmates and probationers?

• Felony inmates had significantly higher rates of rearrest, reconviction, and reincarceration than felony probationers.
• During the three-year release threshold, inmates were rearrested more often than probationers -- inmates had on average three rearrests compared to two rearrests among the probations.
• The highest recidivism rates were among young, male, African American offenders in both groups.
• Inmates and probationers with extensive prior criminal histories were more likely to be rearrested than offenders with less serious criminal histories.
• Inmates and probationers with prior violation of probation offenses had significantly higher rates of rearrest than other offenders.
• In general, there was no consistent pattern of rearrest for a new drug crime among the drug offenders in the inmate cohort group.
• Probationers originally sentenced for a felony drug offense were more likely to be rearrested for the same crime than other types of offenders.
• Most offenders who recidivated were not reincarcerated for long periods of time and, in fact, many were not sent to prison.
• More than half of the rearrests do not result in reconviction; they are ultimately dismissed by the court, the offender found not guilty, or the charges combined into another criminal case against the offender.

How did recidivism rates vary among different categories of offenders?

• Males had significantly higher recidivism rates than females.
• Young offenders were more likely to reoffend than older inmates.
• Minority offenders were more likely to be rearrested than Caucasian offenders.
• Property offenders reoffended more often and were more likely to recommit the same type of crime than violent offenders.
• Inmates released from prison to some form of community supervision such as parole or transitional supervision had a significantly lower rate of rearrest than inmates who returned to their communities under no supervision.

What types of new offenses did repeat offenders commit?

• In general, repeat offenders in both cohort groups committed a variety of new felony and misdemeanor crimes and did not "specialize" in one type of crime.
• Overall, most of the repeat criminal activity among both cohort groups was nonviolent and less serious felonies and misdemeanor property crimes, drug sale and possession offenses, and crimes such as disorderly conduct, breach of peace, and motor vehicle infractions.
• Among inmates and probationers, property and drug offenders were more likely to recommit those same types of offenses.
• Violent inmates were the least likely offenders to recommit another violent crime.
• A previous drug conviction was not a strong predictor of rearrest for a new drug sale or possession crime.

Was recidivism related to an offender's criminal history, demographics, program participation, or other factors?

• The patterns and trends in repeat criminal activity among Connecticut felons were consistent with national research.
• An offender's age, race, and gender were significant predictors of repeat criminal activity among the inmate and probationer samples.
• Young, male, African American, property offenders were the most likely to recidivate.
• The type of primary offense and length of primary sentence were strong predictors of rearrest among the inmate cohort group.
• Offenders with a chronic or serious substance abuse problem were significantly more likely to be rearrested, but most were not rearrested for a new drug crime.
• Participation in prison or community-based rehabilitation, treatment, and service programs did not significantly reduce the rate of recidivism rates among either cohort group.
•  

Detailed Recidivism Analysis

Presented below is the detailed analysis of the rates of recidivism among members of both cohort groups. The analysis of the inmate sample is presented followed by the probationer sample. The following section includes detailed recidivism analyses of:

• overall rates of rearrest, reconviction, and reincarceration;
• rates by specific offender demographic or characteristic;
• patterns of repeat criminal activity;
• time at risk in the community;
• reconviction and sentencing; and
• program participation by the offenders.

Inmate Cohort Group

The inmate rate of recidivism is presented as the percentage of the total of discharged inmates who were rearrested, reconvicted, and reincarcerated during the three-year release threshold. As expected, the rearrest rate exceeds the reconviction rates because not all arrested offenders are prosecuted and convicted and, due to the court's lag time in disposing of cases, not all convictions occurred within the selected follow-up period. Likewise, reconviction rates are greater than reincarceration rates because not all convicted offenders are sent to prison. Some are sentenced to probation, fined, or diverted into alternative sanction programs.

Overall rates of recidivism. As shown in Figure IV-l, 69 percent (2,745) of the 4,006 inmates released from prison in 1997 were rearrested at least once for a new felony or misdemeanor crime, and 46 percent (1,828) were subsequently reconvicted within the three-year release threshold. Twenty-two percent (889) were reincarcerated as a result of that reconviction, and 18 percent received a nonprison sentence of probation supervision, an alternative sanction, or a fine.

Rate based on primary offense category and crime type. This analysis is based on the inmate's primary offense, which is defined as the crime for which an inmate received the prison term for which he or she discharged in 1997. Figure IV-2 shows the recidivism rates for inmates based on five major crime categories, which were discussed on page 15 of this report. (There were no inmates in the sample with a primary offense for a motor vehicle crime.)

As shown, inmates who were in prison for a property offense (74 percent) or a violation of probation (75 percent) were the most likely to be rearrested. These offenders also had the highest reconviction and reincarceration rates. More than half were reconvicted for a new crime, and as a result almost 30 percent were sent back to prison.

Violent offenders had the lowest rate of rearrest (61 percent), although more than half was rearrested. Overall, inmates whose primary offense was the possession of a weapon, risk of injury to a minor, perjury, and conspiracy to commit a crime -- categorized as other crimes -- had the lowest reconviction (about 39 percent) and reincarceration rates (17 percent).

The recidivism rate among inmates in Connecticut by specific types of primary offense are consistent with national research. As shown in Table IV-1, inmates with a prior conviction for burglary, larceny, drug possession, weapon possession, or violation of probation were more likely to be rearrested and reconvicted after being released from prison. About three-quarters within each crime type were rearrested and almost half were reconvicted within the three-year release threshold. Among all of the crime types, these offenders were also among those with the highest rates of reincarceration -- about 25 percent were returned to prison.

Interestingly, inmates who committed certain violent crimes -- homicide, sexual assault, kidnapping, and arson -- had the lowest rearrest and reconviction rates. This may be attributed to more intensive and restrictive community supervision requirements as they discharge from prison, or these inmates may be older at discharge as a result of serving long sentences and, therefore, less likely to recidivate. Whatever the reasons, less than 40 percent of inmates convicted of homicide and sexual assault were rearrested. About 42 percent of those with a

kidnapping conviction and 48 percent of those convicted of arson were rearrested. As a group, these inmates had less than a 30 percent reconviction rate. However, their rate of reincarceration, while slightly lower than offenders convicted of robbery, burglary, larceny, and drug offenses, was still consistent with the overall rate for the inmate sample. Inmates who committed arson had the lowest rate (15 percent) of reincarceration.

Among violent offenders, inmates who committed robbery and assault had the highest rates of recidivism. Over 70 percent of them were rearrested within three years after discharge from prison, half were subsequently reconvicted, and 25 percent were sent back to prison. About 65 percent of those who committed an assault were rearrested, 42 percent were reconvicted, and 21 percent eventually reincarcerated.

Rate based on primary discharge type. The primary discharge is defined as the specific type of release from prison in 1997. As previously stated in Chapter Three, inmates may be released from prison in several ways, including: serving their complete prison term (called "maxing out"); being paroled; or transitioning back to the community under a number of DOC early release programs (e.g., transitional supervision, halfway house placement, and re-entry furlough). Parole and transitional supervision have a community supervision component for a specific period of time prior to the termination of the sentence whereas the others do not. The analysis below examines the recidivism rates based on the manner in which the inmates were discharged from prison.

As shown in Figure IV-3, the overall rates for each type of discharge are similar, but there are some notable variations. Inmates who were released early from prison on parole or transitional supervision were statistically less likely⁹ to be rearrested than those who "maxed out" or were released by DOC to a halfway house or on a re-entry furlough. This may be attributed to the community supervision component of parole and transitional supervision rather than an inmate's predisposition to commit another crime. Inmates under supervision may have their parole or TS release revoked for a technical violation and be returned to prison prior to or in lieu of a new arrest. However, it may also indicate that some form of supervision and/or surveillance decreases the likelihood of a new criminal activity.

Rate based on primary sentence. An inmate's primary sentence is the prison term imposed by the court for the original primary conviction, and it is the sentence from which the inmate discharged in 1997. The data showed inmates with longer court-imposed prison sentences were less likely to be rearrested after being released from prison. These inmates were, on average, older than other inmates at discharge and, therefore, likely to be "aging out" of their criminal careers. In addition, because of their longer sentences they were more likely to be under some form of community supervision (e.g., parole or probation) upon release. The data indicated this has a positive impact on reducing the rate of rearrest.

Total criminal activity. An offender who is arrested may be charged with more than one crime. For the purposes of this study, the first three charges -- or crimes -- per arrest were examined to provide a snapshot of the total criminal activity of inmates and probationers. Because of this approach, the total number of crimes will be greater than the number of offenders.

Within the three-year release window, 2,745 inmates were rearrested at least once. Based on the total charges for the first rearrest, the inmates accounted for 5,573 new crimes. This analysis includes both felonies and misdemeanors.

As shown in Figure IV-4, more than half (68 percent) of the crimes were for the sale or possession of drugs, motor vehicle infractions, a violation of probation, and other crimes such as weapon possession, risk of injury to a minor, stalking, harassment, disorderly conduct, prostitution, and bribery. Only 13 percent of the crimes were violent (i.e., homicide, sexual assault, assault, and robbery) and 19 percent were property offenses such as larceny, burglary, and forgery. When the offenses for multiple rearrests were examined, this pattern was similar.

Rate by demographics. The relationship between recidivism rates and certain offender demographic characteristics such as age, gender, and race was analyzed. Figure IV-5 illustrates the rearrest, reconviction, and reincarceration rates among inmates by age group. As shown, the recidivism rate for all three measures was significantly higher for younger inmates. About 80 percent of inmates between the ages of 16 and 21 were rearrested, more than half were reconvicted, and almost 30 percent reincarcerated for a new crime.

Inmates over 40 were less likely to recidivate, which supports conclusions reached in other studies that older offenders are not rearrested because they "age out" of their criminal career.

Figure IV-6 shows male inmates had a higher rearrest rate than females. However, males and females had similar rates of reconviction and reimprisonment.

Measuring recidivism by the racial group, the data show offenders in a minority group had higher rates of rearrest, reconviction, and reincarceration than Caucasian inmates. As shown in Figure IV-7, African American and Hispanics inmates were significantly more likely to be rearrested. (American Indian, Asian, and other racial groups were not included in the analysis because they represent less than 1 percent of the total inmate sample.) However, the differences for the reconviction rates are not statistically significant.

When the race and age at discharge distributions were analyzed together, an inmate's age and race were strong predictors of recidivism. Young, minority inmates were more likely to be rearrested. Figure IV-8 shows African American and Hispanic inmates between the ages of 16 and 21 years were twice as likely as Caucasians of the same age to be rearrested within the three years after discharge from prison. While older Caucasian inmates had slightly higher rates of recidivism than minorities, the analysis showed they were not significantly more likely to be rearrested.

Rate by other characteristics. The rate of recidivism was further analyzed by the inmates' educational attainment, substance abuse, and mental health levels to determine if they had any relationship to the recidivism rates. Table IV-2 provides the percentage of inmates for various levels of educational attainment, mental health, and substance abuse who were rearrested at least once and subsequently reconvicted and reincarcerated.

Consistent with the national research, the lower the educational grade level or the more chronic or serious the mental illness or substance abuse problem, the higher the overall rate of rearrest and reconviction. However, the analysis showed only an inmate's substance abuse level was a strong predictor of rearrest, while educational attainment and mental health levels were not.

Patterns of repeat criminal activity. Criminal justice researchers have studied the general patterns of criminal behavior in addition to measuring the rates of recidivism. This research is often used to determine whether repeat offenders "specialize" in certain types of crime. The consensus throughout the literature is most recidivists have a varied pattern of offending and typically commit different types of crimes. Therefore, the program review committee analyzed the first rearrest within the three-year period after discharge from prison for any new crime and not just the same crime.

Table IV-3 shows: (1) the total number of inmates discharged from prison in 1997 by their primary type of crime (i.e., homicide, robbery, burglary, sale of drugs, weapon possession, etc.); (2) the number rearrested for the same exact crime type as their primary offense; (3) the number rearrested not for the exact same type of crime, but for one in the same crime category (i.e., violent, property, drug, other) as their primary offense; (4) the number rearrested for any other crime; and (5) the number not rearrested within the three-year release threshold.

There are limitations to these data. First, as previously stated, an inmate may be charged with more than one crime at rearrest. The analysis includes up to three charges per rearrest. The number of crimes listed, therefore, is greater than the number of inmates.

Second, this analysis is based on an inmate's first rearrest after being released from prison. Many of the inmates were rearrested multiple times during the three-year release threshold and those who did not recommit the same type of crime as their primary offense at first rearrest may have done so at a later rearrest. However, a review of subsequent rearrest data showed a similar pattern.

Most inmates were not rearrested for the exact same crime that resulted in their original incarceration. Specifically, less than 30 percent of inmates initially convicted of larceny were rearrested for a new larceny offense, and 19 percent of burglars recommitted a new burglary. Among drug offenders, 23 percent of those incarcerated for drug possession were rearrested for that crime, and 16 percent of drug sellers were rearrested for the sale of illegal drugs. Inmates previously convicted of assault had the highest rate (16 percent) among violent offenders of recommitting the same exact crime. None of the inmates whose primary offense was homicide recommitted murder or manslaughter, and only 5 percent of sexual assault offenders were rearrested for that same crime. About 7 percent of robbery offenders recommitted a robbery.

As shown in Table IV-3, property and drug offenders were more likely to "specialize" in a certain type of crime -- or recommit the exact same crime as their primary offense. Overall, 35 percent of property offenders were rearrested for the first time after their release from prison for a property crime, and 26 percent of drug offenders recommitted another drug crime. A pattern of repeating the same type of crime is less evident among violent offenders. Less than one-quarter of violent offenders were rearrested (for the first time after discharge from prison) for a violent crime.

As noted above, when the types of crimes committed by inmates with multiple rearrests were examined, the data showed a similar pattern. Only property and drug offenders showed a likelihood to be rearrested for new property crimes, while violent offenders were less likely to recommit another violent crime. The vast majority of the new criminal activity involved property crime, drug sale and possession offenses, and a wide range of nonviolent and less serious felonies and misdemeanor crimes such as disorderly conduct, criminal trespass, breach of peace, and motor vehicle infractions.

Severity of repeat crime. A comprehensive analysis of crime includes, in addition to the number and type of offenses, a review of the severity of the offenses. For the purposes of this study, severity is measured by felony and misdemeanor status of the crime. Felony offenses are more serious and under state law punishable by more than one year in prison. Misdemeanors are less serious and punishable by a year or less in prison. Persons convicted of a felony or misdemeanor may also be sentenced to an alternative sanction or a fine. There is a third category of offenses -- a violation or infraction. These crimes do not rise to the level of a felony or misdemeanor offenses, and are typically a breach of a local ordinance, a motor vehicle offense, or other minor offense.

Table IV-4 shows the severity of the crimes for which the inmate group was rearrested. As expected, most of the violent crimes were felonies, except for assault. Almost 80 percent of assaults were misdemeanors. Based on the data, a large percentage of all violent crimes committed during the three-year release threshold were for assault, but were of a less serious nature.

Almost three-quarters of the property crimes, which represented the bulk of repeat criminal activity by the inmate group, were misdemeanors. However, all of the drug sale and most of the drug possession offenses were felonies.

Only a small percentage of the repeat criminal activities were infractions. Most of these were motor vehicle offenses such as license and driving violations.

Number of new arrests. The program review committee database included the total number of rearrests per inmate within the three-year period under review. Thirty-one percent of the inmates were not rearrested. Of the 69 percent rearrested, there were on average almost three (2.7) rearrests per offender. The number of rearrests ranged from one to 24.

The data were analyzed by the number of rearrests and the primary offense. There were no real differences in the percentages of inmates within each crime category with multiple rearrests except property offenders. Property offenders tended to have more rearrests than other types of offenders. As shown in Table IV-5, almost one-third had at least six rearrests during the three-year release threshold.

The rearrest data were also examined by the inmates' age at discharge. The analysis showed younger inmates, especially those between 16 and 21 years, had more rearrests during the three-year release threshold than older inmates. Almost 40 percent of inmates between 16 and 21 years were rearrested five or more times in three years.

In interpreting these data, it is important to acknowledge that those inmates with only one rearrest during the three-year period may have been reincarcerated during the remainder of the at-risk period, which ended at December 21, 2000. Therefore, they would not have the opportunity to reoffend multiple times. This issue will be examined later in this chapter.

Time at risk. The length of time offenders were in the community and at risk of reoffending until the first rearrest was examined. This period of time was defined by the program review committee as the release threshold. For the purposes of this study, the minimum at-risk period is three years. Even those offenders under some form of community supervision such as probation or parole are still considered at risk of reoffending. The committee analyzed the at risk period until the first rearrest for each group and the average length of time between multiple rearrests.

The average period at risk prior to the first rearrest among the inmates was about one year. Figure IV-9 shows 25 percent (1,010) of all inmates (4,006) were rearrested within the first six months after their discharge from prison. By the end of the first year after discharge, 40 percent of the inmates had been rearrested, and within three years after discharge almost two-thirds (65 percent) were rearrested at least once.

Of the 2,745 inmates who were rearrested during the three-year period, 78 percent (2,147) were rearrested more than once. Therefore, the at-risk period between multiple rearrests was calculated as the number of days between dates of rearrest. The at-risk period between multiple rearrests was substantially shorter than the period from discharge to first rearrest, which as stated was about one year. The average at-risk period between the first and second rearrests was six months. The average period drops to about four months for each subsequent rearrest. The at-risk period between rearrests is naturally shorter as the number of rearrests increases because the release threshold under review is only three years.

The data showed less than 3 percent of rearrested inmates were sentenced to three or more years in prison as a result of their first rearrest. The majority of the inmates remained in or returned to the community after a brief period of incarceration at some point during the three-year release threshold.

Reconviction and sentencing. The various types of dispositions were categorized as guilty or not guilty. The criminal sentence that may be imposed for a guilty verdict included the use of prison, probation, fine, or an alternative sanction such as conditional and unconditional discharge, accelerated rehabilitation, youthful offender status, or diversionary program. No sanction was imposed for a not guilty verdict.

Almost half (46 percent) of the inmates who discharged from prison in 1997 were subsequently reconvicted of a new crime. Figure IV-10 shows almost half (45 percent) were reconvicted of a drug sale or possession offense and 20 percent of property crimes, the majority of which were burglaries. Nineteen percent of the inmates were convicted of a violent crime, most of which were for assault and robbery.

The severity of the reconviction offenses was also examined. Again, the status of felony, misdemeanor, and infraction were used to gauge the seriousness of a crime.

When analyzing reconviction data, it is important to note that an offender may not be convicted of the exact crime for which he or she was arrested. As discussed earlier in this report, factors such as combining or dropping charges, plea bargaining, dismissal of charges, and due process or evidentiary issues can have an impact on the ultimate disposition of a criminal case. The recidivism data examined showed in most cases inmates were not reconvicted of the exact crime for which they were rearrested, but were reconvicted of the same category of crime (e.g., violent, property, drug). For example, an inmate may have been charged with a drug sale offense, but convicted of a lesser drug possession crime. Also, many times offenders were charged with a felony, but convicted of a misdemeanor.

Overall, the inmate group was reconvicted of more felony crimes. This may be attributed to their more extensive criminal histories, which may limit plea bargaining options or dismissal of charges.

All of the reconvictions for the violent crimes of homicide, sexual assault, kidnapping, arson, and robbery were at the felony level, but reconvictions for assault were mostly misdemeanors. These violent crimes, however, accounted for a very small percentage of the total repeat criminal activity.

Most of the repeat criminal activity by inmates involved property crimes. About three-quarters of the burglary and forgery reconvictions were felonies, but only half of the larceny reconvictions were felonies. Similarly, most of the reconvictions for a drug sale or possession crime were felonies.

In general, the serious nature of the new crimes for which inmates were reconvicted is a factor in the type of sentences imposed by the court. As previously stated, 22 percent of all discharged inmates (4,006) were sent back to prison and 18 percent were sentenced to a period of probation as a result of a reconviction for a new crime.

However, as Figure IV-11 shows, of the inmates reincarcerated, 42 percent were sentenced to a specific period of incarceration -- a "flat" prison term. One in five were sentenced to a prison term followed by a period of probation -- a "split" sentence. The average prison term was slightly less than two years, and the sentence lengths ranged from 30 days to 55 years.

Almost 30 percent of the reconvicted inmates were sentenced to a period of probation, which averaged about three years, and 10 percent were fined. Finally, 4 percent of the reconvicted inmates did not receive a specific sentence. The lack of sentencing data may be a result of missing or incomplete data or the inmate's case may still be pending for the court to impose a sentence.

Program participation. As discussed in Chapter One, the rate of program participation among a random sample of inmates was examined. Program participation in prison and in the community was reviewed for 423 inmates. Inmates are not required to participate in programs while incarcerated, but they may be required by DOC, the parole board, or the court to participate in a specific community-based program as a condition of their early release from prison.

DOC administers or funds over 300 prison and community-based programs and services for inmates. The programs are categorized as: academic education; addiction services; mental health services; administrative segregation and disciplinary programs; family and parenting services; prison industries; self-improvement programs; sex offender program; transitional services; vocational education; residential community-based programs; nonresidential community-based programs; and other services such as health education, cultural diversity, and religious services. Due to patient confidentiality issues, certain health education programs (e.g., AIDS and HIV awareness and education services) were not included in the analysis.

The program participation data were used to determine whether there was any difference in the recidivism rates among those offenders who participated in programs and those who did not. Table IV-6 compares the rate of rearrest among both groups of inmates -- participants and nonparticipants.

Overall, the data showed program participation had no positive relationship to the inmates' likelihood of rearrest after discharging from prison. The only program that showed a significantly lower rate of rearrest among participants was prison industries. This may be attributed to two factors. First, typically inmates with longer prison sentences are selected to participate in the prison industries program. This allows a sufficient period of time to train and provide actual work experience for the inmates. Inmates with shorter sentences are generally released from prison prior to completing the training and being assigned a job. It should be noted, however, the data showed inmates with longer sentences were less likely to be rearrested even without participating in the prison industries program. Second, prison industries provides marketable skills training, practical work experience, and pays a minimum wage that may better assist in an inmate's transition from prison to the community.

For some types of programs such as academic education, mental health, self-improvement, and transitional services, the inmates who participated actually had a higher rate of rearrest than those who did not participate. This pattern appears to be counter-intuitive.

One explanation is participation in these programs may not be a contributing factor to the likelihood of rearrest. The programs may not be effective enough to overcome the other causes of repeat criminal activity or simply may not be addressing the specific causes of recidivism. For example, inmates with serious mental illness may participate in mental health services while in prison, but they are not cured. Once released from prison and no longer residing in a structured environment, the inmate may fail to continue with treatment and return to the behaviors that resulted in their previous incarceration.

Another example involves inmates who are placed in the administrative segregation and disciplinary programs. The correction department places inmates in these programs because they are chronic or serious disciplinary problems or their primary offense is of such a serious or violent nature that it is difficult and/or unsafe to manage them in the general inmate population. The objective of these programs is management of the inmate's behavior while in prison and not rehabilitation. Almost all of the offenders who participated in this type of program were rearrested with three years after discharging from prison whereas two-thirds of the inmates who did not participate were rearrested. It is not surprising, therefore, that inmates who cannot behave in prison were rearrested once discharged.

Probationer Cohort Group

The analysis of the probationer cohort group parallels that of the inmate group. The same limitations on the data for the inmate sample also apply to interpreting the probationers' data. For example, the rearrest rate exceeds the reconviction rates because not all arrested offenders are prosecuted and convicted and, due to the court's lag time in disposing of cases, not all convictions occurred within the selected follow-up period. Likewise, reconviction rates are greater than reincarceration rates because not all convicted offenders are sent to prison. Some are sentenced to probation, fined, or diverted into alternative sanction programs.

Overall rates of recidivism. Figure IV-12 shows 58 percent (6,021) of the 10,402 probationers were rearrested at least once during the three-year release threshold, and 32 percent were subsequently reconvicted of a new felony or misdemeanor crime.

This cohort group of offenders had originally been sentenced to probation supervision or another sanction without a community supervision component rather than incarcerated as a result of their primary felony conviction. However, 11 percent were sent to prison as a result of a reconviction, but almost half of them received a "split" sentence, which is a period of incarceration followed by a period of probation supervision.

Twenty-one percent of the probationers were again placed under probation supervision or sentenced to another alternative sanction as a result of a new crime. It is interesting to note, however, most rearrested probationers were not sentenced.

Rate based on primary offense category and crime type. Figure IV-13 shows the rates of recidivism for probationers based on the six major crime categories of primary offense. Drug and VOP offenders were more likely to be rearrested, reconvicted, and incarcerated than the other probationers in the group. Motor vehicle offenders were significantly less likely to recidivate within the three-year release threshold and, of those rearrested, none were sent to prison.

The rate of incarceration among the probationers who recidivated was low (about 11 percent), which may be related to their less extensive and serious criminal histories. Only the VOP and drug offenders had a higher rate of incarceration (13 percent).

Repeat criminal activity by probationers was also examined based on the offenders' primary crime type. Table IV-7 shows rearrest, reconviction, and incarceration data. Probationers previously convicted of homicide, sexual assault, and risk of injury to a minor had the lowest rates of rearrest and reconviction.

The highest rates of rearrest were among probationers originally sentenced for a violation of probation, a felony drug sale or possession conviction, robbery, or kidnapping. These offenders were also more likely to be reconvicted of a new crime than other types of probationers.

Total criminal activity. Within the three-year release threshold, 58 percent (6,021) of probationers were rearrested at least once. Based on the total charges for the first rearrest, the probationers committed 11,797 new felony and misdemeanor crimes.

As shown in Figure IV-14, more than half of the criminal activity involved the sale or possession of drugs, possession of a weapon, risk of injury to a minor, disorderly conduct, and minor assaults. Fifteen percent of the crimes were violent, and 20 percent were property offenses.

Rate by demographics. Figure IV-15 shows the recidivism rate based on rearrest, reconviction, and incarceration among the probationers based on their age at sentencing to probation. As shown, recidivism rates for all three measures were higher for younger probationers than any other age group. About 68 percent of all felony probationers between the ages of 16 and 21 were rearrested. Over one-third (37 percent) were reconvicted of a new crime, and 14 percent were sent to prison as a result. Consistent with the pattern among the inmate cohort group, older probationers -- those over 40 years -- had the lowest recidivism rates.

Also similar to the inmate group, male probationers had a significantly higher rate of recidivism than females. As shown in Figure IV-16, 61 percent of male inmates were rearrested for a new crime compared to 45 percent of females. Female probationers had much lower rates of reconviction and incarceration than males. Less than one-quarter of the female probationers were reconvicted of a new crime compared to 34 percent of the males. Male probationers (13 percent) were twice as likely as females (6 percent) to be sent to prison as a result of the conviction.

The recidivism rate when analyzed by the racial group among probationers showed certain minority groups had higher rates of rearrest, reconviction, and reincarceration, as with the inmate group. In Figure IV-17, African American probationers had a 67 percent rearrest rate and Hispanics a 65 percent rate compared to 49 percent for Caucasian probationers. (American Indian, Asian, and other racial groups again were not included in the analysis because they represented less than 1 percent of the total probationer sample.)

About 40 percent of minority probationers and 27 percent of Caucasians were reconvicted of a new crime. Similarly, 15 percent of minority probationers were sent to prison as a result of a conviction compared to 8 percent of Caucasians.

When the probationer's race and age at sentencing were analyzed together, the data showed young minority inmates were more likely to be rearrested. As shown in Figure IV-18, 15 percent of Caucasian probationers between 16 and 21 years old were rearrested compared to 24 percent of African Americans and 23 percent of Hispanic probationers of the same age.

Rate by other characteristics. The rates of recidivism were also analyzed by the probationers' levels of educational attainment, substance abuse, and mental health. Table IV-8 provides the percentage of inmates within each level for education, mental health, and substance abuse who were rearrested at least once. Consistent with the inmate group and national research, the lower the educational grade level or the more serious the substance abuse problem of the probationer, the higher the overall rate of recidivism. The analysis further showed only the probationer's substance abuse level was a strong predictor of rearrest while educational attainment and mental health levels were not.

The data showed one difference between the inmate and probationer groups. Probationers classified with no mental health problem had higher rates of rearrest, reconviction, and incarceration. This may be attributed to the less serious nature of the characteristics and criminal activity of the probationers as a group.

Patterns of repeat criminal activity. The pattern of repeat criminal behavior among the probationer cohort group members was analyzed to determine if they are more likely to "specialize" in certain types of crimes. The definitions and methodology used to conduct this analysis for the inmate group were also used for the probationers.

Table IV-9 shows the: (1) total number of probationers sentenced in 1997 by their primary offense type; (2) number rearrested for the same exact crime as their primary offense; (3) number rearrested not for the exact same crime type, but for one in the same crime category; (4) number rearrested for any other crime; and (5) number not rearrested within the three-year release threshold. The limitations to this data are the same as those set forth in the analysis of the inmate cohort group.

Like the inmate group, most probationers were not rearrested for the exact same crime that resulted in their 1997 conviction and sentence to probation. As the table shows, probationers are even less likely than inmates to "specialize" in a certain type of crime.

Probationers originally convicted of burglary had the highest rate of rearrest (27 percent) for the same crime. One-third of probationers originally sentenced for a weapons violation were rearrested for a similar crime, about 15 percent of those with a prior forgery or fraud conviction were rearrested for a new forgery or fraud crime, less than 10 percent of probationers with a prior arson conviction were rearrested for arson, and only 3 percent of those convicted of selling drugs were rearrested for a drug-sale crime.

While most probationers were not rearrested for a new crime in the same category as their primary offense (e.g., violent, property, drug, other), there were some interesting patterns. About one-quarter of drug offenders were rearrested for a drug sale or possession offense and 21 percent of property offenders recommitted a property crime such as burglary or larceny. The data showed 14 percent of violent probationers were rearrested for another violent crime, predominantly an assault on another person.

There was also no strong pattern of "specialization" among probationers with multiple rearrests during the three-year period under review. Like the inmate group, only property and drug offenders showed a likelihood to recommit the same type of crimes, and violent offenders were less likely to be rearrested for another violent crime. On the whole, the new criminal activity was nonviolent and consisted of less serious felonies or misdemeanor property crimes, drug possession offenses, and a wide range of crimes such as disorderly conduct, breach of peace, and motor vehicle infractions.

Severity of repeat crime. Table IV-10 shows the severity of the crimes for which the probationer group was rearrested. The same categories used for the analysis of the inmate data -- felony, misdemeanor, and infraction -- were used for this group. Like the inmate group, most of the violent crimes committed by probationers were felonies. In fact, the pattern was stronger among this group in that all of the violent crimes, except for arson, were felonies. However, as shown above, only a small percentage of probationers were rearrested for a violent crime.

Most of the probationers were rearrested for a property crime. As the data showed, all were misdemeanors.

All of the drug sale offenses were felonies, but about three-quarters of the drug possession crimes were misdemeanors. Only a small percentage of the repeat criminal activity were infractions and most were motor vehicle offenses.

Number of new arrests. The total number of rearrests during the three-year release threshold for each probationer was calculated and is set forth in Table IV-11. More than 40 percent of the probationers were not rearrested. Of the 58 percent that were rearrested, there were on average almost two (1.8) rearrests per offender. The number of rearrests ranged from one to 36 during the three-year release threshold.

When analyzed by the offenders' primary offenses, about half of the probationers within each crime category had no more than two rearrests during the three-year period. Property offenders and probation violators tended to have the most rearrests. Almost 20 percent of property offender and 16 percent of VOP offenders had six or more rearrests.

As with the inmate data, it is import to consider probationers with only one rearrest may have been incarcerated as a result. They would, therefore, not be in the community and at risk of reoffending during the remainder of the three-year release threshold. Also, the probation cohort group generally committed less serious crimes than the inmate sample, and most were still under probation supervision from their original sentence at the time they reoffended. So, while a probationer may have multiple arrests, the type and severity of the criminal activity may not require the court to impose a term of imprisonment or the court may simply modify the conditions of the primary sentence of probation rather than impose a new sanction.

Time at risk. As for the probationer group, Figure IV-19 shows the time at risk before first rearrest was similar to the inmate group. In general, however, fewer probationers were rearrested than inmates.

Within the first six months after being sentenced to probation, 23 percent of probationers were rearrested. During the first year after being sentenced, 35 percent were rearrested; within three years after being sentenced over 50 percent have been rearrested at least once.

A key difference between the inmates and probationers should be considered when reviewing this analysis. The inmates were discharged from prison near or at the end of their sentence whereas the probationers were just beginning their sentence to probation or other type of supervision sanction. More probationers were under some form of active community supervision during the at-risk period under review.

Reconviction and sentencing. As with the analysis of the inmate data, the criminal court dispositions were categorized as guilty or not guilty. The sentences imposed included period of imprisonment or probation supervision, a fine, or other alternative or diversionary program.

Figure IV-20 shows over 30 percent of the felony offenders who were sentenced to probation in 1997 were subsequently reconvicted of a new crime during the three-year release threshold. Thirty-one percent of the reconvicted probationers were found guilty of a variety of crimes such as weapons offenses, risk of injury to a minor, perjury, tampering with a witness or evidence, and hunting and fishing violations. Twenty-six percent were reconvicted of a drug sale or possession offense, 24 percent of a property crime, and 12 percent of a violent crime.

As previously stated, an offender may not be convicted of the exact crime for which he or she was arrested. The following analysis is based on the reconviction crime and, based on the felony or misdemeanor status, the severity of that crime was analyzed. Overall, probationers were reconvicted of less serious felony and misdemeanor crimes.

Most of the rearrests for a violent crime were for an assault. Three-quarters of the reconvictions for assault were at the misdemeanor level. The majority of crimes for which probationers were rearrested were property crimes such as burglary, larceny, and forgery. About 60 percent of the subsequent reconvictions for those crimes were misdemeanors. All of the reconvictions for a weapon possession, risk of injury to a minor, and drug sale offenses were felonies.

As shown in Figure IV-21, given the less serious nature of the crime committed by the probationer group, only 11 percent of the reconvicted probationers (N=3,328) were incarcerated for a new crime. The average prison term imposed was less than one year (nine months), and ranged from three days to 30 years.

Almost one-quarter were sentenced to another period of probation supervision, which averaged approximately two years. Many probationers reconvicted of a new crime, especially a misdemeanor, did not receive a new sentence. Instead, the court modified or extended the original sentence to probation.

Program participation. Participation in residential and nonresidential community-based programs was reviewed for 1,211 probationers. The court often orders an offender to participate in a program as a condition of release to community supervision. Probation officers can also modify the court order by referring an offender to a program or service to assist him or her to successfully complete the sentence.

The judicial branch contracts for a statewide network of rehabilitative, treatment, and service programs including: alternative and day incarceration centers; alcohol and drug evaluation and treatment; domestic violence, family, and women and children services; intensive youth services; mental health services; post-release supervision; sex offender treatment; academic and vocational education; and residential programs. In addition, offenders may seek private treatment at their own expense.

Table IV-13 compares the rates of rearrest among probationers who participated in a community-based residential or nonresidential program to those probationers who did not participate. Similar to the inmate sample, the data show program participation had no positive effect on the probationer's rate of rearrest, except for the day incarceration center program. A day incarceration center is an alternative sanction program that requires offenders to report to and remain at the center for a specified period each day -- some stay for a few hours and others remain for an extended period. During the day, the offenders participate in structured activities and treatment programs. Often times this program is mandated for offenders who are not employed or attending an educational program. This program significantly reduced the rate of rearrest among probationers who participated -- only one-quarter were rearrested.

Substance abuse treatment, intensive youth services, and residential programs such as halfway houses also had a reduced rate of rearrest among participants. These are structured, intensive programs that typically service offenders with more serious problems and/or criminal histories. Also, the offenders who sought -- and paid for -- private treatment had a lower rate of rearrest.

Similar to the inmate sample, the rearrest pattern among probationers who participated in programs appears to be at odds with what would be expected. Again, an explanation may be participation in most of the programs is neither a factor in increasing or decreasing the likelihood of rearrest. The programs may not be targeting the appropriate offenders or providing the services necessary to overcome the other causes of repeat criminal activity. This analysis highlights the need for a comprehensive evaluation of the efficacy and cost benefits of prison and community-based treatment, rehabilitation, and service programs.

⁹ Committee staff used various methods to analyze the data including a regression analysis. Regression analysis is a method of determining whether there is a statistically significant relationship between two or more variables. The analysis then attempted to determine which offender demographic or criminal history characteristic was a predictor of recidivism.

Return to Year 2001 Studies

Return to Table of Contents