U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Compendium of Federal Justice Statistics, 1998 May 2000, NCJ 180258 ------------------------------------------------------------------ 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/cfjs97.htm This report is one in a series. More recent editions may be available. To view a list of all in the series go to http://www.ojb.usdoj.gov/bjs/pubalp2.htm#cfjs ------------------------------------------------------------------- Jan M. Chaiken, Ph.D. Director, BJS This Bureau of Justice Statistics Report was prepared by the Urban Institute under the supervision of Steven K. Smith and John Scalia, Jr., of the Bureau of Justice Statistics (BJS). The project is supported by BJS grant number 98-BJ-CX-K015. Principal staff for the project at the Urban Institute were William J. Sabol, Ph.D., William Adams, David Kirk, Barbara Parthasarathy, Harvey Meyerson, Yan Yuan, and Christine Arriola. Layout and design were by David Williams. Tom Hester of BJS provided editorial review. This report is made possible through the cooperation of the following Federal agencies and their staffs: The United States Marshals Service (USMS), the Administrative Office of the United States Courts (AOUSC), the Executive Office for U.S. Attorneys (EOUSA), the U.S. Bureau of Prisons (BOP), the United States Sentencing Commission (USSC), and the Federal Judicial Center (FJC). The staff who provided expert advice about the source records include: Steven Schlesinger, Catherine Whitaker, Karen O'Conor, and Pragati Patrick (AOUSC); Barbara Tone (EOUSA); Gerald Gaes and Nancy Miller (BOP); Richard McNeil (USSC); David Rauma and George Cort (FJC); and Joe Briggs (USMS). The contents of this document do not necessarily reflect the views or policies of the Bureau of Justice Statistics or the U.S. Department of Justice. BJS authorizes any person to reproduce, publish, translate, or otherwise use all or any part of the copyrighted material in this publication; citation to source, however, is appreciated. An electronic version of this report and the data analyzed in the report may be found on the Internet at the following address: http://www.ojp.usdoj.gov/bjs/. Or visit the BJS-sponsored Federal Justice Statistics Resource Center (FJSRC) web site at: http://fjsrc.urban.org/, which provides online access to the FJSP database for conducting interactive tabulations of the FJSP data and downloading datasets. Contents Highlights Introduction Organization of the Compendium Modifications to the 1998 Compendium Notes to reader System overview Chapter 1: Arrests for Federal offenses Arrests by offense category Arrests by law enforcement agencies Arrests across demographic groups Chapter notes Chapter 2: Prosecution U.S. attorneys' decisions Suspects in matters received Suspects in matters concluded Chapter notes Chapter 3: Pretrial release Types of pretrial release Factors relating to release or detention Pretrial outcomes by offense categories Pretrial outcomes across demographic groups Length of pretrial detention Chapter notes Chapter 4: Adjudication Defendants in cases filed Defendants in cases terminated Case processing times Convictions by U.S. magistrates Characteristics of convicted defendants Chapter notes Chapter 5: Sentencing Offenders convicted and sentences imposed Average prison sentences imposed Relationship between sentence imposed and mode of conviction Characteristics of offenders sentenced to prison Chapter notes Chapter 6: Appeals Appeals filed Appeals terminated Chapter notes Chapter 7: Corrections Supervision Federal offenders under supervision Outcomes of offenders completing supervision Characteristics of offenders completing supervision Admissions, releases, and standing population of Federal prisoners Federal prisoners: First releases and time served Characteristics of Federal prisoners Chapter notes Methodology The Federal justice database Table construction and interpretation Offense classifications Figure S.2. Source agencies for Compendium data tables Glossary Tables Chapter 1: Arrests for Federal offenses, 15 October 1, 1997 - September 30, 1998 1.1. Arrests for Federal offenses, by offense 1.2. Arrests for Federal offenses, by Federal law enforcement agency 1.3. Characteristics of Federal arrestees Chapter 2: Prosecution, 25 October 1, 1997 - September 30, 1998 2.1. Suspects in matters received by U.S. attorneys, by offense 2.2. Disposition of suspects in matters concluded, by offense 2.3. Suspects in matters concluded and declined, by investigating agency 2.4. Basis for declination of prosecution by U.S. attorneys 2.5. Disposition of matters declined for prosecution by U.S. attorneys, by offense 2.6. Mean and median processing times from receipt to filing or declination, by offense Chapter 3: Pretrial release, 39 October 1, 1997 - September 30, 1998 3.1. Type of pretrial release, by offense 3.2. Type of pretrial release, by defendant characteristics 3.3. Form of pretrial detention, by offense 3.4. Form of pretrial detention, by defendant characteristics 3.5. Pretrial detention hearing outcomes, by offense 3.6. Pretrial detention hearing outcomes, by defendant characteristics 3.7. Behavior of defendants released prior to trial, by offense 3.8. Behavior of defendants released prior to trial, by type of release 3.9. Behavior of defendants released prior to trial, by defendant characteristics 3.10.Length of pretrial detention, by form of release or detention, and most serious offense charged Chapter 4: Adjudication, 53 October 1, 1997 - September 30, 1998 4.1. Defendants in cases commenced, by offense 4.2. Disposition of cases terminated, by offense 4.3. Time from filing to disposition of cases terminated 4.4. Dispositions by U.S. magistrates 4.5. Characteristics of convicted offenders Chapter 5: Sentencing, 65 October 1, 1997 -- September 30, 1998 5.1. Sentence types in cases terminated, by offense 5.2. Type and length of sentences imposed, by offense 5.3. Sentences imposed on convicted offenders, by offense of conviction and method of disposition 5.4. Convicted offenders sentenced to incarceration, by offense and offender characteristics 5.5. Average incarceration sentence lengths imposed, by offense and offender characteristics 5.6. Median incarceration sentence lengths imposed, by offense and offender characteristics Chapter 6: Appeals, 77 October 1, 1997 - September 30, 1998 6.1. Criminal appeals filed, by type of criminal case and offense 6.2. Criminal appeals filed and criminal appeals terminated, by offense 6.3. Criminal appeals terminated, by type of criminal case and offense 6.4. Disposition of criminal appeals ter- minated on the merits, by offense 6.5. Criminal appeals cases terminated on the merits, by nature of offense Chapter 7: Corrections, 89 October 1, 1997 - September 30, 1998 7.1. Federal offenders under supervision, by offense 7.2. Characteristics of Federal offenders under supervision, by offense 7.3. Outcomes of probation supervision, by offense 7.4. Characteristics of offenders termi- nating probation supervision 7.5. Outcomes of supervised release, by offense 7.6. Characteristics of offenders termi- nating supervised release 7.7. Outcomes of parole, by offense 7.8. Characteristics of offenders termi- nating parole 7.9. Admissions and releases of Federal prisoners, by offense 7.10. Characteristics of Federal prison population, by major offense category 7.11. Number of first releases from Federal prison, by release method and sentence length 7.12. Average time to first release and percent of sentence served, for prisoners released by standard methods 7.13. Mean time served to first release, by length of sentence imposed, offense, and type of case, for prisoners released by standard methods 7.14. Percent of sentence served to first release, by length of sentence imposed, offense, and type of case, for prisoners released by standard methods 7.15. Characteristics of first releases from prison, by offense, all releases 7.16. Mean time served to first release from Federal prison for prisoners released by standard methods, by offense and offender characteristics --------------------------------------------------------- Highlights The number of suspects investigations by U.S. attorneys increased by 5% between 1997 and 1998, from 110,034 to 115,692; between 1990 and 1998, the number of investigations increased by 15%. About three-quarters of those for which the investigation was concluded were prosecuted -- either before a U.S. district court judge (61%) or before a U.S. magistrate (12%) -- and 27% of those investigated were not prosecuted by U.S. attorneys. The number of defendants prosecuted in Federal courts increased by 12% between 1997 and 1998, from 69,351 to 78,172. Almost 40% of the increase was attributable to the increase in the number of defendants prosecuted for drug offenses; 29% was attributable to the increase in the number prosecuted for immigration offenses. The number of offenders under Federal correctional supervision increased 42% between 1990 and 1998. At the end of fiscal year 1998, the number of offenders under Federal correctional supervision was 201,693 compared to 141,790 during 1990. At the end of fiscal year 1998, the number of Federal inmates serving a sentence of imprisonment surpassed 100,000 for the first time -- increasing by 9,670 to 108,925. The number under community supervision was 92,768. Almost two-thirds of those under community supervision were on post-incarceration supervised release (54,822) or were on parole (5,827). Arrests During 1998, 106,139 offenders were arrested by Federal law enforcement agencies for violations of Federal law. A third of all arrests were for public order offenses, 29% for drug offenses, 16% for property offenses, 14% for supervision violations, 5% for violent offenses, and 3% to secure and safeguard a material witness. About 73% of all arrests for Federal offenses were made by agencies of the Department of Justice, while Treasury Department agencies accounted for 11% of all arrests. Within the Department of Justice, 37% of arrests were made by the U.S. Marshals Service, one-third of arrests were made by the Immigration and Naturalization Service, 15% were made by the FBI, and 14% were made by the Drug Enforcement Administration. Prosecution During 1998 U.S. attorneys initiated criminal investigations involving 115,692 suspects, and they concluded their investigations of 106,022 suspects. Thirty-five percent of the suspects were investigated for public-order, 32% for drug, 26% for property, and about 7% for violent offenses. The number of suspects investigated for immigration offenses increased by more than 50% between 1997 and 1998, from 9,366 to 14,114. Of the suspects in criminal matters concluded, U.S. attorneys prosecuted 64,993 in U.S. district courts and 12,243 were disposed of before U.S. Magistrates. Declinations by U.S. attorneys have reached an all time low of 27% of matters concluded during 1998. Suspects in criminal matters involving drug or violent offenses were slightly more likely to be prosecuted in a U.S. district court (77% and 63%, respectively) than were the suspects involved in public-order or property offenses (54% and 52%, respectively). Suspects involved in property offenses (such as fraud) or regulatory public-order offenses were more likely to be declined for prosecution (40% and 64%, respectively) than were the suspects investigated for drug or violent offenses (18% and 33%, respectively). Pretrial release During 1998 about 43% of the 60,950 defendants who terminated pretrial services were released at some time prior to their criminal trial. Defendants charged with property offenses such as embezzlement, fraud, and larceny and those charged with regulatory public-order offenses or tax law violations were more likely to be released prior to trial (between 68% and 95% of these defendants were released) than were defendants charged with violent offenses, drug trafficking, weapons, or immigration offenses (between 8% and 43% of these defendants were released). The proportion of defendants released prior to their trial decreased by nearly a third, from 62% during 1990 to 43% during 1998. Violent, drug, weapon, and immigration offenders were considerably less likely to be released during 1998 than during 1990. During 1990, 45% of violent, 54% of drug, 63% of weapon, and 30% of immigration offenders were released at some point prior to trial. By contrast, during 1998, 32% of violent, 35% of drug, 43% of weapon, and 8% of immigration offenders were released. Defendants having a prior criminal history of serious or violent crimes were less likely to be released than those without a prior criminal history; defendants with more criminal history were less likely to be released than those with less criminal history. Twenty-two percent of the defendants with a prior violent felony conviction were released before trial, while 54% of the defendants with no prior convictions were released. Forty-six percent of defendants with one prior conviction were released, as compared to 34% of defendants having two to four prior convictions and 23% of defendants having five or more prior convictions. About 84% of the defendants released prior to trial completed their periods of release without violating the conditions of their release. About 16% of defendants released violated the conditions of their release, and 6% of defendants had their release revoked. Defendants charged with drug and violent offenses were more likely to commit at least one violation of their conditions of release (27% and 18%, respectively) and to have their release revoked (10% and 9%, respectively) than were other defendants. Defendants released during 1998 were more likely to violate the conditions of their pretrial release than those released during 1990. During 1998, 16% of those released at some point prior to trial violated a condition of their release. During 1990, 12% violated their release conditions. Adjudication During 1998, 78,172 defendants were charged in Federal courts with a criminal offense, about 83% of whom were charged with felonies. Of the defendants charged with felonies, 43% were prosecuted for drug trafficking offenses, 28% for public-order offenses, 23% for property offenses, and 6% for violent offenses. The number of defendants charged with a felony immigration offense increased by 38% between 1997 and 1998, from 6,726 to 9,254. The number charged with a felony drug offense increased by 13%, from 24,693 to 28,021. Criminal cases were concluded against 69,769 defendants during 1997, 82% of whom had been charged with felonies. The proportion of defendants convicted in the Federal courts increased from 81% during 1990 to 87% during 1998. Additionally, the proportion of defendants pleading guilty increased from 88% during 1990 to 94% during 1998. Ninety percent of defendants charged with felonies were convicted. The conviction rate was about the same for all major offense categories: 91% of public-order offenders, 90% of violent, property and drug offenders. Sentencing Defendants convicted during 1998 were more likely to be sentenced to prison than those convicted during 1990. During 1998, 71% of defendants were sentenced to prison compared to 60% of those sentenced during 1990. About 92% of felony drug offenders and 91% of violent felony offenders received prison sentences during 1998, as did 83% of felony public-order offenders and 60% of felony property offenders. Average length of sentences imposed, by offense, October 1, 1997 - September 30, 1998 Most serious Average sen- offense of conviction tence length All offenses 58.8 mo Felonies 60.6 Violent offenses 84.2 Property offenses 25.4 Drug offenses 78.8 Public-order offenses 47.3 Misdemeanors 11.4 The 43,041 offenders sentenced to prison received, on average, 58.8 months of imprisonment. Offenders sentenced for violent felony offenses and felony drug offenses received longer average prison terms (84.2 and 78.8 months, respectively) than those convicted of felony property and public-order offenses (25.4 and 47.3 months, respectively). While the proportion of defendants sentenced to prison is at an all time high, average prison sentences have declined from the peak attained during 1992. During 1992 the average prison term imposed was 62.6 months; for violent felony offenses, the average term imposed was 94.8 months; for drug felonies, the average term was 84.1 months. Appeals Between 1994 and 1998, the number of appeals received by the U.S. Courts of Appeals remained relatively constant -- between 10,000 and 11,000 annually. However, the proportion of criminal defendants appealing some aspect of their decreased from 21% of convictions during 1994 to 17% during 1998. During 1998, 10,535 criminal appeals were filed. Forty-seven percent of the appeals filed challenged both the conviction and sentence imposed. Only 4% of appeals were filed by the Government. Of the 10,105 appeals terminated during 1998, 76% (or 8,178) were terminated on the merits. In 82% of the appeals terminated on the merits, the district court ruling was affirmed, at least in part. Corrections Community supervision Between 1990 and 1998, the number of offenders on community supervision increased by 9% from 84,801 during 1990 to 92,768 during 1998. While nearly equal proportions of offenders were serving terms of probation and post-incarceration supervision (parole or supervised release) during 1990, during 1998, almost two-thirds were serving a term of post-incarceration supervision (59% supervised release and 6% parole) while 35% were on probation. Drug offenders comprised 53% of offenders under terms of supervised release during 1998, 52% of offenders on parole, and 11% of offenders on probation. Property offenders comprised 38% of offenders on probation, 27% of offenders serving terms of supervised release, and 11% of offenders on parole. Admissions to Federal Bureau of Prisons, releases, and prisoners at yearend, by offense, October 1, 1997 - September 30, 1998 Most serious Population offense of conviction All admissions All releases at yearend All offenses 58,215 48,545 108,925 Violent offenses 8.9% 9.0% 11.6% Property offenses 17.9 20.4 7.9 Drug offenses 40.2 38.3 57.8 Public-order offense 31.0 30.9 20.4 A total of 16,011 offenders completed their terms of probation during 1998. Most of these offenders (81%) completed their terms of probation successfully,. Eleven percent of probationers terminating supervision during 1998 committed technical violations; 5% committed new crimes. A total of 19,878 offenders completed terms of supervised release during 1998. Of these offenders, 64% successfully completed their terms without violating conditions of release; 21% committed technical violations; and 12% committed new crimes. A total of 3,077 offenders completed terms of parole during 1998. Of these offenders, 56% successfully completed their terms without violating conditions of release; 26% committed technical violations; and 13% committed new crimes. Prison Between 1990 and 1998, the number of inmates serving a sentence of imprisonment increased by 91% from 56,989 during 1990 to 108,925 during 1998. During 1998, the number of sentenced inmates increased by 9,670 to surpass 100,000 for the first time. During 1998, 42,598 prisoners were received by the Bureau of Prisons from U.S. district court commitments. An additional 15,617 prisoners were returned to Federal prison for violating conditions of probation, parole, or supervised release, or were admitted to Federal prison from elsewhere than a U.S. district court. Drug offenders -- who comprised about 40% of persons admitted into Federal prison -- comprised the largest percentage of persons in prison (58%) at the end of 1998. During 1998, 34,504 prisoners were released for the first time from Federal prison after commitment by a U.S. district court. Of these, 31,200 were released by standard methods and 3,304 were released by extraordinary means (death, treaty transfer, sentence commutation, or drug treatment). An additional 14,041 prisoners were released from subsequent commitments to Federal prison. Time served by Federal offenders increased from 17 months during 1990 to 28 months during 1998. Additionally, the proportion of the sentence served increased from 65% during 1990 to 87% during 1998. Violent and drug offenders were among those offenders who served the longest prison terms. However, while time served for violent offenders remained relatively constant between 1990 and 1998, approximately 58 months, time served by drug offenders increased by 57%, from 26 months during 1990 to 41 months during 1998. Average time to first release, standard releases, by offense, October 1, 1997 - September 30, 1998 Most serious original Mean time offense of conviction served All offenses 28.0 mo Violent offenses 53.7 Property offenses 15.5 Drug offenses 41.4 Public-order offenses 18.1 Introduction This Bureau of Justice Statistics (BJS) report presents an overview of case processing in the Federal criminal justice system. The data presented are compiled from the BJS Federal Justice Statistics Program (FJSP) database. The FJSP database includes data provided by the Administrative Office of the U.S. Courts, the Executive Office for the U.S. Attorneys, the Federal Bureau of Prisons, the U.S. Marshals Service, and the U.S. Sentencing Commission. The Administrative Office of the U.S. Courts, in addition to providing data describing defendants in cases processed by the Federal judiciary, provides data describing defendants processed by the Federal pretrial services agencies and the Federal probation and supervision service. The BJS Federal Justice Statistics database is archived on CD-ROM, copies of which are available from the BJS Clearinghouse. The data can be downloaded from the Federal Justice Statistics Resource Center located at http://fjsrc.urban.org. Each agency reports on those defendants it processed during a given year in an annual statistical report. These agency reports are often incomparable due to the varying methods the agencies use to report case processing activities. As reported by an interagency working group, headed by BJS, the differences in the case processing statistics are attributable, in part, to the differing needs and missions of the agencies. The working group found the following differences in reported statistics: * the universe of cases reported during a given period -- some agencies report on those case processing events that occurred during a particular period, whereas others report on those events recorded during a particular period; and * many of the commonly used case processing statistics -- suspect/defendant processed, offense committed, disposition, and sentence imposed -- are defined differently across agencies. BJS, through its FJSP, has recognized the incomparability of these annual statistical reports and has attempted to reconcile many of the differences identified by the working group. For instance, by combining databases from several years, BJS is able to report on those cases that actually occurred during the reporting period. Additionally, commonly used case processing statistics are made comparable across stages by applying uniform definitions to data obtained from each agency. Because definitions in the FJSP are consistent with those categories in BJS programs describing State defendants convicted, sentenced, or imprisoned, the comparison of Federal and State case processing statistics is facilitated. The 1998 Compendium, 13th in a series which includes 1984, 1985, 1986, 1988, 1989, 1990, 1992, 1993, 1994, 1995, 1996, and 1997, describes defendants processed at each stage of the Federal justice system -- arrest by Federal law enforcement agencies (chapter 1), investigation and prosecution by the U.S. attorneys (chapter 2), pretrial release or detention (chapter 3), adjudication in the U.S. district courts (chapter 4), sentencing (chapter 5), appeal of the conviction and/or sentence imposed (chapter 6), and corrections (chapter 7) -- for the 12-month period ending September 30, 1998 (the Federal fiscal year). Prior to 1994, the Compendium was reported on a calendar-year basis. The tables presented report events that occurred during the Federal fiscal year -- October 1, 1997 - September 30, 1998. Generally, the tables include both individual and organizational defendants. Organizational defendants are not included in tables describing pretrial release and detention or tables showing defendants sentenced to incarceration. Juvenile offenders are included in the reported statistics. Felony and misdemeanor distinctions are provided where possible (see "Offense classifications" in Methodology). Organization of the Compendium Each chapter of the Compendium describes a major stage in the processing of criminal suspects and defendants. Each chapter contains Chapter notes that describe the universes of data used in the tables and information relevant to the interpretation of individual tables. The Compendium contains the following: Chapter 1. This chapter describes arrests made by Federal law enforcement agencies for violations of Federal law, including the characteristics of arrestees. Chapter 2. This chapter describes decisions made by Federal prosecutors in screening criminal matters and the characteristics of defendants in cases prosecuted or declined for prosecution. Chapter 3. This chapter describes the pretrial release and detention practices of the Federal judiciary, including the characteristics of defendants detained or released pending trial. Chapter 4. This chapter describes actions by the Federal judiciary in adjudicating defendants in cases filed by the U.S. attorneys, including the offense charged and characteristics of defendants convicted. Chapter 5. This chapter describes the sentences imposed by the Federal judiciary on convicted defendants, including the characteristics of defendants sentenced. Chapter 6. This chapter describes appeals of criminal convictions and sentences imposed in the Federal courts, including the original offense charged. Chapter 7. This chapter describes defendants under Federal correctional supervision -- probation, parole, and supervised release -- including the outcome of the supervision (successful completion or violations), admissions to and releases from Federal prison, and time served by Federal inmates. Methodology. This section describes the procedures followed in analyzing data and developing tables. Glossary. This section contains definitions for terms used in the Compendium. Since many terms used in the text and tables have specialized meanings (either because they refer to Federal law or because of reporting procedures by the Federal agencies supplying the data), readers are encouraged to check the glossary for exact definitions of tabulated data. Comparing Case Processing Statistics. This document, prepared by an interagency working group tasked to reconcile differences in Federal criminal case processing statistics, identifies and describes the major differences in the way Federal criminal justice agencies collect, tabulate, and report criminal case processing events. Modifications in the 1998 Compendium A new chapter (chapter 1), which describes arrests made by Federal agencies for violations of Federal law, has been introduced in the 1998 Compendium. The chapter includes three tables: a table showing arrests for violations of Federal law, by offense; a table showing arrests for violations of Federal law, by the arresting Federal law enforcement agency, by major offense; and a table showing characteristics of suspects arrested for violations of Federal law, by major offense. Due to the introduction of this new chapter on Federal arrests, chapter numbers for the chapters retained from the previous Compendium have been incremented by one. The prosecution chapter is now chapter 2; pretrial release and detention, chapter 3; adjudication, chapter 4; sentencing, chapter 5; appeals, chapter 6; and corrections, chapter 7. A new table has been introduced to the prosecution chapter which shows the declination rate by investigative agency and department (table 2.3). In the corrections chapter, the table showing characteristics of first releases from prison (table 7.15) and the table showing mean time served to first release (table 7.16) have changed. Rather than displaying statistics separately for new law cases as well as all cases, these tables now show only all cases, since new law cases comprise 98% of all first releases. Notes to reader The tables in the Compendium were constructed to permit valid comparisons within each table and to allow the reader to compare percentages (but not raw totals) across tables. It should be understood, however, that the total number of subjects/defendants shown in a particular table may not equal the number of subjects/ defendants involved in a particular stage of processing, since some records could not be linked and some data sources did not include information on particular data elements classified in a particular table. Data notes indicate the exact universe for individual tables. The Compendium is a statistical presentation of Federal criminal justice information with limited analyses of trends or explanatory factors underlying the statistics. Analyses of Federal justice statistics may be found in special reports and other publications, some of which are cited in the Compendium. Assessment of changing patterns in the Compendium tabulations may depend on detailed examination of subcategories not shown in the tabulations or may require other sources of information, such as knowledge of legislation or Federal agency procedures. Chapter 1. Arrests for Federal offenses Chapter notes Sixty-six Federal agencies employ full-time personnel empowered to make arrests. As of June 1998, these agencies employed about 83,000 officers. ***Footnote 1: Brian A. Reaves, Federal Law Enforcement, 1996, Bureau of Justice Statistics, Washington, DC (NCJ-164617.*** More than half (56%) of all law enforcement officers were employed by the Department of Justice. The Department of Treasury employed 24% of all officers. Offenders arrested by Federal agencies are transferred to the custody of the U.S. Marshals Service for processing, transportation, and detention. ***Footnote 2: The Federal agency making the arrest may be different from the Federal agency initiating the investigation involving the arrestee. Statistics describing agencies initiating investigations are shown in table 2.3 of chapter 2, "Prosecution."*** During 1998, the U.S. Marshals Service received 104,422 offenders for processing from the Federal law enforcement agencies, including nearly 29,000 offenders arrested by deputy U.S. marshals. Arrests by offense category (table 1.1) During 1998, 104,422 offenders were arrested by Federal law enforcement agencies for violations of Federal law. A third of all arrests were for public or-der offenses; 29% for drug offenses; 16% for property offenses; 15% for supervision violations; 5% for violent offenses; and 3% to secure and safeguard a material witness. Approximately 41% of drug arrests involved cocaine; 28%, marijuana; 10%, amphetamine (or methamphetamine); 4%, heroin; and the remainder other drugs or drug paraphernalia (figure 1.1). Almost two-thirds of arrests for public-order offenses were immigration offenses. Most (80%) of these were for illegal entry into the United States; 13% involved alien smuggling; 6% involved false claims of citizenship by entrants to the United States; and about 1% involved other immigration violations (figure 1.2). Arrests by law enforcement agencies (table 1.2) During 1998, 72% of all arrests for Federal offenses were made by components of the Department of Justice; the Treasury components accounted for 11% of all arrests. In addition, 3% of arrests were made by State and local agencies and 8% of offenders were arrested after voluntarily reporting to the U.S. marshals following a summons. Of those arrests made by the Department of Justice, the U.S. Marshals Service made 37%; the Immigration and Naturalization Service, 33%; the FBI, 15%; the Drug Enforcement Administration, 14%; and other Department of Justice agencies, less than 1%. Of those arrests made by the Treasury Department, the U.S. Customs Service made 57%; the Bureau of Alcohol, Tobacco, and Firearms, 18%; the Secret Service, 18%; and the IRS, 7%. Arrests across demographic groups (table 1.3) Most (85%) of offenders arrested by Federal law enforcement agencies during 1998 were male. Seventy percent were white and 25% were black. Individuals between 21 and 40 years old comprised nearly 70% of all those arrested. One-third of those arrested were identified as noncitizens. Most (68%) offenders arrested for drug offenses were white, while 29% were black. Of those offenders arrested on charges involving cocaine, 59% were white and 38% were black. Of those offenders arrested on charges involving marijuana, 83% were white and 14% were black (figure 1.1). Chapter notes Tables 1.1-1.3 were derived from the U.S. Marshals Prisoner Tracking System database. Only records of arrests made during October 1, 1997, through September 30, 1998, were selected. Information on individual cases, offenses, and detainees were used to construct the tables. Offenses of arrest were classified by the U.S. Marshals into their four-digit offense codes. These were then aggregated into the offense categories shown in the tables. These categories are similar, but may not be directly comparable, to the categories used in tables in other chapters of this Compendium. Chapter 2 Prosecution Federal criminal cases may be brought by the U.S. Attorney's Office, by the Criminal Division of the U.S. Department of Justice, or by other authorized agencies such as the U.S. Marshals Service. The 93 U.S. attorneys serve as the chief Federal law enforcement officers within their respective districts. Investigations are most commonly referred to a U.S. attorney by a Federal investigative agency (primarily the Drug Enforcement Administration; the Federal Bureau of Investigation; the Postal Inspection Service; the Bureau of Alcohol, Tobacco, and Firearms; and the Secret Service), by the Criminal Division, or by a State or local investigative agency. Investigations may also be initiated -- and cases brought directly -- by U.S. attorneys or by the Criminal Division of the U.S. Department of Justice. This chapter reports only on suspects that were at least partially investigated by U.S. attorneys. U.S. attorneys' decisions After criminal investigations are initiated and criminal suspects are referred to them, U.S. attorneys may file charges against defendants in a U.S. district court, or they may decline to file these charges for reasons such as weak or insufficient evidence, minimal Federal interest, lack of resources, or lack of criminal intent. Matters that are declined may be referred to another authority for prosecution or be settled through alternative resolution procedures. U.S. attorneys may also file cases before U.S. magistrates, who have the authority to adjudicate misdemeanor offenses (18 U.S.C. § 3401). Because of the relatively less serious nature of these cases, and because they are handled by magistrates, the U.S. attorneys count these cases as criminal matters disposed by U.S. magistrates. The decision to prosecute a suspect in a criminal matter depends upon many factors, including the Attorney General's priorities, U.S. attorney priorities and resources, laws governing each type of offense, and the strength of evidence in each case. During 1998, 61% of suspects in all offense categories were prosecuted before U.S. district court, while an additional 12% were handled by U.S. magistrates. Suspects were prosecuted at different rates depending on their offense. Suspects involved in violent and drug offenses were more likely to be prosecuted before U.S. district court judges than were those involved in property and public-order offenses (figure 2.1). Suspects involved in drug offenses were the most likely (77%) to be prosecuted. Suspects involved in property offenses were the least likely to be prosecuted (52%). However, suspects involved in property offenses were more likely to have their matters declined and not referred to other authorities for prosecution (31%) than were suspects involved in other types of criminal matters. These cases were resolved by other means, including restitution, civil/administrative procedures, and pretrial diversion. Suspects in matters received (table 2.1) During 1998 there were 115,692 suspects in new matters received for investigation by U.S. attorneys.**(1 See Chapter notes, Item 1, p. 31. The 1998 data are not directly comparable to the 1993 and prior compendia)**. Of these suspects, about 35% were investigated for public-order offenses, 32% for drug offenses, and 26% for property offenses. Under 7% of all suspects were investigated for violent crimes. Relative percentages of offenses reflect criminal activity as well as Federal investigative, prosecutorial, and statutory priorities. For example, during 1998 priority areas for U.S. attorneys included the detention, educational and rehabilitation of juveniles, drug and violent crimes (including firearms prosecutions), international terrorism, and prosecution of those responsible for hate crimes and attacks on houses of worship (with the National Church Arson Task Force).**(2 Statistical Report, United States Attorneys' Offices, Fiscal Year 1998. U.S. Department of Justice, Executive Office for United States Attorneys, Washington, DC.)** Suspects in matters concluded (tables 2.2-2.6) Upon receiving a matter, a U.S. attorney will either immediately decline it for prosecution or investigate further. Upon conclusion of the investigation, a matter may be filed as a criminal case in a U.S. district court, referred to a U.S. magistrate, or declined for prosecution. Of the 106,022 suspects in criminal matters concluded during 1998, 61% were prosecuted in U.S. district court, 27% were declined for prosecution, and 12% were referred to U.S. magistrates. Suspects prosecuted (table 2.2) -- The likelihood of prosecution in U.S. district court varied widely across offense categories. Drug and violent offenders were more likely to be prosecuted than those involved in property and public-order offenses. U.S. attorneys prosecuted approximately 77% of suspects involved in drug offenses, 63% of those in violent offenses, 54% of those in public-order offenses, and 52% of those in property offenses. Differences in rates of prosecution within major categories were substantial. For example, among violent offenders, 79% of robbery suspects were prosecuted, while 57% of murder, 50% of assault, and 41% of rape suspects were prosecuted. Among property offenders, suspects investigated for counterfeiting were prosecuted at a higher rate than those involved in fraud (70% compared with 51%), and among suspects involved in public-order offenses, immigration (79%), weapons (70%), and tax law violators (65%) were much more likely to be prosecuted than suspects in escape (21%) and civil rights (6%) offenses. Suspects in matters concluded and declined, by investigating agency (table 2.3) -- During 1998, 59% of suspects in matters concluded were investigated by components of the Department of Justice; the Treasury Department investigated 19% of suspects in matters concluded. In addition, 10% of suspects in matters concluded were investigated by independent Federal agencies and State/local agencies. Among the investigating agencies, the Immigration and Naturalization Service (3%) and the National Park Service (10%) had among the lowest declination rates, while the Indian Affairs Bureau (58%), agencies within the Department of Health and Human Services (53%), and the Environmental Protection Agency (53%) were among agencies with the highest declination rates. The declination rate for matters investigated by components of the Department of Justice (28%) was quite close to the overall declination rate (27%). Suspects in matters declined (tables 2.4-2.5) -- Suspects in matters involving violent offenses were less likely to be declined for prosecution (33%) than those involved in property offenses (40%) but more likely than those in drug offenses (18%). Suspects in public-order offenses were slightly less likely to be declined for prosecution than those involved in violent offenses-- 25% compared to 33% (table 2.2). The decision to decline prosecution is based on a number of factors, including the lack of prosecutable offense, alternative resolution, or case- and suspect-related reasons (table 2.4). Of the 28,786 declinations during 1998, 27% occurred because of case-related reasons, mostly due to weak evidence (22%); 24% occurred because there was no crime or criminal intent was lacking; and 18% occurred for other reasons, such as agency requests (7%) and minimal Federal interest (4%). Not all suspects whose matters were declined for prosecution avoided prosecution. Twenty-four percent of the suspects in matters declined were referred to another authority for prosecution (table 2.5). An additional 5% were settled through alternative resolution procedures. While suspects involved in property offenses constituted the largest number of suspects who were referred by U.S. attorneys to other authorities for prosecution or resolution, suspects involved in drug offenses were the most likely of suspects whose matters were declined to be referred to other authorities for prosecution or alternative resolution (figure 2.2). Thirty-six percent of the drug suspects whose matters were declined prosecution by U.S. attorneys were referred for other prosecution or resolution, as were 27% of the public-order, 27% of violent, and 25% of the property offense suspects. Within major offense categories, the likelihood that suspects in matters declined for prosecution were referred to other authorities for prosecution or alternative resolution varied among specific offenses. Within property offenses, referral occurred with 42% of persons involved in motor vehicle theft, compared to 18% of suspects in arson and explosives cases. Within public-order offenses, 39% of weapons suspects and 60% of escape suspects were referred to other authorities or had their case settled through alternative procedures, compared to 6% of suspects in civil rights cases. Defendants in cases concluded by U.S. magistrates (table 2.2) -- Cases concluded by U.S. magistrates are, by statute, misdemeanors. (**3 At the time that U.S. attorneys receive them, matters may not be classifiable as felonies or misdemeanors. If a U.S. magistrate disposes for a case -- or what the U.S. attorneys call a "matter" -- its offense level has been determined to be a misdemeanor.)** Overall, U.S. magistrates disposed of 12% of all criminal matters investigated by U.S. attorneys. For most offenses, the likelihood that a U.S. magistrate concluded the matter was also relatively low. The most notable exception was escape, for which 55% of matters were concluded by U.S. magistrates. The escape offense category accounted for 13% of all cases concluded by U.S. magistrates. Most of the 2,805 escape violations were against defendants who absconded to avoid prosecution in State courts, the so-called "unlawful flight to avoid prosecution." Upon apprehension of the fugitive, these cases are normally dismissed by U.S. magistrates and turned over to the State for prosecution on the original warrant. Other types of offenses having comparably high rates of disposal by U.S. magistrates included migratory birds offenses (85%); conspiracy, aiding and abetting, traffic, and jurisdictional offenses (69%); and other property violations (61%). Processing times (table 2.6) -- For matters concluded during 1998, the average number of months from receipt of a matter to its conclusion by a U.S. attorney as a case filing or declination or its disposal by a U.S. magistrate was 8.2 months. However, the time to process matters varied with the outcome of the matter. Matters concluded by U.S. magistrates in which the suspect was convicted took the least amount of time -- an average of 3.0 months, and at least 50% of these convictions occurred in about 40 days. Similarly, 50% of the matters declined for prosecution were declined within 12.5 months. Overall, suspects in violent and drug offenses were processed more quickly than suspects in other offenses. For key decisions, such as whether to file a case or decline a matter for prosecution, drug offenses were concluded slightly quicker than violent offenses (5.3 months compared to 5.8 months). At an average of 13.6 months, fraudulent property offenses took longer to conclude than other offenses. Chapter notes 1) Tables 2.1-2.6 were created from the Central System data files of the EOUSA. For table 2.1, only records which showed a matter received during 1998 were selected. For tables 2.2-2.6, only records which show a matter terminated by reason of declination, disposition by a U.S. magistrate, or filed as a case in U.S. district court from October 1, 1997, through September 30, 1998, were selected. In all of these tables, matters "declined immediately" -- those cases in which the U.S. attorney spent less than 1 hour investigating -- were excluded. 2) The number of suspects in matters investigated in table 2.1, and the number of suspects in matters concluded in tables 2.2 and 2.6 are not directly comparable to counts in the 1993 and prior compendia. In those years, these statistics included appellants in appeals filed or appeals terminated. In table 2.1 of this Compendium and its corresponding tables in 1994 and subsequent compendia, the number of suspects in matters received is limited to suspects in criminal matters that were filed as cases in U.S. district courts, handled by U.S. magistrates, or declined for prosecution. In the 1993 and prior compendia, the number of suspects in criminal matters received included criminal appeals. To obtain a number of suspects comparable to the statistic reported in the 1993 and prior compendia, add to the data in table 2.1 of this Compendium, the number of appellants in appeals that were handled by U.S. attorneys. In tables 2.2 and 2.6 of this Compendium and their corresponding tables in the 1994 and subsequent compendia, the number of suspects in matters concluded also is limited to those suspects whose matters were concluded by a case filing in a U.S. district court, by declination, or by disposition by a U.S. magistrate. In the 1993 and prior compendia, this number included appellants in criminal appeals terminated. To obtain a number that is comparable to the number reported in the 1993 and prior compendia, the number of appellants in appeals terminated and handled by U.S. attorneys needs to be added to the totals in table 2.2. 3) Offenses in the Central System data files are classified by the title and section of the United States Code for the most serious offense investigated, as determined by the assistant U.S. attorney responsible for the matter. For tables in this Compendium, these citations were translated into the corresponding four-digit offense codes used by the Administrative Office of the United States Courts (AOUSC). These four-digit codes were then aggregated into the categories shown in the tables. U.S. Code citations often do not permit detailed classification of drug offenses by the type of criminal activity involved. A person charged with conspiracy is classified under the substantive offense alleged (for example, conspiracy to defraud is classified as fraud) unless type of conspiracy cannot be determined from the U.S. Code citation. Unclassified conspiracies are included among "conspiracy, aiding and abetting, traffic, and jurisdictional offenses" in tables 2.1, 2.2, 2.5, and 2.6. The most serious offense investigated is based on the criminal lead charge as determined by the assistant U.S. attorney responsible for the criminal matter. 4) In the tables in this chapter, the "all offenses" row or "total declinations" row (in table 2.5) includes records whose offense category (basis for declination) is missing or indeterminable. The percentage distributions in these tables, however, are based on the number of observations with nonmissing offenses (basis for declination) values. The number of missing values are reported either on a separate line or in a footnote. Chapter 3 Pretrial release The Bail Reform Act of 1984 (18 U.S.C. § 3141) sets the terms for release and detention of defendants facing charges in Federal courts. At the time defendants first appear before a judicial officer, they may be (1) released on personal recognizance or unsecured bond; (2) released subject to conditions imposed by the court, possibly including traditional bail; (3) temporarily detained without bail (limited to 10 working days) to permit deportation, exclusion, or the revocation of previously granted conditional release; or (4) detained pending the outcome of a detention hearing. **(*18 U.S.C. sec. 3142(e) (1984)** According to the statute, a defendant must be brought before a judicial officer without "unnecessary delay" upon arrest. The judicial officer, who may be a judge but is generally a U.S. magistrate, determines whether the defendant will be released or detained prior to trial. If a hearing is required, the release decision may be delayed no longer than 3 days (or 5 days if requested by the defendant). At the initial appearance a recommendation to detain the defendant is made by the U.S. attorney and/or the pretrial services officer. If the court concurs with the recommendation to detain the defendant, a pretrial detention hearing is scheduled -- typically within 1 week of the initial appearance. At the pretrial detention hearing evidence is presented to show cause why the defendant should be detained pending adjudication of the charges. Recommendations to detain a defendant are limited to those instances where the defendant was charged with (1) a crime of violence, (2) an offense with a statutory maximum sentence of life imprisonment or death, (3) a drug offense with a statutory maximum sentence of 10 years or more imprisonment, or (4) any felony offense if the defendant had been convicted on two or more occasions of an offense described above or a similar State-level offense. Additionally, a detention recommendation may be made if there is a serious risk that the defendant would flee in an attempt to avoid prosecution or would obstruct justice or intimidate witnesses or jurors. Types of pretrial release Defendants may be released without financial conditions according to the following: Personal recognizance -- defendant is released subject to no financial or other conditions. Unsecured bond -- no money is required to be posted before release, but defendant is liable for full bail amount if he or she fails to appear. Conditional release -- any combination of restrictions that are deemed necessary to guarantee the defendant's appearance at trial or the safety of the community. Non-financial conditions commonly place restrictions on the defendant's movements, associations, and/or actions. They may also involve employment or treatment for medical, psychological, or substance abuse conditions. Defendants may also be released on financial conditions. Financial conditions include (1) deposit bond (the defendant is required to post a percentage of the total bail amount, usually 10%), (2) surety bond (the defendant is released subject to guarantees by a third person that the full amount will be paid), or (3) collateral bond (collateral equal to the full bail amount required to be posted by the defendant before release). Financial conditions may occur in combination with nonfinancial conditions. Factors relating to release or detention In deciding whether to release a defendant and in setting release conditions, the court is directed to consider the nature and circumstances of the offense charges, the weight of evidence against the defendant, the defendant's character, physical and mental condition, family ties, employment, financial resources, length of residence in the community, community ties, past conduct, history relating to drug or alcohol abuse, criminal history, and record concerning appearance at court proceedings. During 1998, 26,260 defendants were released for some period of time prior to trial (table 3.1). Those defendants released made up 43% of the 60,950 defendants who terminated pretrial services during 1998. Of the defendants released, 84% completed pretrial services without violating the conditions of their release (table 3.8). In 1998, 40,669 defendants were detained for some period of time prior to the disposition of their case (table 3.3). Those defendants detained comprised about 67% of all defendants who terminated pretrial services during 1998. Included are those who were detained prior to having a preliminary hearing with the magistrate, at which point they could have been either released or ordered to have a detention hearing. Of the defendants detained, 55% were ordered detained by the court until case disposition. Pretrial detention hearings were held for 30,204 defendants. Of these, 22,143, or 73%, were ordered detained (table 3.5). Of the 26,246 defendants released prior to trial, 16% violated a condition of their release (table 3.7). The majority (88%) of these violations were technical violations of the bail conditions. Defendants released on deposit bond were more likely (25%) than other defendants to incur some violation of the conditions of their release (table 3.8). Conversely, defendants given conditional release violated their release at the lowest rates (5%). Five percent of all released defendants had their release revoked. Pretrial outcomes by offense categories Releases (table 3.1) -- Defendants charged with property offenses were more likely than other defendants to be released prior to criminal trial; in 1998, 77% of the 13,463 property offenders were released (figure 3.1). However, the percent released varied greatly across offense categories. Among violent offenders, 18% of those charged with robbery were released compared to 55% of those charged with assault and 66% of those charged with sex offenses other than rape. Among public order offenders, 8% of those charged with immigration offenses were released compared to 95% of those charged with tax law violations and 43% of those with weapons offenses. Property defendants also made up the largest part of those defendants who were released, while drug defendants made up the largest part of those defendants who were detained. Violent defendants composed the smallest part of defendants in both groups. About 35% of all drug defendants were released. Drug trafficking defendants were less likely to be released than those charged with nontrafficking offenses (33% compared to 53%), but defendants charged with trafficking offenses outnumbered those charged with nontrafficking offenses by more than 9 to 1. The release types for those defendants who were released prior to trial varied among offense categories (table 3.1). Within each major offense category, the highest percentage of defendants released was on unsecured bond, except for violent defendants. Of the 32% of violent defendants released, 40% were released on unsecured bond and 43% on personal recognizance compared to 77% of all property defendants released, of whom 54% were released on unsecured bond and 34% on personal recognizance. Detentions (tables 3.3 and 3.5) -- Among the major offense categories, defendants charged with drug offenses were the most likely (80%) to be detained for some period of time (table 3.3, figure 3.1). Seventy-seven percent of defendants charged with violent offenses were detained for some period of time. But violent defendants were more likely than drug offenders (62% versus 55%) to be detained by the court. Property and public-order defendants were less likely to be detained than were drug and violent defendants. During 1998, 74% of public-order defendants were detained. Of public-order defendants, those charged with immigration offenses were the most likely to be detained (96%). In 1998, 30% of defendants charged with property offenses were detained. Of those defendants detained, 56% of public-order and 43% of property defendants were detained by the court. For 66% of defendants charged with violent offenses, it was decided at the preliminary hearing that a detention hearing was warranted (table 3.5, figure 3.2). This rate varied by offense category: 62% of those charged with drug-related offenses, 50% of those charged with public-order offenses, and 22% of those charged with property offenses were ordered to have a detention hearing. Of those charged with violent offenses, the most likely to have a pretrial detention hearing were those charged with kidnaping, robbery, threats against the President, or murder. The least likely to have a pretrial detention hearing were those charged with sex offenses other than rape and those charged with assault. Once a defendant was ordered to have a detention hearing, the chances of being detained were less variable across offense categories (table 3.5). The highest rate of defendants ordered detained were those charged with public-order offenses (83%) -- primarily immigration (89%) and weapons offenders (70%). Of the 66% of defendants charged with violent offenses and given a pretrial detention hearing, 73% were ordered detained. In 1998, 70% of those charged with drug offenses and 59% of those charged with property offenses were ordered detained following a pretrial hearing. Violations (table 3.7) -- Of defendants released prior to trial and terminating pretrial services during 1998, those charged with drug offenses were more likely (27%) than other defendants to incur at least one violation during the release period. Among major offense categories, those charged with property offenses were less likely (11%) than others to violate conditions of their release. Those charged with drug offenses and violent offenses were more likely to have their release revoked (10% of all released drug defendants, 9% of all released violent defendants), and public-order defendants were slightly more likely than property defendants to have their release revoked (4% versus 3%). Pretrial outcomes across demographic groups Releases (table 3.2) -- Females were more likely (68%) than males (38%) to be released during 1998 (figure 3.3). Males were more likely to have a financial condition imposed on them as females (17% versus 11%). Fifty-eight percent of all defendants identified as non-Hispanic were released during 1998 compared to 19% of Hispanics. Non-Hispanic releasees had more than twice the chance of being released on personal recognizance as Hispanics. The higher the education level of the defendant, the greater the probability of release. A third (34%) of defendants with less than a high school education were released in 1998, while 75% of those who had completed college were released. Releasees with a college degree were less likely to receive financial conditions (13%) and more likely to be released on personal recognizance (31%) than those who had less education. Defendants with a history of drug abuse were released less frequently than those defendants with no known drug history. However, among all released defendants, drug abuse history does not appear to have had a major effect on the form of pretrial release. Detentions (tables 3.4 and 3.6) -- Males were more likely (71% versus 44%) than females to be detained (table 3.4). Of all detainees, 56% of the men and 42% of the women were detained following a detention hearing. Hispanics had a substantially higher probability of being detained (nearly 9 out of 10) as compared with non-Hispanics (5 out of 10). Higher detention rates for Hispanics reflects a perceived greater risk of pretrial flight. In a previous BJS report, 70% of Hispanics prosecuted in the Federal courts were identified as noncitizens and 64% lived in the community in which they were arrested for less than 1 year. (Federal Pretrial Release and Detention, BJS Special Report, NCJ 168635.) Additionally, almost all of those charged with an immigration offense -- of which Hispanics comprise the majority -- were detained because of the perceived risk of pretrial flight. Younger defendants were more likely than older ones to be detained, with the percentage peaking at defendants in their twenties. Seventy-two percent of defendants between the ages of 21 and 30 years were detained compared to 70% of defendants between 16 and 18 years and 48% of those over 40. However, among all defendants detained, the chances of being detained following a detention hearing were similar across all age groups. These patterns were more dramatic with education levels. Those defendants with lower levels of education were more likely to be detained. Seventy-seven percent of those who did not graduate from high school were detained versus 34% of college graduates. Of those detained, 54% of defendants without a high school diploma were detained following a detention hearing compared to 48% of college graduates. Not only were male defendants nearly twice as likely as female defendants to have a detention hearing (53% versus 31%), they were also more likely to be ordered detained as a result of the hearing (table 3.6). The likelihood of having a detention hearing increased with the number of prior convictions, severity of the defendant's criminal history, and current sentence of supervision. Further-more, 55% of those with a history of drug abuse had detention hearings as compared to 42% of those without such a past. Those with a history of drug abuse were slightly more likely to be detained. Violations (table 3.9) -- Released males were more likely than females to incur a violation during the pretrial release period (18% versus 13%) and were slightly more likely to have their release revoked. Blacks were more likely to incur a violation than whites (21% versus 14%). Defendants with no prior criminal history were less likely to have a pretrial release violation than those with a prior conviction. Released defendants with a known drug history had more than 3 times the likelihood of incurring a violation than those with no known drug history. Also, those with a history of drug use were substantially more likely to have their release revoked. Length of pretrial detention (table 3.10) Among defendants detained for any time prior to case disposition, those charged with drug offenses were detained the longest (an average of 88.7 days) compared to violent defendants (81.2 days), public-order defendants (62.5 days), and property defendants (50.0 days) -- not in a table. Of defendants released, those released on conditional means spent the greatest average number of days detained (65.2 days) compared to those released on personal recognizance (50.1 days), unsecured bond (46.1 days), or financial conditions (35.8 days). This trend varied only slightly across all major offense categories -- defendants released on conditional means were detained more days than defendants who were released by any other means. Of defendants not released, those detained following a detention hearing were detained longer, on average, than those held on financial conditions (105.0 days compared with 77.4 days). Only public order defendants held on financial conditions (74.2 days) were detained longer than those detained by the court (72.0 days). Chapter notes 1) All tables in chapter 3 were created from data in the Pretrial Services Information Act System data base, which is maintained by the Pretrial Services Agency (PSA) within the AOUSC. The data describe 60,950 defendants who terminated pretrial services during October 1, 1997, through September 30, 1998, and whose cases were filed by complaint, indictment, or information. In these tables, the totals (e.g., "all offenses") included records whose offense or other attributes were missing or indeterminable. The percentage distributions were based on nonmissing values, and missing values were reported in a separate row or in a footnote. Offenses in the PSA are based on the most serious charged offense, as determined by the probation officer responsible for the interview with the defendant. The probation officer classifies the major offense charged into AOUSC four-digit offense codes. For defendants charged with more than one offense on an indictment, the probation officer chooses as the most major charged offense the one carrying the most severe penalty or, in the case of two or more charges carrying the same penalty, the one with the highest offense severity. The offense severity level is determined by the AOUSC, which ranks offenses according to the maximum sentence, type of crime, and maximum fine amount. These four-digit codes are then aggregated into the same offense categories as those used in chapter 2. For drug offenses, the type of drug activity -- trafficking or possession -- is obtained by the probation officers from their reading of the indictment or other charging documents. The AOUSC citation manual provides probation officers with detailed instructions on how to code drug crimes. 2) In tables 3.13.4, the percentages showing the methods of release or methods of detention were based on the number of defendants released or the number of defendants detained. In tables 3.5 and 3.6, the percentages were based on the number of defendants who had hearings and were ordered detained. This method departs from the 1993 and prior compendia, in which the percentages were based on the number of defendants terminating pretrial services. Chapter 4 Adjudication Federal criminal cases may result in conviction of the defendant following plea or trial, acquittal of the defendants after trial, or dismissal of the case.**(1Unless otherwise noted, data describe felony and misdemeanor cases in U.S. district courts and include Class A misdemeanors handled by U.S. magistrates. The data also include Class B misdemeanors that are handled by U.S. district court judges, which occur infrequently.)** Defendants in cases commenced (table 4.1) During 1998, 78,172 defendants had criminal cases commenced**(2Cases are commenced on or after the date that charges are filed against defendants.)** against them in U.S. district courts, and 64,769 (83%) of those defendants were charged with felonies. The 28,021 felony drug defendants comprised 36% of all criminal defendants in cases filed and 43% of all felony defendants. Felony property and public-order defendants comprised 19% and 23%, respectively, of all defendants in cases filed. The 13,254 misdemeanor defendants constituted 17% of all defendants in cases filed, with 37% of those misdemeanor defendants charged with traffic violations. Defendants in cases terminated (table 4.2) Cases were terminated against 69,769 defendants during 1998, 57,054 (82%) of whom were felony defendants. During 1998, the 24,317 drug-related felony defendants comprised nearly 35% of all defendants in cases terminated and 43% of all felony defendants. Over 87% of all defendants were convicted; 90% of all felony defendants and 75% of misdemeanor defendants. For major felony offense categories, conviction rates hovered around the 90% range. Within major offense categories, however, the conviction rates varied more widely. For example, within the violent offense category, the conviction rates ranged from 80% for those found guilty of assault, to 96% for robbery. Within the public-order category, the conviction rate ranged from 66% for civil rights violations to 95% for immigration and gambling offenses. Most defendants who were convicted pleaded guilty. Overall, over 93% of those convicted pleaded guilty, while only 6% were convicted at trial. For felony offenses, 94% pleaded guilty to their charges. For the major felony offense categories, guilty pleas were registered for 92% of violent offend-ers, 93% of drug offenders, and 95% each for public-order and property offenders. During 1998, 4,882 defendants exercised their right to a trial. A higher percentage of violent offenders went to trial than drug, property, and public-order offenders. Ten percent of violent offenders went to trial as compared with 6% of both property and public-order offenders, and 7% of drug offenders. Of defendants who exercised their right to a trial, 3,692 (76%) were convicted either by a jury or a bench trial. The felony trial conviction rate was 83%, while the misdemeanor conviction rate was 50%. Among felony offenses, trial conviction rates ranged from 79% for violent defendants to 86% for drug defendants. Property and public-order offenders fell in between, as 80% were convicted at trial for each respectively. Case processing times (table 4.3) The data in table 4.3 cover the inter-val from the time a case is filed in U.S. district court through sentencing for those convicted as well as the interval from case filing through disposition for those not convicted, or those whose cases are dismissed. Title I of the Speedy Trial Act of 1974, as amended, sets time requirements for processing criminal cases in Federal courts. Except for certain exclusions, indictment must occur within 30 days of arrest; defendants are guaranteed at least 30 days to prepare for trial; and the Government must be ready for trial within 70 days. The Speedy Trial Act does not specify a time interval from trial to sentencing,**(3The sentencing reform Act of 1984 specifies minimum intervals for the disclosure and objections to the presentnce report. See rule 31, Federal Rules of Criminal Procedure)** and the time limits set by the Act exclude several enumerated periods of pretrial procedure such as time spent awaiting rulings on motions.**(418 U.S.C. 3161 et. seq.)** Additionally, defendants may waive their rights within the Speedy Trial Act. As a result, the actual processing time for most defendants can be longer than the statutory limits without violating the provisions of the Act. Overall, the average time for processing defendants was 8.7 months (figure 4.1). For felony defendants, the overall processing time average was 9.7 months; for those charged with misdemeanor crimes, the average processing time was 3.9 months. Defendants who pleaded guilty were processed faster (7.9 months) on average, than defendants who went to trial (11.6 months). Overall, defendants whose cases were ultimately dismissed took longest to process (12.7 months), on average. Among major felony offense categories, however, case processing times were not similar to the overall pattern. Only those convicted of drug offenses took a greater amount of time for dismiss-al, on average, than those who went to trial. Convictions by U.S. Magistrates (table 4.4) During 1998 U.S. magistrates disposed of 11,685 misdemeanor criminal defendants, almost 75% of whom were convicted. Traffic violators comprised 42% of the defendants disposed by U.S. magistrates, with property offenders comprising another 32%. Drug offenses comprised almost 10% of U.S. magistrates' cases. Characteristics of convicted defendants (table 4.5) Among defendants convicted, there were over 5 times as many men as women (85% versus 15%); over twice as many whites as blacks (67% versus 28%); and almost twice as many non-Hispanics as Hispanics (63% versus 37%). Most were U.S. citizens (67%), had graduated from high school or completed some higher education (57%), and had some criminal history (54%). Persons over the age of 30 represented 55% of defendants. Chapter notes 1) Tables 4.1-4.4 were derived from the Administrative Office of U.S. Courts (AOUSC) criminal master data files. Only records with cases filed in U.S. district court (table 4.1) or cases that terminated in U.S. district court during October 1, 1997, through September 30, 1998, were selected. Offenses were classified according to the most serious offense charged. In the case of multiple offenses, the offense carrying the most severe potential penalty was selected. In this Compendium, carjacking offenses are classified as robberies, based on title and section of the U.S. Code. In compendia prior to the 1997 Compendium, they were classified as motor vehicle thefts, based on the AOUSC offense classifications. 2) Table 4.5 was created by matching the AOUSC master data files with the U.S. Sentencing Commission (USSC) monitoring system files and the Pretrial Services Agency (PSA) data files. These latter two data files contain information on the characteristics of defendants. The USSC monitoring system files are limited to records of defendants sentenced under the Federal sentencing guidelines only. These include defendants convicted of felonies or Class A misdemeanors. Excluded from the USSC data were defendants convicted only of Class B or C misdemeanors or infractions, defendants whose offenses were committed before November 1, 1987, and juvenile offenders. Juveniles are not subject to guidelines, and they were not included in table 4.5. Some of the defendants excluded from the USSC data files were included in the PSA data. (See Chapter 3 Chapter notes for more information on the PSA data.) Table 4.5 indicates the number of records for which relevant data were available. Percentage distributions were based on records with known values of defendant characteristics and offenses. 3) A criminal record, as reported in table 4.5, is limited to prior adult convictions. For some defendants in this table, it is further limited to the portion that is relevant for calculating sentences under the Federal sentencing guidelines. In general, this is limited to sentences imposed within a 15-year period prior to the current offense and offenses committed within the United States. For most defendants, the criminal history used to calculate sentencing guideline ranges includes their entire adult criminal history. Chapter 5 Sentencing Depending upon the type and severity of the offense, convicted offenders may be sentenced to incarceration, probation, a fine, or a combination of sanctions such as a split or mixed sentence (imprisonment as well as a period of probation supervision). (For a definition of mixed sentences, see Glossary, p. 111.) The Federal sentencing guidelines require a term of supervised release following service of any prison sentence of more than 1 year. In addition, courts have the discretion to impose supervised release in any other case. Except where otherwise indicated, tables in this chapter are based on the most serious offense of conviction. They are not directly comparable with tables in earlier chapters that are based on the most serious offense investigated or most serious offense charged (see "Offense classifications" in Methodology, p. 107). Offenders convicted and sentences imposed (table 5.1) Of the 60,958 offenders sentenced during 1998, 43,041 (71%) were sentenced to prison; 12,744 (21%) were sentenced to probation; and 3,199 (5%) were ordered to pay only a fine. (Offenders given an intermediate sanction such as intermittent confinement or community confinement that also included probation supervision are counted among offenders given probation.) Of the 43,041 offenders sentenced to prison, 41,420 (96%) were convicted of felonies, and 82% of convicted felons were sentenced to prison. Defendants convicted of drug (92%) and violent (91%) felonies were more likely to receive prison sentences than were defendants convicted of property (60%) or public-order (83%) offenses (figure 5.1). Among public-order offenders, persons convicted of weapons (92%), immigration offenses (91%), and escape (91%) were as likely as violent and drug offenders to receive prison sentences. Among property offenders, persons convicted of arson or the use of explosives (87%) were also almost as likely as violent and drug offenders to receive prison sentences. Overall, 21% of convicted offenders were sentenced to probation. The percentage of misdemeanants sentenced to probation (48%) was three times that of convicted felons (15%). Among felons, persons convicted of communication violations (86%), postal law violations (84%), and agriculture violations (68%) were the most likely to be given probation. While less than 1% of all convicted felons were ordered to only pay a fine, 45% of the felons convicted of antitrust violations were so ordered. Fines generally were reserved for misdemeanor offenders, 28% of whom received fines. Average prison sentences imposed (table 5.2) For all offenders given prison terms, the average term of imprisonment imposed was 58.8 months. Persons convicted of felonies received an average prison term of 60.6 months. The average length of prison sentence imposed varied among major offense categories (figure 5.2). Property offenders received the shortest sentence, on average, while violent offenders received the longest (25.4 months compared to 84.2 months). Drug offenders received 78.8 months, on average, and public-order offenders received 47.3 months. The average prison term for public-order offenders was primarily attributable to sentences imposed for weapons offenders. The 2,914 weapons offenders sentenced to prison comprised 24% of the 12,218 public-order offenders sent to prison; weapons offenders received an average prison term of 100.6 months, and public-order offenders other than weapons offenders received an average prison term of 30.6 months. Relationship between sentence imposed and mode of conviction (table 5.3) Offenders convicted by plea were less likely to receive prison sentences than those convicted at trial. Seventy percent of the offenders convicted by guilty plea received some prison time, while 84% of the offenders convicted at trial received prison sentences. For violent offenders, 91% of those convicted by plea received prison sentences, as did 97% of those convicted at trial. For fraudulent property offenders, 59% of those convicted by plea received prison sentences, compared to 86% of those convicted at trial. And, for regulatory public-order offenders, 41% of those convicted by plea went to prison, but 70% of those convicted at trial did. The average prison term imposed on defendants convicted at trial was longer than the term imposed on defendants convicted by plea. Defendants convicted at trial received 148.9 months, on average, while those convicted by plea received an average of 52.3 months (figure 5.3). Drug offenders convicted at trial received an average of 188.4 months as compared to the 71 months for drug offenders convicted by a guilty plea. Violent offenders who went to trial received an average sentence of 180.4 months as compared to the 75.8 months for those convicted by plea. Characteristics of offenders sentenced to prison (tables 5.4-5.6) The majority of persons sentenced to prison during 1998 were male(85%), white (67%), non-Hispanic (63%), and U.S. citizens (67%). The percentage sentenced to prison was not uniform across all categories of age, race, sex, education, and ethnicity. Similarly, the average sentences imposed were not uniform across these categories (table 5.5). However, conclusions about the effect that characteristics may have had on sentencing cannot be drawn from the aggregate patterns in the data. Such an assessment would require detailed information for each category showing the presence or absence of factors that are legally intended to affect sentencing, such as offense severity, weapon use, role in the offense, victim injury, dollar loss, and so forth. One such study of the factors affecting sentencing found that "nearly all of the aggregate differences among sentences for whites, blacks, and Hispanics during 1989-90 can be attributed to characteristics of offenses and offenders that current law and sentencing guidelines establish as legitimate considerations in sentencing decisions." **(* Douglas C. McDonald and Kenneth E. Carlson, Sentencing in the Federal Courts: Does Race Matter? Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, December 1993 (NCJ145328).)** Average sentences imposed were longest for males (61.8 months), blacks (90.2 months), non-Hispanics (69.1 months), and U.S. citizens (68.8 months). These longer terms were due in large part because these categories of offenders (males, blacks, non-Hispanics, and U.S. citizens) are more highly concentrated in the offense types (table 5.4) that also are associated with longer average sentences (table 5.2). For example, incarceration rates for blacks exceeded those of whites in the violent, drug, and nonregulatory public-order offense categories. These offenses are associated with longer average sentences (table 5.2, figure 5.2) than are property offenses or misdemeanors -- the categories in which whites' incarceration rates exceeded those of blacks. In general, within categories of offender characteristics, average prison terms were longer for violent and drug offenses than for property and regulatory public-order offenses. These two offense categories generally are considered to be more serious under the Federal sentencing guidelines than are property and regulatory public-order offenses. As a result, for aggregate tables like the ones presented in this chapter, persons with a certain characteristic may appear to receive more severe sentences on average. The primary reason for the larger averages is the larger number of serious offenses that made up the average sentences. Chapter notes 1) Tables 5.1-5.3 were derived from the AOUSC criminal master data files. Only records of defendants sentenced during October 1, 1997, through September 30, 1998, were selected. Offenses shown in these tables -- offenses of conviction -- are based on the longest actual sentence imposed. 2) Tables 5.4-5.6 were created from AOUSC criminal master data files, supplemented with linked data on offender characteristics from two sources: the United States Sentencing Commission (USSC) monitoring system files (which are limited to records of defendants sentenced under the Federal sentencing guidelines); and the Pretrial Services Information Act System database, maintained by the Pretrial Services Agency (PSA). These two supplemental data files contain information on the characteristics of defendants. Table 5.4 indicates the number of records for which relevant data were available. Percentage distributions were based on records with known values of defendant characteristics and offenses. Means and medians in tables 5.5 and 5.6 were based on the number of records shown in table 5.4. 3) Criminal record, as reported in tables 5.4-5.6, is limited to prior adult convictions. For some defendants in these tables, it is further limited to the portion of their criminal record that is relevant for calculating sentences under the Federal sentencing guidelines. In general, this is limited to sentences imposed within 15 years of the current offense and sentences for offenses committed within the United States. For most defendants, the criminal history used to calculate sentencing guideline ranges is the same as their adult criminal history. Chapter 6 Appeals Prior to implementation of the Sentencing Reform Act of 1984, only criminal convictions could be appealed. However, the Sentencing Reform Act provided for the appellate review of sentences imposed given that the sentence was (1) imposed in violation of the law; (2) imposed as the result of an incorrect sentencing guideline application; (3) outside the recommended guideline sentencing range; or (4) imposed for an offense for which no sentencing guideline exists and is plainly unreasonable. Both the defendant and the Government have the right to appeal an imposed sentence (18 U.S.C. § 3742). Appeals filed (tables 6.1 and 6.2) During 1998 the U.S. Court of Appeals received 10,535 criminal appeals. Of all appeals, 4% were filed by the Government (not shown in a table). Forty-seven percent of appeals filed in 1998 challenged both the conviction and the sentence imposed. Twice as many appeals challenged only the sentence imposed as those that challenged only the conviction (24% versus 12%). Fifteen percent of appeals filed were appeals of convictions for crimes committed before 1987, when the sentencing guidelines took effect (figure 6.1). The distribution of appeals by type of appeal filed (sentence, conviction, or both) was relatively similar across the major offense categories underlying the appeal. However, 56% of appeals filed for immigration offenses appealed both the sentence and the conviction, compared to 48% for public-order offenses and 47% for all criminal categories. Nearly half of all appeals filed were for drug convictions (48%). Appeals for public-order convictions comprised 25% of all appeals filed, property offenses comprised 19%, and the remaining 7% of appeals filed were for violent offenses (figure 6.2). Appeals terminated (tables 6.3-6.5) A total of 10,105 appeals terminated during 1998. Of these, 47% were appeals of both the sentence and conviction, 24% were appeals of the sentence only, and 14% were appeals of the conviction only. Fourteen percent were appeals of convictions for offenses sentenced under laws in effect prior to the sentencing guidelines. In nearly half of all appeals terminated in 1998, drug offense convictions underlie the appeal (46%); the under- lying offense in 24% of the appeals was a public-order offense; 19% were property offense appeals; and 7% were appeals for violent offense convictions. For 43% of public-order appeals, the underlying offense of conviction was for weapons. Robbery offenses composed 42% and murder composed 16% of the underlying offenses for appeals of a violent conviction. Sixty-two percent of property offense appeals were for fraud convictions. Of the 10,105 appeals terminated during 1998, 76% were terminated on the merits, while the remainder were terminated on procedural grounds (figure 6.3). For appeals based on a conviction of a drug offense or a public-order offense, 78% were terminated on the merits. Also terminated on the merits were 79% of appeals based on a violent conviction and 72% of appeals based on a property conviction. Within these major offense categories, the appeals having the highest rates of termination on the merits were those based on a conviction for a weapons offense (83%), a rape or a murder conviction (81%), or a conviction for racketeering and extortion (80%). In 82% of appeals terminated on the merits, the district court ruling was affirmed. In another 4% it was partially affirmed. Eighty-three percent of all drug offense appeals were affirmed, as were 80% of violent offense appeals and property appeals. Eighty-two percent of public-order offense appeals were affirmed. District court decisions were reversed or remanded back to the court in 8% of the cases. Violent offenses had the highest rate of reversal or being remanded back to the court (12%). Nine percent of property offense appeals were reversed or remanded, as were 8% of public-order and 7% of drug offense appeals. Overall, appeals were dismissed in 5% of cases. Across major offense categories, this dismissal rate remained relatively constant. Chapter notes Tables in chapter 6 were created from the AOUSC appeals database. Records of appeals filed or terminated during October 1, 1997, through September 30, 1998, were selected. Data include records of 10,535 criminal appeals filed during fiscal year 1998 and 10,105 appeals terminated during the same period. The unit of analysis in the appeals data is a single case. Most records in the appeals data report on a single appellant. Appeals were classified into the offense categories that represent the underlying offense of conviction. Offenses represent the statutory offense charged against a defendant in a criminal appeal. Offenses were classified by the AOUSC into their four digit offense codes. These were then aggregated into the offense categories shown in the tables. Chapter 7 Corrections Federal offenders under supervision (table 7.1) As of the end of the fiscal year 1998, there were 92,768 offenders under active Federal supervision, of which most (88%) were felons. About 65% of these offenders were receiving one of two forms of post-incarceration supervision: supervised release (54,822) or parole (5,827). The remainder (32,119) were under probation supervision. Most of the probationer population had been convicted of either a property felony (38%) or some type of misdemeanor offense (33%). Nearly 53% of offenders under supervised release and over 52% of parolees had been convicted of a drug offense. Outcomes of offenders completing supervision (tables 7.3, 7.5, and 7.7) Probation (table 7.3) -- During 1998, 16,011 offenders completed one or more terms of active probation. Overall, 81% of offenders successfully completed their term of probation, another 17% violated their conditions of probation, and the remaining 2% were administrative closures. Of offenders terminating probation, 5% committed a new crime (figure 7.1). Most committed technical violations, including drug use (4%) and absconding (2%). Offenders convicted of violent offenses were less likely than others to successfully complete a term of probation without a violation (figure 7.2). During 1998, 30% of probationers convicted of violent offenses violated their conditions of probation, as did 16% of drug, 14% of property, and 9% of public-order offenders. Violent offenders were also more likely to commit new crimes (9%) than were probationers convicted of property (4%), drug (5%), and public-order (4%) offenses. Supervised release (table 7.5) -- During 1998, 19,878 offenders completed terms of supervised release. Overall, 64% of them successfully completed their term of supervised release; 21% committed technical violations, such as drug use (8%) and absconding (4%); and 12% of these offenders violated their probation supervision by committing a new crime. The remaining 3% had their supervision administratively terminated. Violent offenders were less likely than others to complete a term of supervised release without a violation. Fifty-four percent of violent offenders violated conditions of supervised release, while 32% of drug, 32% of public-order, and 28% of property offenders violated supervised release. Violent offenders were also more likely to commit new crimes (18%) than offenders convicted of property (10%), drug (12%), or public-order (13%) offenses. Parole (table 7.7) -- During 1998, 3,077 offenders completed terms of parole. Overall, 56% of these offenders successfully completed their term of parole. During 1998, 13% of these offenders violated their supervision by committing a new crime; 26% committed technical violations, such as drug use (12%) and absconding (5%); and 5% completed their parole term through an administrative case closure. Offenders convicted of violent offenses were less likely than others to complete a term of parole without a violation. Sixty-two percent of violent offenders violated conditions of parole, while 35% of public-order, and 33% of property and 29% of drug offenders violated parole. Violent offenders were also more likely to commit new crimes (20%) than offenders convicted of property (14%), public order (12%), or drug (10%) offenses. Characteristics of offenders completing supervision (tables 7.4, 7.6, and 7.8) Probation (table 7.4) -- Among offenders who completed a term of probation, 74% were male; 68% were white and 88% were of non-Hispanic origin; 25% were less than age 31 (compared to 53% over age 40); 25% had less than a high school diploma (compared to 38% whose highest education was a high school diploma and 37% who had at least some college education); and 80% had no known drug history. Younger offenders, those with a history of drug abuse, and those with a lower educational background had higher probabilities of violating the conditions of probation. Offenders ages 16 to 20 were more likely (41%) to violate conditions of supervision than probationers in other age groups -- 25% of those ages 21 to 30 and 13% of those over age 30. Probationers with a history of drug abuse were almost 9 times more likely to violate probation for drug use than offenders who were not drug abusers. Probationers with known histories of drug abuse were also more likely to violate supervision for other reasons (fugitive status and new crimes) than were other probationers. Probationers with lower levels of education were more likely to violate the conditions of probation than those with higher levels of education. Those who did not graduate high school had a 25% violation rate, those who were high school graduates had a 17% violation rate, those with some college had a 13% violation rate, and those with a college degree had only a 5% violation rate. Supervised release (table 7.6) -- Among those who completed a term of supervised release, 85% were male; 66% were white; 83% were of non-Hispanic origin; 20% were less than age 31 (52% were over age 40); 37% had only some high school (34% had a high school diploma and 28% had at least some college); and 56% had no known drug history. Younger offenders, those with a history of drug abuse, and those with a lower educational background had higher probabilities of violating the conditions of supervised release. Offenders under age 30 were more likely to violate conditions of their supervision than offenders in other age groups -- 46% of those under age 30 violated a condition of their supervision compared to 23% of those over age 40. Releasees with a history of drug abuse were 4 times more likely to terminate their supervision for technical violations of drug use as were releasees who were not drug abusers. Releasees with known histories of drug abuse were also more likely to violate supervision with technical violations or new crimes. Similarly, releasees with lower levels of education were more likely to violate conditions of supervised release than those with higher levels of education. Those who did not graduate from high school had a 41% violation rate, those who were high school graduates had a 33% violation rate, those with some college had a 25% violation rate, and those with a college degree had a 12% violation rate. Parole (table 7.8) -- Among those who completed a term of parole, 96% were male; 63% were white; 86% were of non-Hispanic origin; 79% were over age 40; 43% had only some high school (33% had a high school diploma and 18% had at least some college); and 48% had no known drug history. Offenders with a history of drug abuse and those with a lower educational background had higher probabilities of violating the conditions of parole. Releasees with a history of drug abuse were three times more likely to terminate their supervision for technical violations of drug use than were releasees who were not drug abusers. Releasees with known histories of drug abuse were also more likely to violate supervision with technical violations or new crimes. Similarly, releasees with lower levels of education were more likely to violate conditions of supervised release than those with higher levels of education. Those who did not graduate from high school had a 45% violation rate, those who were high school graduates had a 42% violation rate, those with some college had a 28% violation rate, and those with a college degree had a 15% violation rate. Admissions, releases, and standing population of Federal prisoners (table 7.9) The Federal prison population grew by 9,670 persons during fiscal year 1998, increasing from 99,255 sentenced prisoners to 108,925. The greatest growth was among drug and public-order offenders, particularly immigration offenders. The number of drug offenders in prison grew by 4,847 in 1998. The number of immigration offenders increased by 1,970. In contrast, the number of weapons offenders grew at a slower rate, increasing by 680. Violent offenders increased by 815 offenders. Federal prisoners: First releases and time served (tables 7.12-7.14) During 1998, 34,504 prisoners were released from Federal prison for the first time after serving a sentence imposed by a U.S. district court. **(1Tables 7.12-7.14 include only prisoners first released by the BOP during 1988.)** Of these, 31,200 were released by standard means. These standard releases served an average of 28.0 months (table 7.12) in prison -- 91% of the sentence imposed, on average (table 7.14).**(2Average time to first release is the number of months in Bureau of Prison facilities. Time to first release does not include time detained pending adjudication and/or sentencing. Percent of total sentence obligation served until first release.)** The remaining 3,304 first releases were released by extraordinary means: 2,085 were released early due to participation in a drug treatment program; 223 had their sentences commuted; 400 were de-ported; 205 died; and 391 were released for other reasons including escape and an intensive confinement program (table 7.12). Drug offenders made up the vast majority of those released by extraordinary means (81%). Time served until first release varied by offense: Violent offenders served an average of 53.7 months; drug offenders served an average of 41.4 months; public-order offenders served an average of 18.1 months; and property offenders served an average of 15.5 months (figure 7.4). During 1998 the average time to first release for public-order offenders was influenced heavily by the relatively short time served by immigration offenders. Immigration offenders composed 61% of all released public-order offenders and, on average, served 11.1 months until first release. This was only half of the average time served for all public-order offenders (18.1 months). Exclusive of immigration offenders, public-order offenders served an average of 28.9 months until first release (not shown in a table). While violent and drug offenders served more time before first release, property and public-order offenders served more of their sentences than violent and drug offenders (figure 7.4). Old law offenders -- those who committed crimes before the implementation of Federal Sentencing Guidelines in November 1987 -- that were released in 1998 are not shown separately in any table. Only 3%, or 996, of released offenders were old law offenders. New law offenders released during 1998 served, on average, 26.3 months in prison (92% of the sentence imposed). Since the guidelines became effective as of November 1987, new law offenders sentenced to terms of imprisonment greater than 149 months could not have been released by standard means as of the end of fiscal year 1998.***Footnote 3: The Federal Sentencing Guidelines allow a prisoner up to 54 days per year for good behavior, so a prisoner sentenced to 149 months in prison in November 1987 could have served 127 months of his or her sentence and been released in 1998.)*** They could have been released for extraordinary reasons such as death, commutation, or treaty transfer. ***Footnote 4: Prisoners released for extraordinary reasons have been excluded from all time served and percent of sentence served calculations.*** Characteristics of Federal prisoners (tables 7.10, 7.15, and 7.16) Standing population (table 7.10) -- Of prisoners in Federal prison at the end of 1998, 93% were male, 57% were white, 40% were black, 28% were Hispanic, and 74% were U.S. citizens. Fifty-seven percent of the yearend population were 31 years of age or older. This portrait of the Federal prison population was similar across offenses and for new law offenders, except for the ethnicity and citizenship of violent and property offenders. More than 88% of these offenders were non-Hispanic and U.S. citizens. Prisoners released (table 7.15) -- Of prisoners released during 1998, 88% were male, 73% were white, 24% were black, 40% were Hispanic, and 61% were U.S. citizens. Fifty-five percent were 31 years of age or older at the time they were committed to Federal prison. These patterns were similar across offenses and for new law offenders, except for nonregulatory public-order offenses. Of released offenders convicted of those offenses, 64% were Hispanic and 66% were not U.S. citizens. Time served to first release (table 7.16)-- Of prisoners released by standard methods, males served more time than females, blacks served more time than whites, non-Hispanics served more time than Hispanics, older persons served more time than younger (until over age 40), and U.S. citizens served more time than noncitizens (figure 7.5). However, among the offenses, the overall pattern does not apply. For example, blacks served more time on average than whites, but among property offenders released from prison, whites served more time than blacks. Chapter Notes 1) Offenders were classified according to their most serious offense at conviction. In cases involving multiple offenses, the AOUSC offense severity hierarchy was applied. The most serious offense was the one with the most severe penalty imposed. If equal prison terms were imposed, or there was no imprisonment, the offense with the highest severity code, as determined by the U.S. Title and Code Criminal Offense Citations Manual, was selected. 2) Tables 7.1 and 7.2 were created using the probation, parole, and supervision data files of the Federal Probations Supervision Information System (FPSIS), which is maintained by the AOUSC. Yearend pending cases (that is, records with offenders who were under active supervision as of the end of the fiscal year, September 30, 1998) were selected. Corporate defendants were excluded from table 7.1. 3) Tables 7.3-7.8 were created using the FPSIS data files as well. Only records with one or more terminations of active supervision during October 1, 1997, through September 30, 1998, were selected. Each termination was counted separately. Technical violations and terminations for new crimes are shown only if supervision terminated with incarceration or removal from active supervision for reasons of a violation. Corporate defendants were excluded from tables 7.3-7.8. 4) Table 7.9-7.16 were created from the Federal Bureau of Prisons (BOP) data files. Tables 7.9 and 7.10 show sentenced prisoners in the custody of the BOP at the end of fiscal year 1998. Tables 7.11-7.16 include only prisoners committed by U.S. district courts or released by the BOP during fiscal year 1998. In these tables, a single person may be counted more than once if that person was committed into, or released from, the BOP more than one time during fiscal year 1998, or if that person appears in more than one column in a table, e.g., commitment and population. Offense categories in these tables were based on combinations of offense designations used by the BOP. They are similar, but may not be directly comparable, to the categories used in other tables of this Compendium. Offenses for some prisoners admitted or released were not able to be classified; these offenders were included in the totals and are shown as "other prisoners" on the last line of tables. In these tables, offenses were classified according to the offense associated with the longest single sentence actually imposed. Classifications in other tables may have been based on the longest potential sentence allowed by law. 5) Tables 7.9-7.16 distinguish between prisoners committed by U.S. district courts for violations of the U.S. Code and other prisoners. Prisoners released after commitment from U.S. district court are called "first releases." The other admissions and releases include offenders who returned to prison after their first release (such as probation, parole, or supervised release violators), offenders convicted in other courts (such as military or District of Columbia courts), and persons admitted to prison as material witnesses or for purposes of treatment, examination, or transfer to another authority. Offenders who entered or left a prison temporarily — such as for transit to another location, for health care, or to serve a weekend sentence — were not counted as admitted or released. Persons who were detained for deportation by the Immigration and Naturalization Service and who were not criminal offenders were also not included in the tables. Table 7.11 distinguishes between standard and extraordinary methods of release from prison. "Standard" releases include the usual methods of exiting from prison, such as full-term sentence expirations, expirations with good time, mandatory releases, and releases to parole. Extraordinary releases are unusual exits from prison, such as death, commutation, and transfer to another facility. 6) Tables 7.11-7.16 include only prisoners committed by U.S. district courts and released by the BOP during fiscal year 1998. Prisoners committed by U.S. district court — but not for a violation of the U.S. Code, or whose offense could not be classified — were included in the total but are not shown separately. Table 7.9, however, shows these prisoners separately. Other prisoners — such as probation and parole violators and prisoners committed by other courts, such as courts martial or District of Columbia Superior Court — were excluded from tables 7.11-7.16, as were other persons admitted to Federal prison but not committed from a U.S. district court. 7) In tables 7.11-7.16, time served was calculated for prisoner's arrival into jurisdiction of the BOP until first release from prison, plus any jail time served and credited. The calculation is the same as that currently used by the BOP. Prisoners serving consecutive sentences may have total imposed sentences exceeding the longest single sentence length. Accordingly, time actually served may exceed the longest single imposed sentence. The percent of sentenced served (in tables 7.12 and 7.14) is the average of each individual prisoner's percent of sentence served. Because other publications may include different groups of prisoners, calculate time served differently, or use a different offense classification, data in tables in this Compendium may differ from estimates of time served in previous publications by the BOP or in publications based on other data sources. Time served, as reported in tables 7.11-7.16 in this Compendium, may not be directly comparable to the calculation of time served in the 1993 and prior compendia. The methodology in this report uses additional information to identify prison commitment dates and account for jail credit. In previous reports, jail credits were overestimated in some cases. In addition, time served estimates reported in tables 7.12-7.14 and 7.16 include only prisoners released by standard methods (expirations, mandatory releases, and releases to parole). Prisoners released by extraordinary means (death, commutation, transfer, etc.) are excluded from the time served calculations in these tables. 8) Table 7.9 shows all persons admitted to, or released from, the jurisdiction of the BOP during fiscal year 1998. Prisoners counted in this column are the same as prisoners included in tables 7.11, 7.12, and 7.13. Table 7.9 shows in separate columns, as well as in the total, prisoners committed by a U.S. district court — but not for a violation of the U.S. code, or whose offense could not be classified. Other columns in table 7.9 include prisoners who were committed by other courts, returned to prison for violation of the conditions of supervision, or were received for examination, treatment, or transfer to another jurisdiction. Offenders who returned to prison for a violation of the conditions of supervision without a new court commitment, were classified according to the offense with the longest single sentence originally imposed at conviction. Methodology The Federal justice database Source of data The source of data for all tables in this Compendium is the Bureau of Justice Statistics (BJS) Federal justice database. The database is presently constructed from source files provided by the U.S. Marshals Service, the Executive Office for U.S. Attorneys, the Administrative Office of the U.S. Courts, the United States Sentencing Commission, and the U.S. Bureau of Prisons. The Administrative Office of the U.S. Courts also maintains the data collected by the Pretrial Services Agency, the U.S. Court of Appeals, and the U.S. Probation Service. Federal law prohibits the use of these files for any purposes other than research or statistics. A description of the source agency data files is provided in the table at the end of this section. Some records in the Federal justice database are matched according to a statistically weighted combination of names, other personal identifiers, dates of court appearances, types of offenses, and other relevant information contained in the files. Using the matched data files, it is possible to combine information about two or more stages of the processing of a criminal matter or case, for example from adjudication to probation or parole supervision. Reporting period Wherever possible, matters or cases have been selected according to some event which occurred during fiscal year 1998 (October 1, 1997, through September 30, 1998). Some data files provided by source agencies are organized according to a calendar year time frame; these have been combined and divided into fiscal years for purposes of the Compendium. Files which are organized by their source agencies according to fiscal year nonetheless include some pertinent records in later years' files. For example, tabulations of suspects in matters concluded during fiscal year 1998 in this Compendium have been assembled from source files containing records of 1998 matters concluded which were entered into the data system during fiscal years 1998 or 1999. Availability of data items The availability of particular items of information is affected by the data source. For example, data on prosecutors' decisions prior to court filing are provided for cases investigated by U.S. attorneys but not for those handled by other litigating divisions of the U.S. Department of Justice. Criminal Division cases enter the data base once they are filed in U.S. district court, however. Many items of social and demographic information come from presentence investigation records, supervision records, or sentencing records, and are available only for arrested defendants who were convicted and/or began serving a sentence involving supervised release. This particularly affects sex, race, ethnicity, and prior record information. Table construction and interpretation The tables presented report events that occurred during the Federal fiscal year — October 1, 1997 - September 30, 1998. Generally, the tables include both individual and organizational defendants. Organizational defendants are not included in tables describing pretrial release and detention or tables showing defendants sentenced to incarceration. Juvenile offenders are included in the reported statistics. Unit of analysis The unit of analysis in chapters 1 through 6 is a combination of a person (or corporation) and a matter or case. For example, if a single person is involved in three different criminal cases during the time period specified in the table, he or she is counted three times in the tabulation. Similarly, if a single criminal case involves a corporate defendant and four individual defendants, it counts five times in the tabulation. In chapter 7, the unit of analysis for incarceration, probation, parole, or other supervised release is a person entering custody or supervision, or a person leaving custody or supervision. For example, a person convicted in two concurrent cases and committed once to the custody of the Federal Bureau of Prisons in the indicated time period is counted as one admission to a term of incarceration. A person who terminates probation twice in the indicated time period, such as with a violation and again after reinstatement, is counted as two terminations of probation. Interpretation The tables in the Compendium are constructed to permit the user to make valid comparisons of numbers within each table and to compare percentage rates across tables. The total numbers of subjects in Compendium tables that are based on records linked between two files are generally less than the total number of records in either source file. Accordingly, comparisons of absolute numbers across two or more tables in this volume and other data sources are not necessarily valid. Offense classifications Procedure The offense classification procedure used in this Compendium is based on the classification system followed by the Administrative Office of the U.S. Courts. Specific offenses combined to form the BJS categories shown in the Compendium tables.*(These categories correspond to the Bureau of Justice Statistics crime definitions and, to the extent possible, are organized and presented consistent with BJS publications on State criminal justice systems.) Offense categories for Federal arrestees in chapter 1 are based on the FBI's National Crime Information Center (NCIC) offense classifications, which are converted into U.S. Marshals' four-digit offense codes, which, in turn are aggregated into the offense categories shown in the tables. These categories are similar, but may not be directly comparable to the BJS offense categories used in other chapters of this Compendium. For data from the Executive Office for U.S. Attorneys, which include United States Code citations but do not include the Administrative Office offense classifications, United States Code titles and sections are translated into the Administrative Office classification system and then aggregated into the offense categories used in the tables. Offense categories for prisoners in chapter 7 are based on combinations of offense designations used by the Bureau of Prisons. They are similar to the categories in other chapters and other tables in chapter 7, but may not be directly comparable. Felony/misdemeanor distinctions Felony and misdemeanor distinctions are provided where possible. Felony offenses are those with a maximum penalty of more than 1 year in prison. Misdemeanor offenses are those with a maximum penalty of 1 year or less. Felonies and misdemeanors are further classified using the maximum term of imprisonment authorized. Section 3559, U.S. Code, Title 18 classifies offenses according to the following schedule: Felonies Class A felony — life imprisonment, or if the maximum penalty is death. Class B felony — 25 years or more. Class C felony — less than 25 years but more than 10 years. Class D felony — less than 10 years but more than 5 years. Class E felony — less than 5 years but more than 1 year. Misdemeanors Class A misdemeanor — 1 year or less but more than 1 month. Class B misdemeanor — 6 months or less but more than 30 days. Class C misdemeanor — 30 days or less but more than 5 days. Infraction — 5 days or less, or if no imprisonment is authorized. In this Compendium, felony and misdemeanor distinctions are provided where the data permit these distinctions. Chapter 2 does not use this distinction because many suspects cannot be so classified at the investigation stage in the criminal justice process. Chapter 3 no longer reports this distinction because the Pretrial Services Agency no longer gathers this information. Chapters 4 and 5 distinguish between felony and misdemeanor offenses, as do tables 7.1–7.6. Tables 7.7–7.11 follow the convention of other BJS publications by separately tabulating offenders whose actual imposed sentences are less than or equal to 1 year and those whose actual sentences are greater than 1 year. Most serious offense selection Where more than one offense is charged or adjudicated, the most serious offense (the one that may or did result in the most severe sentence) is used to classify offenses. The offense description may change as the criminal justice process proceeds. Tables indicate whether investigated, charged or adjudicated offenses are used. In chapter 2, the most serious offense is based on the criminal lead charge as determined by the assistant U.S. attorney responsible for the criminal proceeding. In chapter 3, the major charged offense is based upon the Administrative Office's offense severity classification system, as determined by the pretrial officer responsible for the case. To select this offense, the officer ranks offenses according to severity based on maximum imprisonment, type of crime, and maximum fines. In chapter 4, the most serious offense charged is the one that has the most severe potential sentence. For chapter 5, conviction offenses are based on statutory maximum penalties. In chapter 6, offenses are classified by the offense of conviction. In tables 7.1–7.6, the most serious offense of conviction is either the one having the longest sentence imposed or, if equal sentences were imposed or there was no imprisonment, it was the offense carrying the highest severity code as determined by the Administrative Office's offense severity code ranking. In tables 7.7–7.13, prisoners are classified according to the offense which bears the longest single incarceration sentence. Offense categories For offense categories in all text tables, the following conditions apply: "Murder" includes nonnegligent manslaughter. "Other sex offenses" may include some nonviolent offenses. "Fraud" excludes tax fraud. "Larceny" excludes transportation of stolen property. "Other property felonies" excludes fraudulent property offenses, and includes destruction of property and trespass. "Tax law violations" includes tax fraud. "Obscene material" denotes the mail or transport thereof. "All other felonies" includes felonies with unknown or unclassifiable offense type. "Misdemeanors" includes misdemeanors, petty offenses, and unknown offense levels. "Drug possession" also includes other drug misdemeanors. Figure S.2. Figure S.2. uses data from the Executive Office for U.S. Attorneys and the Administrative Office of the U.S. Courts to provide an overview of case processing in the Federal criminal justice system. The figure uses data from two separate cohorts of persons involved in criminal proceedings. The data on suspects in criminal matters concluded and defendants prosecuted come from the cohort of suspects in matters concluded during October 1, 1997, through September 30, 1998. (The filing of charges in a criminal case is one conclusion of a matter; the other is the declination of the matter for prosecution.) For the other three outcomes in figure S.2.— offenders convicted, sentenced to prison, or given a term of probation — the data come from the cohort of defendants in cases terminating during fiscal year 1998 in U.S. district courts (by either a judge or a magistrate). A total of 106,022 suspects were investigated in matters concluded by U.S. attorneys during 1998 (table 2.2). Seventy-three percent of these suspects were prosecuted in U.S. district court, either before U.S. district court judges (64,993/106,022 = 61.3%) or before U.S. magistrates (12,243/ 106,022 = 11.5%). The conviction data in figureS.2. refer to defendants in cases terminated in U.S. district court during 1998, whether before U.S. district court judges or U.S. magistrates. Fifty-seven percent (60,958/106,022) of the defendants in cases terminating during 1998 were convicted (table 4.2). Of the defendants investigated, 41% (43,041/106,022) were reported as sentenced to a term of imprisonment, and 12% (12,744/106,022) were sentenced to a term of probation without imprisonment (table 5.2). These figures are based on all charges, whether they eventually were felony or misdemeanor offenses. Source agencies for Compendium data tables Data source agency -- data file United States Marshals Service (USMS) — Prisoner Tracking System (PTS) Description of data files contents Contains data on suspects arrested for violations of Federal law, by Federal enforcement agencies. The data include information on characteristics of Federal arrestees. Compendium tables Arrest: 1.1, 1.2, 1.3 Data source agency -- data file Executive Office for U.S. Attorneys (EOUSA) — Central System and Central Charge Files Description of data files contents Contains information on the investigation and prosecution of suspects in criminal matters received and concluded, criminal cases filed and terminated, and criminal appeals filed and handled by U.S. attorneys. The central system files contain defendant-level records about the processing of matters and cases; the central charge files contain the records of the charges filed and disposed in criminal cases. Data are available on matters and cases filed, pending, and terminated. Compendium tables Prosecution: 2.1, 2.2, 2.3, 2.4, 2.5, 2.6 Data source agency -- data file AOUSC: Pretrial Services Agency (PSA) — Pretrial Services Act Information System Description of data files contents Contains data on defendants interviewed, investigated, or supervised by pretrial services. The information covers defendants' pretrial hearings, detentions, and releases from the time they are interviewed through the disposition of their cases in district court. Compendium tables Pretrial release: 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 3.10 Adjudication: 4.5 (defendant characteristics )Sentencing: 5.4, 5.5, 5.6 (defendant characteristics) Data source agency -- data file Administrative Office of the U.S. Courts (AOUSC) — Criminal Termination Files Description of data files contents Contains information about the criminal proceedings against defendants those cases were filed in U.S. district courts. Includes information on felony defendants, Class A misdemeanants — whether handled by U.S. district court judges or U.S. magistrates — and other misdemeanants provided they were handled by U.S. district court judges. The information in the data files cover criminal proceedings from case filing through disposition and sentencing. Data are available on criminal defendants in cases filed, pending, and terminated. Compendium tables Adjudication: 4.1, 4.2, 4.3, 4.4, 4.5 Sentencing: 5.1, 5.2, 5.3, 5.4, 5.5, 5.6 Data source agency -- data file United States Sentencing Commission (USSC) — Monitoring Data Base Description of data files contents Contains information on criminal defendants sentenced pursuant to the provisions of the Sentencing Reform Act of 1984. It is estimated that more than 90% of felony defendants in the Federal criminal justice system are sentenced pursuant to the SRA of 1984. Data files are limited to those defendants whose records have been obtained by the U.S. Sentencing Commission. Compendium tables Adjudication: 4.5 (defendant characteristics) Sentencing: 5.4, 5.5, 5.6 (defendant characteristics) Data source agency -- data file AOUSC: Court of Appeals Description of data files contents Contains information on criminal appeals filed and terminated in U.S Courts of Appeals. Records of appeals filed, pending, or terminated include information on the nature of the criminal appeal, the underlying offense, and the disposition of the appeal. Compendium tables Appeals: 6.1, 6.2, 6.3, 6.4, 6.5 Data source agency -- data file AOUSC — Federal Probation and Supervision Information System (FPSIS) Description of data files contents Contains information about supervisions provided by probation officers for persons placed on probation or supervised release from prison. The files contain records of individuals entering, or currently on supervision, as well as records of offenders terminating supervision. Compendium tables Corrections: 7.1, 7.2, 7.3, 7.4, 7.5, 7.6, 7.7, 7.8 Data source agency -- data file Bureau of Prisons (BOP): Extract from BOP's online Sentry System Description of data files contents The data extracts contain information on all offenders released from prison over a specific period of time plus information about the offenders in prison when the data extracts are made. The information covers the time that offenders enter prison until their release from the jurisdiction of the Bureau of Prisons. Compendium tables Corrections: 7.9, 7.10, 7.11, 7.12, 7.13, 7.14, 7.15, 7.16 Glossary Acquittal — legal judgment that a criminal defendant has not been proven guilty beyond a reasonable doubt of the charges against him. Affirmed — in the appellate courts, the decree or order is declared valid and will stand as rendered in the lower court. Agriculture violations — violations of the Federal statutes on agriculture and conservation: for example, violations of the Agricultural Acts, Insecticide Act, and the Packers and Stockyards Act, 1921; also violations of laws concerning plant quarantine and the handling of animals pertaining to research (7 U.S.C., except sections on food stamps (fraud); also 16 U.S.C. sections relating generally to violations in operating public parks, such as trespassing for hunting, shooting, and fishing). Antitrust violations — offenses relating to Federal antitrust statutes, which aim to protect trade and commerce from unlawful restraints, price fixing, monopolies (for example, 15 U.S.C. §§ 1, 3, 8, 20, and 70(i)), and