U.S. Department of Justice Office of Justice Program Bureau of Justice Statistics Special Report Federal Justice Statistics Program Offenders Returning to Federal Prison, 1986-97 September 2000, NCJ 182991 ------------------------------------------------------------------ This file is text only without graphics and many of the tables. A zip archive of the tables in this report in spreadsheet format (.wk1) and the full report including tables and graphics in .pdf format are available from: http://www.ojp.usdoj.gov/bjs/abstract/orfp97.htm ------------------------------------------------------------------ by William J. Sabol, Ph.D. William P. Adams Barbara Parthasarathy Yan Yuan The Urban Institute -------------------------------------------------------------------- Highlights * 16% of the 215,263 offenders released from Federal prison between 1986 and 1994 returned to Federal prison within 3 years of release. * The proportion of offenders returning to Federal prison within 3 years increased from 11.4% of those released during 1986 to 18.6% of those released during 1994. * 60% returned following a technical violation of release conditions, 30% following a conviction for a new offense, and 10% for other violations. * Of the 33,855 offenders returning to prison between 1986 and 1997, 54% returned within 1 year of being released; an additional 34% returned within 2 years of being released. About 12% returned after 2 to 3 years. * Offenders convicted of a violent offense returned to prison at a higher rate (32% of releases) than those convicted of a property (17%), public-order (15%), or drug (13%) offense. * Offenders who served 5 years or more in Federal prison were more likely to return to Federal prison (25%) than those who served terms of less than 5 years (15%). * Offenders returning to prison for any reason served an additional 16 months, on average, in prison. Those committing technical violations served 12 months, on average; those convicted of a new offense, 25 months. * Violent offenders returning to prison for any reason served an additional 28 months; drug offenders, 20 months; property offenders, 12 months; and public-order offenders, 11 months. -------------------------------------------------------------------- Between 1986 and 1994, 215,263 offenders were released from Federal prison for the first time. Of those offenders, 16% (or 33,855) returned to Federal prison within 3 years. The proportion of offenders returning to prison increased from 11.4% of those released during 1986 to 18.6% of those released during 1994. Among offenders released between 1986 and 1992, the proportion that returned to Federal prison within 3 years increased from 11.4% to 18.2%; of those released between 1992 and 1994, the proportion increased from 18.2% to 18.6%. Information is not available on offenders who, when released from Federal prison, were subsequently imprisoned for a State offense without an intervening Federal revocation. During the 1980's Federal sentencing policy and practice underwent a series of changes. The Sentencing Reform Act of 1984 (SRA), which went into effect on November 1, 1987, in part, established sentencing guidelines that abolished parole, required a determinate prison term, reduced the amount of good conduct time that Federal offenders could earn, and required Federal offenders to serve the entire sentence imposed, less up to 54 days of any good conduct time earned per year served. Before implementation of the SRA, Federal offenders served about 58% of the sentence imposed, on average, in prison. ***Footnote 1: Time Served in Prison by Federal Offenders, 1986-97, BJS Special Report. NCJ 171682, 1999.*** After serving a certain portion of their sentence in prison, offenders could be paroled by the U.S. Parole Commission and supervised in the community by Federal probation officers. The Parole Commission could reimprison offenders violating the conditions of their parole or committing new crimes for up to the unserved portion of their sentence. (If convicted of a subsequent crime, the offender could receive a separate sentence for the new crime.) Parole guidelines, however, provide for re-parole following revocation (see 28 CFR Sec. 2.21). Following implementation of the SRA, Federal offenders could expect to serve approximately 87% of the sentence imposed, assuming they earned all available good conduct time. Following release from prison, most of these offenders are required to serve a term of "supervised release." Federal supervised release is similar to parole insofar as the offender is required to abide by court-imposed conditions and Federal probation officers monitor compliance with those conditions. Unlike parole, supervised release has a term that must be established at the time of sentencing in accordance with the Federal sentencing guidelines (see U.S.S.G., Chapter 7, Part D). Offenders on supervised release can be reincarcerated for violations of release conditions or new crimes. The prison term imposed for the violation is in addition to any sentence previously imposed and served for the original offense and any sentence imposed for a subsequent crime. During 1998, 98% of offenders sentenced to a term of imprisonment were also sentenced to a term of supervised release (not shown in a table). The average term of supervised release imposed during 1998 was 41 months. Offenders released from prison during 1986 and 1987 were sentenced under laws in effect prior to the implementation of the Federal sentencing guidelines ("old law") and, therefore, were eligible for parole. The first cohort of 3,202 offenders sentenced pursuant to the SRA ("new law") entered supervised release during 1988. By 1994, 90% of the offenders released from Federal prison had been sentenced pursuant to the SRA and were not eligible for parole (not shown in a table). -------------------------Box on page 2-------------------------------- New-law offenders had higher return rates compared to old-law offenders Offenders sentenced pursuant to the Federal sentencing guidelines returned to Federal prison at a greater rate than other offenders. Overall, 18% of the 103,686 new-law offenders released from Federal prison between 1988 and 1994 returned within 3 years of release (see figure below). Of the 111,577 old-law offenders who were released between 1987 and 1994, 14% returned. For both new- and old-law offenders, the annual return rate increased over time. For new-law offenders the proportion returning to Federal prison increased from 12% of those released during 1988 to 19% of those released during 1994. For old-law offenders, the proportion who returned increased from 11% of those released during 1986 to 16% of those released during 1994. New-law offenders returned at a greater rate for each offense category than old-law offenders. New-law offenders convicted of fraud and weapons offenses were almost twice as likely as old-law offenders to return to Federal prison within 3 years of release: 17.9% of new-law fraud offenders returned compared to 9.2% of old-law offenders; and 31.3% of new-law weapon offenders returned compared to 16.3% of old-law offenders -------------------------End of box--------------------------------- Violent offenders returned to Federal prison at higher rates than other offenders Offenders originally convicted of violent offenses (such as robbery) were more likely to return to Federal prison within 3 years of their release than offenders who had been convicted of property, drug, or public-order offenses. About 32% of violent offenders released from Federal prison during this period returned within 3 years of release compared to 13% of drug offenders. Defendants convicted of robbery were among those most likely to return to Federal prison: 36% of the 8,880 robbery offenders released from prison between 1986 and 1994 returned within 3 years. By contrast, about 13% of those convicted of either fraud or drug offenses returned. Within public-order offenses, offenders convicted of weapon offenses returned to prison (24%) at a higher rate than those convicted of immigration (15%) or other public-order (11%) offenses. Of the 10,354 offenders who returned to Federal prison for new crimes committed while on parole or supervised release, 81% committed the same type of offense for which they were originally imprisoned (not shown in a table). Violent offenders of all racial groups returned to prison at about the same rate – about 33% returning within 3 years. For other offense categories,however, nonwhite offenders returned to prison at a higher rate than whites. For instance, among property offenders, 26% of blacks and 19% of other nonwhites returned to prison compared to 13% of whites. Among drug offenders, 22% of blacks and 15% of other nonwhites returned compared to 11% of whites. Men, blacks, and younger adults returned to Federal prison at higher rates than other offenders; Hispanics at lower rates Comprising about 10% of the 215,263 offenders released from Federal prison between 1986 and 1994, women returned to prison at a lower rate than men. About 12% of the women released from Federal prison returned within 3 years compared to 16% of the men. Within each offense category, a higher percentage of men than women returned to Federal prison. Blacks comprised 19% of all offenders released between 1986 and 1994. Overall, blacks and other nonwhite offenders returned to Federal prison at a higher rate than white offenders: 24% of black offenders and 22% of other nonwhite offenders returned, compared to 13% of white offenders. Hispanics comprised 38% of offenders released. Hispanic offenders returned to Federal prison at a lower rate than did non-Hispanics: 14% of Hispanic offenders and 17% of non-Hispanics returned. More than half of the 81,093 Hispanic offenders released from Federal prison were convicted of immigration offenses; about a third, of drug offenses (not shown in a table). About 15% of those convicted of immigration offenses and 11% of those convicted of drug offenses returned to Federal prison. Persons between ages 21 and 40 comprised 69% of offenders released. These offenders returned to prison at a higher rate than offenders of other age groups: 18% of those age 21 to 40, 14% of those under age 21, and 11% of those older than 40 at the time of release returned to Federal prison within 3 years. Noncitizens, more than a third of offenders released from Federal prison, returned to prison at a lower rate than citizens Noncitizens comprised almost 38% (80,992) of the 215,263 offenders released from Federal prison. Noncitizens returned to prison at a lower rate than did U.S. citizens: about 11% of noncitizens returned to Federal prison within 3 years of release compared to 19% of citizens. Except for immigration offenses, noncitizens were about half as likely to return to Federal prison within 3 years as citizens (not shown in a table). Among violent offenders, 33% of citizens returned to Federal prison compared to 15% of noncitizens; among property offenders, 18% compared to 9%; and among drug offenders, 17% compared to 6%. Several factors account for noncitizens' lower rate of return At least two factors unrelated to the behavior of offenders on release contribute to a generally lower risk of return for noncitizens than citizens. Following release from prison, noncitizens may be deported or detained pending deportation by the Immigration and Naturalization Service. Deported noncitizens are not under the supervision of Federal probation officers and, therefore, actions that would otherwise constitute violations do not lead to reincarceration. Similarly, noncitizens cannot be returned for new crimes unless they reenter the United States, and that reentry itself would constitute a new Federal offense. Noncitizens were generally less likely to be released to supervision than other prisoners. Between 1986 and 1994, about 39% of noncitizens were released to post-incarceration supervision compared to 74% of citizens (not shown in a table). Of those released to supervision, 9% returned to Federal prison within 3 years. Of those released without a supervision requirement, 12% returned within 3 years. Return rates for technical violations of post-release supervision exceeded the return rates for new crimes About 61% of the 215,263 offenders released from prison were released to some form of post-incarceration supervision – either parole (48%) or supervised release (52%). Of the 33,885 offenders who returned to prison between 1986 and 1997, 60% violated their conditions of supervision; 30% returned following a conviction for a new offense; and the remaining 10% returned for other violations. Eighteen percent of the offenders released to supervision returned to prison, as compared to 12% of those who were released without supervision. Of those released to supervision, 16% returned for technical violations and 3% returned for new crimes. Some of those offenders who were identified as returning to prison for technical violations may have actually committed new crimes. A condition of supervision is that offenders not commit any new crimes. Those who are arrested for new crimes are often initially returned for technical violations. About 3% of those offenders whose records indicated a return for a technical violation actually committed a new crime. Eight percent of offenders released without a supervision requirement returned for new crimes and 4% returned for other reasons. Old-law offenders who were released to parole returned to prison at a lower rate than new-law offenders on supervised release: 16% of old-law offenders returned compared to 20% of new-law offenders. The difference in the rate of return for old-law and new-law offenders was primarily due to differences in return rates for technical violations. Less than 14% of the old-law offenders returned to prison for technical violations compared to 17% of the new-law offenders (not shown in a table). ------------------------Box on p4--------------------------------- Technical violations of release conditions Between 1992 and 1998, more than a third of offenders whose supervision was terminated unsuccessfully for a technical violation failed at least one drug test (Compendium of Federal Justice Statistics, 1992-98, and Federal Offenders Under Community Supervision, 1987-96). Additionally, 17% failed to keep scheduled appointments with their probation officer or make scheduled court appearances. Violent and weapons offenders were more likely than other offenders to be returned for use of controlled substances. The remaining 45% of offenders had their supervision terminated for other technical violations of release conditions. While no data are available to describe the nature of these violations, they may include failure to comply with any of the standard conditions of supervision such as having regular employment, payment of fines and/or special assessments, and limiting association with known felons (see U.S.S.G. Sec. 5B1.3). ---------------------End of box------------------------------------ As length of time served in prison increased, so did the rate of return to prison Imposed prison sentences reflect both the seriousness of the offense committed and the risk that the offender will recidivate, based on the offender's criminal history. The rate at which offenders returned to Federal prison increased with the amount of time that they had served in Federal prison. Overall, offenders who served shorter prison terms were less likely to return to Federal prison within 3 years of release than those who served prison sentences of more than 5 years. About 14% of those who served less than 1 year in prison returned to Federal prison compared to 25% of those who served 5 or more years. New-law offenders who served less than 3 years were more likely to return to Federal prison than old-law offenders who served an equal amount of time. Between 1988 and 1997, 18% of the released new-law offenders returned to prison compared to 11% of old-law offenders. Offenders released to supervision remained free longer than other offenders Offenders who returned to Federal prison remained free an average of 15.5 months. Generally the number of months that returning offenders were free did not vary greatly by offense category. Violent offenders who returned were free for an average of 14.7 months; property offenders, 15.4 months; and public-order offenders, 14.1 months (not shown in a table). However, drug offenders who returned to Federal prison were free for the longest (17.7 months). Offenders who were released to supervision remained free longer than those offenders who were not released to supervision. Offenders released to either parole or supervised release returned to prison after being free 16 months, on average. Of these, offenders who returned for a new crime returned after 17.1 months, and those who returned for a technical violation returned after an average 15.8 months. By contrast, offenders who were released without a supervision requirement, such as old-law offenders who served their full term, were returned for a new crime after 13.3 months. Offenders returning to prison following revocation served an additional 16 months in Federal prison Old-law offenders returning to prison are eligible to serve that portion of their original sentence that was not served in prison. Federal parole guidelines, however, base the revocation term according to the type of violation: offenders committing technical violations could be returned for up to 9 months; offenders engaging in new crimes could be required to serve a sentence corresponding to the parole guideline's offense severity score (see 28 CFR Sec. 2.21). For instance, an offender returning for a lower-level drug offense (involving 1 to 5 grams of cocaine) or a fraud offense (involving up to $20,000) could be returned to Federal prison for a minimum of 1 to 14 months. By contrast, an offender committing another robbery could be returned for a minimum of 24 to 36 months. Because new-law offenders have served the entire sentence imposed, those returning to prison as a result of a revocation of supervision are sentenced to additional term of imprisonment in accordance with the Federal sentencing guidelines (see, generally, U.S.S.G. chapter 7). Offenders whose supervision was revoked for technical violations or misdemeanors received shorter sentences upon revocation than those committing felonies, generally, and violent or drug offenses, more specifically (see U.S.S.G. Sec. 7B1.4). For instance, the minimum time an offender returning for a technical violation or a misdemeanor crime could expect to serve is 3 to 9 months. Offenders returning for felonies other than violent crimes or drug trafficking offenses could expect to serve a minimum of 4 to 10 months. Offenders returning for violent crimes or drug offenses could expect to serve a minimum of 12 to 18 months or 24 to 30 months, depending on the severity of the original offense. Pursuant to both the Federal parole and sentencing guidelines, offenders with prior convictions would be eligible for longer terms upon return than offenders for whom the current conviction was their first. Offenders who returned to Federal prison between 1986 and 1997 served an additional 16 months in prison, on average. Violent offenders served the longest (28 months) prison terms upon return; drug offenders served 20 months; property offenders, 12 months; and public-order offenders, 11 months. Within all offense categories except immigration offenses, offenders who returned for technical violations served fewer months (12 months) than those who returned for new crimes (25 months). Robbery offenders who returned for a new crime served the longest additional terms (114 months), followed by returning drug offenders (71 months). Overall, new-law offenders who returned to prison served longer terms than old-law offenders. New-law offenders served about 17 months, on average, upon return compared to 15 months for old-law offenders. For each offense type other than property, new-law offenders served longer sentences following revocation than old-law offenders. For example, * New-law violent offenders served 39 months upon return compared to 21 months for old-law violent offenders. * New-law drug offenders served 24 months upon return compared to 15 months for old-law drug offenders. New-law fraud offenders, however, served less time upon revocation than old-law offenders: 11 months for new-law offenders compared to the 14 months for old-law offenders. New-law offenders who returned for new crimes served 26 months as compared to 22 months for old-law offenders who returned for new crimes. New-law offenders who returned for committing a technical violation served shorter terms than old- law offenders. New-law offenders who returned for technical violations served 10 months compared to 14 months for old-law offenders. Methodology Data sources The primary source of the data for the tables presented in this report is the extracts from the BJS Federal Justice Statistics Program database, which is constructed from source data files provided by the Executive Office for United States Attorneys, the Administrative Office of the United States Courts, the United States Sentencing Commission, and the Federal Bureau of Prisons. The data for offenders "first released" from Federal prison are limited to those inmates who served a term of imprisonment resulting from a conviction in U.S. district court. Arrestees, boarders, and other detainees are excluded from the counts of first releases. Unless specified in the table, the data for offenders returning to Federal prison include all offenders who were returned for any reason, including those who entered prison as the result of a supervision violation or conviction for a new Federal offense. Scope of the analysis This study counts the number of Federal prisoners who returned to Federal prison after first release from a U.S. district court commitment. It therefore provides a limited measure of overall recidivism by Federal offenders. The measure excludes prisoners who, after first release from Federal prison, subsequently enter State prisons or jails. To minimize the impact of limited data on returns to prison, offenders released during 1994 represent the final cohort included in the study. By ending with the 1994 cohort, offenders can be tracked for at least 3 years following release. Identifying and counting "first returns" to prison In this report, offenders released from Federal prison for the first time between 1986 and 1994 are identified. For each such first release, the FJSP database on Federal prisoners was searched from the date of release forward to find the next record for a released offender. Personal identifying information such as names, register numbers, and other individual information was used to link individual release records with their first return record. Data of return records were inspected to ensure that they came after a release date; other information was analyzed to ensure tht it was consistent with the data in the first release record. First returns were defined as returns to Federal prison that occurred within 3 years of a first release. Alternative observation windows were selected to observe first returns. A 1-year and a 2-year observation window provided shorter time frames for analyzing returns to prison. The 3-year window was selected for two reasons: (1) to ensure consistency with prior BJS publications on recidivism of State prisoners in which a 3-year window was used to measure recidivism; and (2) the return rate increases with the length of the observation window.***Footnote 2: See Recidivism of Prisoners Released in 1983, BJS Special Report. NCJ 116261, 1989.*** However, as the difference between the 2- and 3-year rates is smaller than the difference between the 1- and 2-year rates, the rate of increase in return rate decreases with the length of the observation window. This difference diminishes further when comparing the 3- to 4-year return rate (appendix figure 1). Hence, the 3-year rate provides a tradeoff between a marginal increase in the precision of the estimate of the number of returns and the length of the period of time to observe first releases from prison. Unit of analysis The unit of analysis in the study is a first release from Federal prison. Return rates are calculated based on the number of first releases. A person may be counted more than once if that person has more than one first release. For example, an offender first released during 1986 may return to prison for a violation of parole and subsequently be released for the second time. If this person were to return to prison again but on a U.S. district court commitment and then were released from this commitment, this second episode of entry into prison would be counted as a first release. This particular offender would have two first releases. Were he/she to return on the second first release, he/she would be counted a second time as a first return. Conversely, if the offender returns to prison because of a violation, is released from that return, and returns a second time on another violation of supervision, that offender is counted once as a first release and once as a first return. His/her release from the technical violation (his/her second release) is not counted, nor is his/her return for the second violation not counted. Measuring time served upon return For persons who returned to Federal prison, time served is the number of months that an offender served before being released from prison following revocation. For offenders who were released during the study period, time served is the number of months they actually served. For offenders still in prison, time served is estimated as follows -- * For new-law offenders, time served is the difference between the projected release date and the commitment date. If the projected release date is missing, time served is 87% of the prison term imposed. * For old-law offenders, time served is the difference between the projected release date and the commitment date. If the projected release date is missing, time served is the product of the proportion of the sentence served for the original commitment and the term imposed for the second commitment. Comparing returns to prisons with unsuccessful terminations of supervision Other BJS reports, such as the Compendium of Federal Justice Statistics and Federal Offenders Under Community Supervision, 1987-96, provide data on the number of offenders terminating supervision for technical violations or new crimes. The data in those reports are not directly comparable with the data in this report for several reasons. * This report describes returns to prison for annual cohorts of offenders released from Federal prison for the first time from a U.S. district court commitment. Other reports use other data to describe a different universe of offenders, the entire population of offenders at risk of terminating post-incarceration supervision. * The population of offenders on which this report is based includes offenders released from prison for the first time in a given year, but other reports also include offenders who were released from prison on other than their first release and offenders released from prison during previous years. * The methods for classifying the type of violation of supervision differ between those used in this report and those used in other reports. In other reports, termination of supervision by reason of a new crime includes offenders who violated State laws and offenders who receive probation or a fine for the new crime. In this report, new crimes refer to offenders who violated Federal statutes and returned to prison for this reason. * In other reports, termination of supervision for technical violations can result in the imposition of new conditions of supervision. In this report, terminations of supervision for reason of technical violation refers only to those offenders who returned to prison for this reason. * Some offenders who return to prison for a new crime are classified initially in the Federal Bureau of Prisons information system as technical violators while they await a hearing to determine if they violated conditions of supervision or pending the outcome of their new trial. After the final disposition of their hearing or case, their records are updated to reflect the technical violation or new crime as the reason for commitment. Definitions of concepts First release. A first release is defined as the first release from a U.S. district court commitment. Technical violation or supervision violation. A technical violation is a violation of a particular condition of supervision. For serious or repeated violations of his supervision conditions, an offender can be returned to prison. New crime. A new crime is the commission of a new criminal offense, whether the offender is under supervision at the time for a different offense, or when the offender is not under supervision. Data from the Federal Justice Statistics Program are compiled by the Bureau of Justice Statistics from the source files of several Federal agencies. Data can be obtained on CD-ROM from the Bureau of Justice Statistics Clearinghouse, 1-800-732-3277, or from the Federal Justice Statistics Resource Center located on the Internet: http://fjsrc.urban.org The Resource Center, as well as the report and supporting documentation, are also accessible through the BJS website: http://www.ojp.usdoj.gov/bjs/ Data from the FBI's Uniform Crime Report program can be obtained from the National Archive of Criminal Justice Data at the University of Michigan. The National Archive is accessible through the BJS website. Recent editions of the FBI's annual report Crime in the United States are accessible through the FBI's website: http://www.fbi.gov/ucr.htm The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. Jan M. Chaiken, Ph.D., is director. BJS Special Reports address a specific topic in depth from one or more datasets that cover many topics. William J. Sabol, Ph.D., William P. Adams, Barbara Parthasarathy, and Yan Yuan of The Urban Institute prepared the report. Gerry Gaes, Ph.D., of the Federal Bureau of Prisons provided substantive review. Tom Hester and Ellen Goldberg produced the report, and Jayne Robinson prepared it for final publication. September 2000, NCJ 182991 End of file 9/7/00 pm