U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Recidivism of Sex Offenders Released from Prison in 1994 November 2003, NCJ 198281 -------------------------------------------------------------- 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/rsorp94.htm -------------------------------------------------------------- Lawrence A. Greenfeld, Director Patrick A. Langan, Erica L. Schmitt, and Matthew R. Durose, all BJS statisticians, wrote this report. Carolyn Williams and Tom Hester edited and produced it. Contents Introduction and highlights Definitions Imprisonment offense Sex offender Rapist Sexual assaulter Child molester Statutory rapist First release Prior arrest Rearrest Reconviction Returned to prison -- with a new sentence -- with or without a new sentence Demographic characteristics All sex offenders (table 1) Rapists and sexual assaulters (table 1) Child molesters and statutory rapists (table 2) Sentence length and time served All sex offenders (table 3) Rapists and sexual assaulters (table 3) Child molesters and statutory rapists (table 4) Prior criminal record All sex offenders (table 5) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 5) Child molesters and statutory rapists (table 6) Four measures of recidivism Four measures All sex offenders (table 7) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 7) Child molesters and statutory rapists (table 8) Time to recidivism All sex offenders (table 9) Rapists and sexual assaulters (table 9) Child molesters and statutory rapists (table 10) Rearrest for any type of crime Demographic characteristics All sex offenders (table 11) Rapists and sexual assaulters (table 11) Child molesters and statutory rapists (table 12) Time served before 1994 release All sex offenders (table 13) Rapists and sexual assaulters (table 13) Child molesters and statutory rapists (table 14) Prior arrest for any type of crime All sex offenders (table 15) Rapists and sexual assaulters (table 15) Child molesters and statutory rapists (table 16) Number of prior arrests for any type of crime All sex offenders (table 17) Rapists and sexual assaulters (table 17) Child molesters and statutory rapists (table 18) State where rearrested for any type of crime All sex offenders (table 19) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 19) Child molesters and statutory rapists (table 20) Rearrest and reconviction for a new sex crime Rearrest and reconviction All sex offenders (table 21) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 21) Child molesters and statutory rapists (table 22) Time to rearrest All sex offenders (table 23) Rapists and sexual assaulters (table 23) Child molesters and statutory rapists (table 24) Demographic characteristics All sex offenders (table 25) Time served before 1994 release All sex offenders (table 26) Prior arrest for any type of crime All sex offenders (table 27) Rapists and sexual assaulters (table 27) Child molesters and statutory rapists (table 28) Number of prior arrests for any type of crime All sex offenders (table 29) Prior arrest for a sex crime All sex offenders (table 30) Rapists and sexual assaulters (table 30) Child molesters and statutory rapists (table 31) State where rearrested for a sex crime All sex offenders (table 32) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 32) Child molesters and statutory rapists (table 33) Rearrest for a sex crime against a child Undercounts of sex crimes against children No data on precise ages of molested children Rearrest All sex offenders (table 34) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 34) Child molesters and statutory rapists (table 35) Prior arrest for a sex crime against a child All sex offenders (table 36) Rapists and sexual assaulters (table 36) Child molesters and statutory rapists (table 37) Molester's and child's age at time of imprisonment offense Child molesters (table 38) State where rearrested for a sex crime against a child All sex offenders (table 39) Sex offenders compared to non-sex offenders Rapists and sexual assaulters (table 39) Child molesters and statutory rapists (table 40) Rearrest for other types of crime All sex offenders (table 41) Rapists and sexual assaulters (table 41) Child molesters and statutory rapists (table 42) Victims of sex crimes Characteristics of victims of rape or sexual assault (unnumbered table) Methodology Number of sex offenders released from State prisons in 1994 and number selected for this report, by State (Appendix table) Introduction and highlights Introduction In 1994, prisons in 15 States released 9,691 male sex offenders. The 9,691 men are two-thirds of all the male sex offenders released from State prisons in the United States in 1994. This report summarizes findings from a survey that tracked the 9,691 for 3 full years after their release. The report documents their "recidivism," as measured by rates of rearrest, reconviction, and reimprisonment during the 3-year followup period. This report gives recidivism rates for the 9,691 combined total. It also separates the 9,691 into four overlapping categories and gives recidivism rates for each category: * 3,115 released rapists * 6,576 released sexual assaulters * 4,295 released child molesters * 443 released statutory rapists. The 9,691 sex offenders were released from State prisons in these 15 States: Arizona, Maryland, North Carolina, California, Michigan, Ohio, Delaware, Minnesota, Oregon, Florida, New Jersey, Texas, Illinois, New York, and Virginia. Highlights The 15 States in the study released 272,111 prisoners altogether in 1994. Among the 272,111 were 9,691 men whose crime was a sex offense (3.6% of releases). On average the 9,691 sex offenders served 31/2 years of their 8- year sentence (45% of the prison sentence) before being released in 1994. Rearrest for a new sex crime Compared to non-sex offenders released from State prisons, released sex offenders were 4 times more likely to be rearrested for a sex crime. Within the first 3 years following their release from prison in 1994, 5.3% (517 of the 9,691) of released sex offenders were rearrested for a sex crime. The rate for the 262,420 released non- sex offenders was lower, 1.3% (3,328 of 262,420). The first 12 months following their release from a State prison was the period when 40% of sex crimes were allegedly committed by the released sex offenders. Recidivism studies typically find that, the older the prisoner when released, the lower the rate of recidivism. Results reported here on released sex offenders did not follow the familiar pattern. While the lowest rate of rearrest for a sex crime (3.3%) did belong to the oldest sex offenders (those age 45 or older), other comparisons between older and younger prisoners did not consistently show older prisoners' having the lower rearrest rate. The study compared recidivism rates among prisoners who served different lengths of time before being released from prison in 1994. No clear association was found between how long they were in prison and their recidivism rate. Before being released from prison in 1994, most of the sex offenders had been arrested several times for different types of crimes. The more prior arrests they had, the greater their likelihood of being rearrested for another sex crime after leaving prison. Released sex offenders with 1 prior arrest (the arrest for the sex crime for which they were imprisoned) had the lowest rearrest rate for a sex crime, about 3%; those with 2 or 3 prior arrests for some type of crime, 4%; 4 to 6 prior arrests, 6%; 7 to 10 prior arrests, 7%; and 11 to 15 prior arrests, 8%. Rearrest for a sex crime against a child The 9,691 released sex offenders included 4,295 men who were in prison for child molesting. Of the children these 4,295 men were imprisoned for molesting, 60% were age 13 or younger. Half of the 4,295 child molesters were 20 or more years older than the child they were imprisoned for molesting. On average, the 4,295 child molesters were released after serving about 3 years of their 7-year sentence (43% of the prison sentence). Compared to the 9,691 sex offenders and to the 262,420 non-sex offenders, released child molesters were more likely to be rearrested for child molesting. Within the first 3 years following release from prison in 1994, 3.3% (141 of 4,295) of released child molesters were rearrested for another sex crime against a child. The rate for all 9,691 sex offenders (a category that includes the 4,295 child molesters) was 2.2% (209 of 9,691). The rate for all 262,420 non-sex offenders was less than half of 1% (1,042 of the 262,420). Of the approximately 141 children allegedly molested by the child molesters after their release from prison in 1994, 79% were age 13 or younger. Released child molesters with more than 1 prior arrest for child molesting were more likely to be rearrested for child molesting (7.3%) than released child molesters with no more than 1 such prior arrest (2.4%). Rearrest for any type of crime Compared to non-sex offenders released from State prison, sex offenders had a lower overall rearrest rate. When rearrests for any type of crime (not just sex crimes) were counted, the study found that 43% (4,163 of 9,691) of the 9,691 released sex offenders were rearrested. The overall rearrest rate for the 262,420 released non-sex offenders was higher, 68% (179,391 of 262,420). The rearrest offense was a felony for about 75% of the 4,163 rearrested sex offenders. By comparison, 84% of the 179,391 rearrested non-sex offenders were charged by police with a felony. Reconviction for a new sex crime Of the 9,691 released sex offenders, 3.5% (339 of the 9,691) were reconvicted for a sex crime within the 3-year followup period. Reconviction for any type of crime Of the 9,691 released sex offenders, 24% (2,326 of the 9,691) were reconvicted for a new offense. The reconviction offense included all types of crimes. Returned to prison for any reason Within 3 years following their release, 38.6% (3,741) of the 9,691 released sex offenders were returned to prison. They were returned either because they received another prison sentence for a new crime, or because of a technical violation of their parole, such as failing a drug test, missing an appointment with their parole officer, or being arrested for another crime. Denfinitions Imprisonment offense The 9,691 prisoners were men released from State prisons in 1994 after serving some portion of the sentence they received for committing a sex crime. The sex crime they committed is referred to throughout the report as their "imprisonment offense." Their imprisonment offense should not be confused with any new offense they may have committed after release. Sex offender The 9,691 released men were all violent sex offenders. They are called "violent" because the crimes they were imprisoned for are widely defined in State statutes as "violent" sex offenses. "Violent" means the offender used or threatened force in the commission of the crime or, while not actually using force, the offender did not have the victim's "factual" or "legal" consent. Factual consent means that, for physical reasons, the victim did not give consent, such as when the offender had intercourse with a sedated hospital patient or with a woman who had fallen unconscious from excessive drug taking. "Legal" consent means that the victim willingly participated but, in the eyes of the law, the victim was not old enough or not sufficiently mentally capable (perhaps due to mental illness or mental retardation) to give his or her "legal" consent. State statutes give many different names to violent sex offenses: "forcible rape," "statutory rape," "object rape," "sexual assault," "sexual abuse," "forcible sodomy," "sexual misconduct," "criminal sexual conduct," "lascivious conduct," "carnal abuse," "sexual contact," "unlawful sexual intercourse," "sexual battery," "unlawful sexual activity," "lewd act with minor," "indecent liberties with a child," "carnal knowledge of a child," "incest with a minor," and "child molesting." "Violent" sex offenses are distinguished from "nonviolent" sex offenses and from "commercialized sex offenses." Nonviolent sex offenses include morals and decency offenses (for example, indecent exposure and peeping tom), bestiality and other unnatural acts, adultery, incest between adults, and bigamy. Commercialized sexual offenses include prostitution, pimping, and pornography. As used throughout this report, the terms "sex crimes" and "sex offenders" refer exclusively to violent sex offenses. Each of the 9,691 sex offenders in this report is classified as either a rapist or a sexual assaulter. Classification was based on information about the imprisonment offense contained in prison records supplied for each sex offender released from prison in 1994. Also based on imprisonment offense information, an inmate could be categorized as a child molester and/or a statutory rapist. Classification to either of these two categories is in addition to, not separate from, classification as a rapist or sexual assaulter. For example, of the 3,115 sex offenders classified as rapists, 338 were child molesters. Or, to put it another way, the imprisonment offense for 338 of the 4,295 child molesters identified in this report was rape. Similarly, 3,957 of the 4,295 child molesters were also sexual assaulters. The report gives statistics for all sex offenders and each of the four types -- rapists, sexual assaulters, child molesters, and statutory rapists. (See Methodology on page 37 for details on how sex offenders were separated into categories.) Rapist "Violent sex crimes" are separated into two categories: "rape" (short for "forcible rape") and "other sexual assault." As used throughout this report the term "rapist" refers to a released sex offender whose imprisonment offense was defined by State law as forcible intercourse (vaginal, anal, or oral) with a female or male. Rape includes "forcible sodomy" and "penetration with a foreign object." Rape excludes statutory rape or any other nonforcible sexual act with a minor or with someone unable to give legal or factual consent. As used throughout this report, "rape" always means "forcible rape." "Statutory rape" is not a type of forcible rape. A total of 3,115 sex offenders are identified in the report as released rapists -- about a third (32%) of the 9,691 released sex offenders. However, enough information to clearly distinguish rapists from other sexual assaulters was not always available in the prison records used to categorize sex offenders into different types. Consequently, the number of rapists among the 9,691 was almost certainly greater than 3,115; how much greater is unknown. An obstacle to identifying rapists from penal code information is that the label "rape" is not used in about half the 50 States. However, released sex offenders whose imprisonment offense was rape could still be identified. To illustrate, in one State, the term criminal sexual conduct refers to all types of sex crimes. The statutory language was consulted to determine if an offender's imprisonment offense involved "intercourse" that was "forcible," in accordance with the definition of rape used in this report. If the offense was not found to involve intercourse (or penetration), then the inmate was not classified as a rapist. The same was true of force; if the statutory language did not include a reference to force (or coercion), the offense was not categorized as rape. Sexual assaulter By definition in the report, all sex offenders are either "rapists" or "sexual assaulters." Sex offenders whose imprisonment offense could not be positively identified as "rape" were placed in the "sexual assault" category. To the extent that rapists were reliably distinguished from sexual assaulters, "sexual assaulters" identified in this report were released sex offenders whose imprisonment offense was 'sexual assault," defined as one of the following: 1. forcible sexual acts, not amounting to intercourse, with a victim of any age, 2. nonforcible sexual acts with a minor (such as statutory rape or incest with a minor or fondling), or 3. nonforcible sexual acts with someone unable to give legal or factual consent because of mental or physical reasons (for example, a mentally ill or retarded person or a sedated hospital patient). A total of 6,576 sex offenders are identified in this report as released sexual assaulters. The 6,576 sexual assaulters made up about two-thirds (68%) of the 9,691 released sex offenders. Child molester Many of the 9,691 sex offenders were released prisoners whose imprisonment offense was the rape or sexual assault of a child. Throughout the report, released sex offenders whose forcible or nonforcible sex crime was against a child are referred to as "child molesters." The sex crime did not have to involve intercourse to fit the definition of child molestation. Of the 9,691 sex offenders, 4,295 were identified as child molesters based on prison records made available for the study. However, because complete information was not always supplied, not every child molester could be identified. Of the 9,691 released sex offenders, undoubtedly more than 4,295 were child molesters, but 4,295 represent all who could be identified from the information available. One reason child molesters were not easily identified from penal code information is that most States do not use the term "child molester" in their penal code. Nevertheless, all States have laws against sexual activity with children, which does facilitate identification. As a result of the uncertainty regarding the number of child molesters among the 9,691 sex offenders, the study cannot say what percentage of the victims of the 9,691 sex offenders' offenses were children, and what percentage were adults. In short, the 4,295 released child molesters in this report were men who -- a. had forcible intercourse with a child or b. committed "statutory rape"(meaning nonforcible intercourse with a child)or c. with or without force, engaged in any other type of sexual contact with a child. Of the 4,295, at least 338 (about 8%) had forcible intercourse, and at least 443 (10%)committed statutory rape. Statutory rapist State laws define various circumstances in which intercourse between consenting partners is illegal: for example, when one of the partners is married or when the two are blood relatives or when one is a "child." Laws that criminalize consensual intercourse based solely on the marital status of the partners are called "adultery laws." Those that criminalize it based solely on blood relationship are "incest laws." Laws that prohibit consensual sexual intercourse based solely on the ages of the partners are called "statutory rape laws." Statutory rape pertains exclusively to consensual intercourse, as opposed to other types of sexual contact with a child, such as forcible intercourse, forcible fondling, or consensual fondling. Statutory rape is one specific form of what this study calls "child molestation." The child victim of statutory rape can be male or female, and the offender can be male or female. The offender can be almost any relative ("statutory rape" includes incest with a child), an unrelated person well known to the child (such as a school teacher, neighbor, or minister), someone the child hardly knows, or a stranger. Statutory rape laws define a "child" as a person who is below the "age of consent," meaning below the minimum age at which a person can legally consent to having intercourse. Age of consent in the 50 States ranges from 14 to 18. Most States set age of consent at 16. In those States, consensual intercourse with someone age 16 or older is usually not a criminal offense, but intercourse with someone below 16 generally is. However, all States make exceptions to their age rules. Consequently, consensual intercourse with children below the age of consent is not always a crime, and consensual intercourse with children who are old enough to give consent is not always legally permissible. Exceptions for children below age of consent Certain statutory exceptions exist to legal prohibitions against nonforcible intercourse with children who are below the age of consent. One way exceptions are made in statutes is by specifying the minimum age the offender must be (for example, at least age 18, at least age 20) for intercourse to be unlawful. Persons below this minimum age generally cannot be prosecuted. Another common way exceptions are made (virtually every State has these provisions in its laws) is by specifying how much older than the victim the perpetrator must be for criminal prosecution to occur. For example, by law in one State where age of consent is 16, no prosecution can occur unless the age difference is at least 3 years. In that State it is legal for a 17-year-old to have consensual intercourse with a 15-year-old, even though 15 is below the age of consent; but the same act with a 15-year- old is illegal when the other is 18. That is because the 17- year-old is not 3 years older than the 15-year-old, whereas the 18-year-old is. The aim of such exceptions is to distinguish teen behavior from exploitative relationships between adults and children. Another exception is consensual intercourse between husband and wife; no prosecution can occur if one spouse is below the age of consent. Exceptions for children old enough to give consent Certain adults can be prosecuted for having consensual intercourse with a child who has reached the age of consent. For example, in one State it is a third degree felony for a psychotherapist to have intercourse with a 17-year-old client even though 17 is over the minimum age of consent in that State. In another State, where an adult generally cannot be prosecuted for having consensual intercourse with a 16-year-old, an exception is made when the adult is the child's school teacher. In that case the teacher can be prosecuted for a "class A" misdemeanor. Exceptions are made for other professions as well (clergy, for example). In this report, 443 of the 9,691 released sex offenders are identified as statutory rapists based on information supplied by the prisons that released them. There were more than 443 statutory rapists among the 9,691 released male sex offenders, but the 443 are all that could be positively identified with the limited information available. One reason statutory rapists are not easily identified from penal code information available on the released sex offenders is that most States do not use the term "statutory rape" in their laws. First release Though all 9,691 sex offenders in the study were released in 1994, for a fourth of the offenders 1994 was not the first year of release since receiving their prison sentence. This group had previously served a portion of the sentence and were released, then violated parole and were returned to prison to continue serving time still left on that sentence. For the remaining 75% of sex offenders released, the 1994 release was their "first release," meaning their first discharge from prison since being convicted and sentenced to prison. "First release" should not be confused with first ever release from a prison. "First release" pertains solely to the sentence for the imprisonment offense (as defined above). It does not pertain to any earlier prison sentences offenders may have served for some other offense. Attention is drawn to first releases because certain statistics in the report -- for example, "average time served," "percent of sentence served," "child molester's age when he committed the sex crime for which he was imprisoned" -- could only be computed for those prisoners classified as first releases. For such statistics, date first admitted to prison for their imprisonment offense was needed. Since prison records made available for the study only provided this admission date on first releases, first releases necessarily formed the basis for the statistics. Prior arrest Statistics on prior arrests were calculated using arrest dates from the official criminal records of the 9,691 released sex offenders. Only dates of arrest were counted, not the number of arrest charges associated with that arrest date. To illustrate, one man was arrested on March 5, 1970, and that one arrest resulted in 3 separate arrest charges being filed against him. In this study, that March 5 arrest is considered one prior arrest. Prior arrests were measured two different ways in this report. The first way did not include the imprisonment offense for which the sex offender was in prison in 1994. Prior arrest statistics that did not include the imprisonment offense are found in sections of the report that describe the criminal records of the 9,691 sex offenders at the time of release from prison. In this case, any arrest that had occurred on a date prior to the sex offender's arrest for his imprisonment offense was considered a prior arrest. For example, one released sex offender was found to have four different dates of arrest prior to the date of arrest for his imprisonment offense. Those four arrests resulted in 17 different charges being brought against him. When describing this released prisoner's criminal record, he is considered to have four prior arrests. The second way of measuring prior arrests did include the imprisonment offense of the released sex offender. Prior arrest statistics that did include the imprisonment offense are found in sections of the report that describe the recidivism rates of the 9,691 sex offenders following their release from prison. In this case, any arrest that had occurred on a date prior to the sex offender's release from prison was considered a prior arrest. By definition, all 9,691 sex offenders had at least one arrest prior to their release, which was the sex crime arrest responsible for their being in prison in 1994. This means that the sex offender who was arrested on four different dates prior to the arrest for his imprisonment offense under the first definition of prior arrest was, under this second definition, classified as having five prior arrests, once his imprisonment offense is included. Thirteen tables in the report provide statistics on prior arrests (and, in 2 of the 13, prior convictions and prior imprisonments). In tables 15, 16, 17, 18, 27, 28, 29, 30, 31, 36, and 37, "prior arrests" includes the sex crime arrest for the imprisonment offense; these tables have the heading "prior to 1994 release." In tables 5 and 6,"prior arrests" excludes that arrest; these tables have the heading "prior to the sex crime for which imprisoned." In all tables, the same counting rule was used: arrest dates, not arrest charges, were counted to obtain the number of prior arrests. Rearrest Unless stated otherwise, this recidivism measure is defined as the number or percentage of released prisoners who, within the first three years following their 1994 release, were arrested either in the same State that released them (in this report those arrests are called "in-State" arrests) or in a different State (those arrests are referred to as "out-of- State" arrests). Data on arrests came from State RAP sheets and FBI RAP sheets. RAP sheets (Records of Arrest and Prosecution) are law enforcement records intended to document a person's entire adult criminal history, including every arrest, prosecution and adjudication for a felony or serious misdemeanor offense. Arrests, prosecutions and adjudications for minor traffic offenses, public drunkenness, and other petty crimes are not as fully recorded as those for serious crimes. The "percent rearrested" is calculated by dividing the number rearrested by the number released from prison in 1994. All measures of recidivism based on criminal records are subject to two types of errors. Type 1 errors arise when the arrest or the conviction in the released prisoner's record is for a crime that person did not commit. Type 2 errors arise when the released prisoner commits a crime but he is not arrested for it, or, even if he is, the arrest does not result in his conviction. Some amount of type 1 and type 2 error is inevitable, however recidivism is measured. But that does not mean that all recidivism measures are equally suitable, no matter the purpose they are intended to serve. The main purpose of this recidivism study was to document the percentage of sex offenders who continued their involvement in various types of crime after their release from prison in 1994. The more suitable measure for that is the one with the fewest type 2 errors: the one, in other words, less prone to saying someone is not committing crimes when he actually is. Between rearrest and reconviction as the recidivism measure, the one less likely to make that type of error is rearrest. One reason is that the rigorous standard used to convict someone -- "proof beyond a reasonable doubt" -- makes it certain that guilty persons will sometimes go free. Another reason is record keeping: the justice system does better at recording arrests than convictions in RAP sheets. For such reasons, this study uses rearrest more often than reconviction as the measure of recidivism. Rearrest forms a conservative measure of reoffending because many crimes do not result in arrest. Not all types of crime are alike in this regard. Crimes committed in nonpublic places (such as in the victim's home) by one family member against another (such as by the husband against his wife, or by the father against his own child) are a type that is less likely than many other types to be reported to police and, consequently, less likely to result in arrest. Sex crimes, particularly those against children, are a specific example of this type. While some sex offenders in this study probably committed a new sex crime after their release and were not arrested or convicted, the study cannot say how many. As mentioned above, one reason why sex offenders are not arrested is that no one calls the police. Results from the National Crime Victimization Survey indicate that the offenses of rape/sexual assault are the least likely crimes to be reported to the police. (See Reporting Crime to the Police, 1993-2000, March 2003, .) Reconviction Except where stated otherwise, this recidivism measure pertains to State and Federal convictions in any State (not just convictions in the State that released them) in the three years following release. Information on convictions came from State and FBI RAP sheets. RAP sheets are intended to document every conviction for a felony or serious misdemeanor, but not every conviction for a minor offense. "Percent reconvicted" is calculated by dividing the number reconvicted by the number released from prison in 1994. (It is not calculated by dividing the number reconvicted by the number rearrested.) Return to prison Two recidivism measures are returned to prison -- with a new sentence with or without a new sentence. Recidivism defined as Returned to prison with a new sentence pertains exclusively to sex offenders who, within 3 years following release, were reconvicted for any new crime in any State following their release and received a new prison sentence for the new crime. Recidivism defined as Returned to prison with or without a new sentence includes resentenced offenders plus any who were returned to prison within 3 years because they had violated a technical condition of their release. Technical violations include things such as failing a drug test, missing an appointment with their parole officer, or being arrested for a new crime. Offenders returning to prison for such violations are sometimes referred to as "technical violators." Prisons should not be confused with jails. A prison is a State or Federal correctional facility reserved for convicted persons with relatively long sentences (generally over a year). A jail is a local correctional facility for convicted persons with short sentences or for persons awaiting trial. Returns to prison refer to any prison, not necessarily the same prison that released the offender in 1994. The "percent returned to prison with a new sentence" is calculated by dividing the number returned to prison with a new sentence by the number released from prison in 1994. The "percent returned to prison with or without a new sentence" is calculated by dividing the number returned to prison with or without a new sentence by the number released from prison in 1994. Data on returns with a new sentence are based on State and FBI RAP sheets. Data on returns with or without a new sentence are based on State and FBI RAP sheets plus prison records. Demographic characteristics All sex offenders Of the 9,691 released sex offenders, approximately -- * 6,503 (67.1% of the 9,691) were white males * 3,053 (31.5%) were black males * 136 (1.4%) were males of other races (Asian, Pacific Islander, American Indian, and Alaska Native). The vast majority of sex offenders were non-Hispanic males (80.1%). Half were over the age of 35 when released. Rapists and sexual assaulters As defined in this report, all sex offenders are either "rapists" or "sexual assaulters." Of the 9,691 released sex offenders, 3,115 were rapists and the remaining 6,576 were sexual assaulters. Of the 3,115 rapists, 1,735 (55.7% of 3,115) were white males and 1,327 (42.6%) were black males. Of the 6,576 sexual assaulters, 4,768 (72.5% of 6,576) were white males and 1,723 (26.2%) were black males. Rapists and sexual assaulters were close in age at time of release: over 70% were age 30 or older. Median age at time of release was about 35 years for both rapists and sexual assaulters. Child molesters and statutory rapists Some of the 9,691 sex offenders were men whose imprisonment offense was a sex offense against a child. Precisely how many is unknown. In this report, the 4,295 who could be identified are called "child molesters" The 4,295 identified child molesters included some (443 out of the 4,295) whose specific sex offense against a child was non-forcible intercourse. These 443 are called "statutory rapists. "There were more than 443 among the 4,295, but 443 were all that could be identified from the limited information obtained for the study. Both the 4,295 child molesters and the 443 statutory rapists were predominantly non-Hispanic white males. Nearly three-fourths of the child molesters (73.2%) were age 30 or older. Just over half the statutory rapists (54%) were 30 or older at the time they were released from prison. Among the released child molesters there were 3,333 white men (77. 6% of 4,295) and 889 black men (20.7%). The 443 statutory rapists included 324 white men (73.2% of 443) and 110 black men (24.8%). Sentence length and time served All sex offenders All 9,691 sex offenders selected to be in this study had a prison sentence greater than 1 year. The shortest terms were a day over 1 year; the longest were life sentences. The fact that sex offenders with a life sentence (18 offenders in the study) were among the 9,691 released in 1994 should not be surprising because only rarely do life sentences in the United States literally mean imprisonment for the remainder of a person's life. Most felons receiving a life sentence are eventually paroled (unpublished tabulation of data from the 1997 BJS Survey of Inmates in State Correctional Facilities). On average, a sex offender released from prison in 1994 had an 8-year term and served 31/2 years of that sentence (45%) before being released. Half of the released sex offenders had a sentence length of 6 years or less. Half had served no more than a third of their sentence before being released. When released, the majority (54.5%)had more than 3 years of their sentence remaining to be served. Rapists and sexual assaulters Rape always involves forcible intercourse, whereas sexual assault (as the term is used here)never does, although it can involve other types of forcible sexual assault. Because forcible intercourse is considered to be a more serious offense than other forms of forcible sexual assault, penalties for rape are generally more severe than those for sexual assault. Consistent with the more serious nature of rape -- * on average a released rapist had a longer sentence (just over 11 years) than a sexual assaulter (just under 7 years) * on average a rapist spent more time in confinement before being released (51/2 years)than a sexual assaulter (just under 3 years) * median sentence length was longer for rapists (half of the rapists had a sentence of 9 years or more, while half of the sexual assaulters had a sentence of 51/2 years or more) * 39.2% of the 3,115 rapists were in prison for over 5 years prior to release, while 12.5% of the 6,576 sexual assaulters served 61 months or more * rapists served 49% of their sentence before being released, compared to 43% for sexual assaulters. Depending on the length of their sentence and the amount of time they had served before being released, some of the released sex offenders would have been on parole (or some other type of conditional release) throughout the full 3 years they were tracked in this study. For example, when released, 63.3% of rapists had more than 3 years left to serve on their sentence. In their case, any new crimes they committed during this 3-year followup period were offenses committed while still on parole. By comparison, just over half of released sexual assaulters had more than 3 years left to serve. Child molesters and sexual assaulters On average, child molesters were released after serving nearly 3 years (33.7 months) of their nearly 7-year sentence (81.1 months). Statutory rapists were released after serving a little over 2 years of their approximately 4-year sentence. Upon release, almost half of the child molesters still had at least 3 years of their sentence remaining to be served, compared to 15% of statutory rapists. Prior criminal record All sex offenders Arrests and convictions for minor traffic offenses, public drunkenness, and other petty crimes are often not entered into official criminal records. Since official records formed the basis for this study's statistics on arrests and convictions, these statistics understate levels of contact with the justice system. Statistics shown throughout this report on arrests and convictions pertain mostly to arrests and convictions for felonies and serious misdemeanors. Statistics on prior arrests in this section of the report do not include the imprisonment offense for which the sex offender was in prison in 1994. At the time the 9,691 male sex offenders were arrested for the sex crime that resulted in their imprisonment -- * 78.5% (7,607 of the 9,691 men) had been arrested at least one earlier time * half had 3 or more prior arrests for some type of crime * 58.4% (5,660 men) had at least one prior criminal conviction * 13.9% (1,347 men) had a prior conviction for a violent sex offense * 4.6% (446 men) had been convicted for a sex crime against a child * nearly a quarter had served time in a State or Federal prison at least once before for some type of crime. All 9,691 were in prison in 1994 because they had been arrested and convicted for a sex offense. For 71.5% of the 9,691 men (6,929), that arrest was their first ever for a violent sex crime. In other words, these 6,929 men had no previous arrest for a sex offense. For the remaining 28.5% (2,762 men), that arrest was not their first sex offense arrest. Some had been arrested once before for a sex crime and some two or more times before. To illustrate, one of the 9,691 sex offenders in this study had his first arrest for a sex crime in 1966, when he was age 19; he was also arrested for sex crimes in the 1970's and 1980's, in three different States. The arrest for his imprisonment offense was in 1982. In the early part of 1983, 4 months after his arrest, he was convicted of sexual assault and began serving a 25-year prison term. Eleven years later, in 1994 at age 47, he was released. For 75% of the 9,691 sex offenders, their 1994 release represents their first release since being sentenced for their sex offense. The remaining 25% had previously served time under the same sentence, had been released, had violated one or more conditions of their parole and, consequently, were returned to prison to continue serving time still remaining on their sentence. Sex offenders compared to non-sex offenders A total of 262,420 non-sex offenders were released from State prisons in 1994 in the 15 States. Of the 262,420 non-sex offenders, 94% had at least 1 prior arrest and 82% had at least 1 prior conviction (not in a table). Overall, the 9,691 sex offenders had a shorter criminal history than the 262,420 non- sex offenders. Before the arrest that resulted in their prison sentence, sex offenders had been arrested 4.5 times, on average. This prior arrest record was about half that of non-sex offenders (8.9 prior arrests). In addition, among the 1994 prison releases, 23.7% of the sex offenders (2,297), compared to 44.3% of non-sex offenders (116,252), had served prior prison sentences. Sex offenders were more likely to have been arrested (28.5%) or convicted (13.9%) for a sexual offense than non-sex offenders (6.5% with a prior arrest for a sex crime; 0.2% with a prior conviction for a sex crime). The same is true for child molesting -- about 1 in 10 sex offenders had a prior arrest for a sex offense against a child, compared to about 1 in 100 non-sex offenders. Rapists and sexual assaulters For approximately 71% of the 3,115 rapists, the arrest for rape that resulted in their imprisonment was their first for a sex crime. The remaining 29% had one or more prior sex crime arrests. Likewise, for sexual assaulters, the sexual assault arrest that led to their imprisonment was the first arrest for a sex crime for 72% of the 6,576 sexual assaulters. The remaining 28% had been arrested at least once before for some type of sex crime. Child molesters and sexual assaulters The 4,295 child molesters had at least 1 arrest for child molesting (the arrest that led to their imprisonment). For 3,509 (81.7%) of them, that arrest was their first ever arrest for child molesting. For the other 786 men (18.3% of the 4,295), that was not their first. Some had one prior arrest for a sex offense against a child, some had two, and others had three or more. Among those with three or more priors was a man whose first arrest for child molesting was in 1966, when he was age 20. When released in 1994, he was serving an 11-year sentence for molesting a child under age 14. The prior criminal record of this serial pedophile spanned three decades, with arrests for child molesting in the 1970's, the 1980's, and the 1990's. Four measures of recidivism This section measures recidivism four ways: * percent rearrested for any type of crime * percent reconvicted for any type of crime * percent returned to prison with a new prison sentence for any type of crime * percent returned to prison with or without a new prison sentence. "Percent rearrested" is calculated by dividing "the number rearrested" by "the number released from prison in 1994." "Percent reconvicted" is obtained by dividing "the number reconvicted" by "the number released from prison in 1994." (It is not calculated by dividing "the number reconvicted" by "the number rearrested.") "Percent returned to prison with a new sentence" is calculated by dividing "the number returned to prison with a new sentence" by "the number released from prison in 1994.' (It is not calculated by dividing "the number returned to prison with a new sentence' by 'the number reconvicted.') Except where stated otherwise, all four recidivism measures -- * refer to the full 3-year period following the prisoner's release in 1994 * include both "in-State" and "out-of-State" recidivism. "In-State" recidivism refers to new offenses committed within the State that released the prisoner in 1994. "Out-of-State" recidivism is any new offenses in States other than the one that released him in 1994. Not all 4 of the recidivism measures are based on data from 15 States * "Percent rearrested' is based on 15 States * "Percent reconvicted" is based on 14 of the 15 States participating in the study * "Percent returned to prison with a new sentence" is based on 13 of the 15 States * "Percent returned to prison with or without a new sentence" is based on 9 of the 15. Three of the four recidivism measures were calculated from data on fewer than 15 States because the information needed to perform the calculations was not available (or not readily available) from each of the 15 participating States. Notes at the bottom of the tables alert readers to such missing data. Four measures All sex offenders The 9,691 sex offenders in this study were all released from prison in 1994. Within the first 3 years following their release -- * 43% (4,163 of the 9,691) were rearrested for at least 1 new crime * 24% (2,326 of the 9,691) were reconvicted for any type of crime * 11.2% (1,085 of the 9,691) were returned to prison with another sentence * 38.6% (3,741 of the 9,691) were returned to prison with or without a new sentence. For approximately three-fourths of the 4,163 men who were rearrested for some new crime, their most serious rearrest offense was a felony; for the remaining fourth, the most serious was a misdemeanor (not shown in table). Of the 4,163 men rearrested for some new offense, nearly 9 in 10 (87%) were still on parole when taken into custody (not shown in table). The 2,326 reconvicted for a new crime consisted of 1,672 (71.9%) whose most serious conviction offense was a felony, and 654 (28.1%) whose most serious offense was a misdemeanor (not shown in table). Of the 2,326 reconvicted for any new crime after their release, 1,085 were resentenced to prison, and the remaining 1,241 were placed on probation or ordered to pay a fine or sentenced to short-term confinement in a local jail. The 1,241 not resentenced to prison made up a little over half (54%) of the total 2,326 reconvicted. A major reason why over half were not resentenced to prison was that the new conviction offense for about 900 of the 2,326 newly convicted men (approximately 40%) was a misdemeanor rather than a felony, and State laws usually do not permit State prison sentences for misdemeanors. Altogether, 3,741 (38.6%) of the 9,691 released sex offenders were returned to prison either because of a new sentence or a technical violation. Of the 3,741, 2,656 (71%) were returned for a technical violation, such as failing a drug test, missing an appointment with the parole officer, or being arrested for another crime; and 1,085 were returned with a new prison sentence. The 2,656 consisted of 670 who were reconvicted but not resentenced to prison, plus 1,986 not reconvicted. As previously explained, a total of 1,241 released sex offenders were reconvicted but not resentenced to prison for their new crime. The 1,241 included 670 (described immediately above) who were returned to prison for a technical violation. The 670 were 54% of the 1,241, indicating that most of those who were reconvicted but not given a new prison sentence were, nevertheless, returned to prison. Sex offenders compared to non-sex offenders The 15 States in this study released 272,111 prisoners altogether in 1994. The 9,691 released sex offenders made up 3.6% of that total. The remaining 262,420 released prisoners were non-sex offenders. Of the 262,420 non-sex offenders, 68% (179,391 men and women out of the 262,420) were rearrested for a new crime within 3 years (not shown in table). The 43% overall rearrest rate of the 9,691 released sex offenders (4,163 out of 9,691) was low by comparison. Another difference was the rearrest charge. The rearrest offense was a felony for about 3 out of 4 (75%) of the 4,163 rearrested sex offenders (not shown in table). By comparison, about 84% of the 179,391 non-sex offenders were charged by police with a felony (not shown in table). Of the 4,163 sex offenders rearrested for a new crime, nearly 9 in 10 (87%) were on parole when taken into custody; of the 179,391 rearrested non-sex offenders, also about 9 in 10 (85%) were on parole (not shown in table). There was a difference in reconvictions. The reconviction rate for the 9,691 released sex offenders was 24.0%, compared to 47.8% for 262,420 non-sex offenders released in 1994 (not shown in table). The 2,326 sex offenders reconvicted for any new crime included 1,672 (71.9%) whose most serious conviction offense was a felony (not shown in table). Of the 262,420 non-sex offenders, 125,437 (47.8%) were reconvicted, which included 94,078 (75.0%)whose most serious reconviction offense was a felony (not shown in table). Rapists and sexual assaulters Within the first 3 years following release -- * 46.0% of the 3,115 rapists (1,432 men) and 41.5% of the 6,576 sexual assaulters (2,731 men) were rearrested for all types of crimes * 27.3% of the 3,115 rapists (850 men) were reconvicted, compared to 22.4% of the 6,576 sexual assaulters (1,473 men)for all types of crimes * 12.6% of the 3,115 rapists (392 men) and 10.5% of the 6,576 sexual assaulters (690 men) were resentenced to prison for their reconviction offense * 43.6% of the 3,115 rapists (1,358 men) and 36.1% of the 6,576 sexual assaulters (2,374 men) were returned to prison either because of a new sentence or because of a technical violation of their parole. For approximately three-fourths of the 1,432 rapists who were rearrested for a new crime, the crime was a felony; for the remainder, the most serious was a misdemeanor (not shown in table). As indicated earlier, 2,731 sexual assaulters were rearrested for a new offense after their release, and for about three-fourths, their most serious rearrest offense was a felony; for the remainder, the most serious crime was a misdemeanor (not shown in table). The 850 rapists reconvicted for any new crime included 617 (72.6%) whose most serious reconviction offense was a felony; the 1,473 reconvicted sexual assaulters included 1,052 (71.4%) who were reconvicted for a felony (not shown in table). Child molesters and statutory rapists Of the child molesters and statutory rapists released from prison in 1994 -- * 1,693 of the 4,295 child molesters (39.4%) and 221 of the 443 statutory rapists (49.9%) were rearrested for a new crime (not necessarily a new sex crime) * 876 of the 4,295 child molesters (20.4%) and 145 of the 443 statutory rapists (32.7%) were reconvicted for any type of crime * 9% of the 4,295 child molesters and 13% of the 443 statutory rapists were resentenced to prison for their new conviction offense * 38% of the 4,295 child molesters and 46% of the 443 statutory rapists were back in prison within 3 years as a result of either a new prison sentence or a technical violation of their parole. The most serious offense for three-fourths of the 1,693 child molesters who were rearrested was a felony, and a misdemeanor for the remainder (not shown in table). Following their release in 1994, 221 statutory rapists were rearrested for a new crime. The most serious offense that approximately three-fourths were charged with was a felony (not shown in table). The 876 child molesters reconvicted for any type of crime included 643 (73.4%) whose most serious reconviction offense was a felony; the 145 reconvicted statutory rapists included 97 (66.7%) whose most serious was a felony (not shown in table). Time to recidivism All sex offenders Within 6 months following their release, 16% of the 9,691 men were rearrested for a new crime (not necessarily another sex offense). Within 1 year, altogether 24.2% were rearrested. Within 2 years the cumulative total reached 35.5%. By theend of the 3-year followup period, 43% (4,163 of the 9,691) were rearrested for some type of crime. These statistics indicate that most recidivism within the first 3 years following release occurred in the first year (56%, since 2 4.2% / 43% = 56%). While the bulk of rearrests occurred in the first year, that period did not account for the bulk of reconvictions or reimprisonments. This is largely because a sizable number of those rearrested in the first year were not reconvicted and reimprisoned until sometime in the second year, due to the additional time needed to prosecute, convict, and sentence a criminal defendant. For example, by the end of the first year, 8.6% of the 9,691 released sex offenders were reconvicted, and by the end of the third year, a cumulative total of 24% were reconvicted, indicating that the first year accounted for a relatively small percentage of all the reconvictions in the 3 years (36%, since 8.6% / 24% = 36%). Rapists and sexual assaulters Forty-six percent of released rapists were rearrested within 3 years, and over half of those rearrests (56%) occurred in the first year (since 25.8% / 46.0% = 56%). Similarly, 41.5% of released sexual assaulters were rear-rested within the first 3 years following their 1994 release, and over half of those rearrests (56%)occurred in the first year (since 23.4% / 41.5% = 56%). Child molesters and statutory rapists Of the 4,295 released child molesters, 1,693 (39.4%) were rearrested during the 3-year followup period. The majority of those charged (approximately 982 of the 1,693, or 58%) were charged in the first 12 months. While 49.9% of released statutory rapists were rearrested within 3 years, nearly three- fifths of those rearrests occurred within the first year following release (29.8% / 49.9% = 60%). Rearrest for any type of crime Demographic characteristics All sex offenders Race Black men (56.1%) released in 1994 were more likely than white men (36.7%) to be rearrested for a new crime (not limited to just a new sex crime) within the first 3 years following their release. Hispanic origin Among released sex offenders, non-Hispanics (45.9%) were more likely than Hispanics (42.2%) to have a new arrest within the 3-year followup period. Age The younger the prisoner when released, the higher the rate of recidivism. For example, of all the sex offenders under age 25 at the time of discharge from prison, 59.8% were rearrested for some type of crime within 3 years, or more than double the 23.5% of those age 45 or older. Rapists and sexual assaulters Race Among releasees whose imprisonment offense was sexual assault, 57% of black men and 35.8% of white men were rearrested for all types of crimes. A higher rearrest rate for blacks was also found among released rapists. Hispanic origin Among released rapists, non-Hispanics (50.2%) were more likely than Hispanics (47.7%) to be rearrested within the 3-year followup period. The same was true among released prisoners whose imprisonment offense was sexual assault. Age For both rapists and sexual assaulters, younger releasees had higher rearrest rates than older releasees. Child molesters and statutory rapists Race The rearrest rate among released child molesters was 51.7% for black men and 36.2% for white men. Among statutory rapists, black men (61.5%) had a higher rearrest rate than white men (46.0%). Hispanic origin Among released prisoners whose imprisonment offense was statutory rape, Hispanics (56.9%) were more likely than non-Hispanics (48.8%) to be rearrested within the 3-year followup period. The opposite was true of child molesters, as Hispanics had a lower rearrest rate (37.1%) than non-Hispanics (41.9%). Age The younger the sex offender was when released, the higher was his likelihood of being rearrested. For example, the rearrest percent for statutory rapists younger than 25 was higher (70.0%) than the rearrest percent for statutory rapists ages 25 to 30 (56.4%). The same was true among child molesters. Time served before 1994 release All sex offenders Sex offenders who served the shortest amount of time in prison before being released (6 months or less) had a higher rearrest rate (45.7%) than those who served the longest (over 5 years, 39.9% rate). Similarly, prisoners who served 6 months or less had a higher rearrest rate (45.7%)than those who served 7 months to 1 year (42.1%). However, other comparisons did not indicate a connection between serving more time and lower recidivism. For example, among sex offenders who served 1 to 11/2 years in prison before being released, 38.9% were rearrested for all types of crimes, compared to 46.7% of sex offenders who served a bit longer -- 11/2 to 2 years. Similarly, released prisoners who served between 3 and 5 years in prison had a higher rate of rearrest (38.9%) than released prisoners who served 21/2 to 3 years (35.7%). Because of these mixed results, and others illustrated below, the data do not warrant any general conclusion about an association between the level of recidivism and the amount of time served. Rapists and sexual assaulters Among sexual assaulters who served no more than 6 months, 45.0% were rearrested for all types of crimes. Those who served a little longer -- from about 6 months to 1 year -- had a lower rearrest rate, 43.1%. Those released after serving even more time -- 1 to 11/2 years -- had an even lower rate, 39.2%. However, there are numerous instances where serving more time was not linked to lower recidivism. For example, rapists released after about 1 to 11/2 years in prison had a 37.6% rearrest rate, while those imprisoned a little longer -- from about 11/2 to 2 years -- had a higher rate, 51.1%. Child molesters and statutory rapists Among released statutory rapists and child molesters, the results continued to be mixed regarding an association between the rate of recidivism and the amount of time served. For example, child molesters released after serving about 2 to 21/2 years had a higher rate of rearrest for all types of crimes (39.4%) than those who served somewhat longer -- about 21/2 to 3 years (27.2%). However, the rearrest rate rose (31.5%) among molesters who served more time -- 3 to 5 years. Prior arrest for any type of crime All sex offenders For 2,084 sex offenders (21.5% of the 9,691 total), their only arrest prior to being released in 1994 was the arrest for their imprisonment offense (a sex offense). Among these 2,084 released sex offenders with just 1 prior arrest, 24.8% were rearrested for a new crime (not necessarily a new sex crime). For the remaining 7,607 (78.5% of 9,691), their prior record showed an arrest for the sex offense responsible for their current imprisonment plus at least 1 earlier arrest for some type of crime. Of these 7,607 prisoners, 47.9% were rearrested, or about double the rate of their counterparts with 1 prior arrest (24.8%). Rapists and sexual assaulters Of the 3,115 released rapists, 83.1% (2,589 rapists) had more than 1 arrest for some type of crime prior to their release from prison in 1994, and 16.9% (526 rapists) had just 1 prior arrest, the arrest for the sex crime that resulted in their being in prison in 1994. The multiple prior arrests for the 2,589 rapists included the arrest for their imprisonment offense plus at least 1 other arrest for any type of crime. The 2,589 with more than 1 prior arrest had a rearrest rate (49.6%) nearly double that of the 526 with just 1 prior (28.3%). Child molesters and statutory rapists Of the 4,295 child molesters, 76.8% (3,299 men)had more than 1 prior arrest. These 3,299 child molesters had a rearrest rate (44.3%) nearly double the 23.3% rate of the 996 molesters with just 1 prior arrest (996 is 23.2% of 4,295). The 357 statutory rapists with more than 1 prior arrest (357 is 80.6% of 443) had a rearrest rate (55.7%)more than double the 25.6% rate of the 86 statutory rapists with 1 prior arrest (86 is 19.4% of 443). Number of prior arrests for any type of crime Statistics on prior arrests in this section of the report do include the imprisonment offense of the released sex offender. All sex offenders The number of times a prisoner was arrested in the past was a relatively good predictor of whether that prisoner would continue his criminality after re- lease. Prisoners with just one prior arrest for any type of crime had a 24.8% rearrest rate for all types of crimes. With two priors, the percentage rearrested rose to 31.9%. With three, it increased to 36.9%. With four, it went up to 42.6%. With additional priors, there were further increases, ultimately reaching a rearrest rate of 67.0% for released prisoners with the longest criminal record (more than 15 prior arrests). Rapists and sexual assaulters Both rapists and sexual assaulters followed the pattern described immediately above: the more prior arrests they had, the more likely they were to have a new arrest for some type of crime after their release in 1994. Child molesters and statutory rapists Among released prisoners with the smallest number of prior arrests (1 prior arrest), 23.3% of child molesters and 25.6% of statutory rapists were rearrested for all types of crimes within 3 years. Rearrest rates generally rose with each increase in the number of prior arrests. Among released prisoners with the largest number of prior arrests (more than 15), 62.0% of child molesters and 76.2% of statutory rapists had at least 1 new arrest after being released in 1994. State where rearrested for any type of crime The State where the rearrest occurred was not always the State that released the prisoner. In some cases, the released sex offender left the State where he was imprisoned and was rearrested for a new crime in a different State. For example, a sex offender released from prison in California may have traveled to Nevada, where he was arrested for committing another crime. Sex offenders A total of 4,163 sex offenders were rearrested for some type of new crime after their 1994 release. Of the 4,163 arrests, 16.0% -- or 1 in 6 -- were outside the State where the prisoner was released. The rest (84.0%) were made in the State that released them. Sex offenders compared to non-sex offenders The 15 States in this study released 262,420 non-sex offenders in 1994, of whom 179,391 were rearrested for a new crime within 3 years (not shown in table). Of the 179,391 arrests for any type of crime, 11.2%, or 20,092 arrests, were arrests that occurred outside the State that released them. Rearrested sex offenders had a higher percentage: 1 in 6 of their rearrests for any type of crime were in a State other than the one that released them. Rapists and sexual assaulters Following their 1994 release, 1,432 rapists and 2,731 sexual assaulters were rearrested for any new crime. For 17.4% of the 1,432 rearrested rapists, and 15.2% of the 2,731 rearrested sexual assaulters, the place where the arrest occurred was in a different State than the one that released them. Child molesters and statutory rapists Out of the 4,295 child molesters, 1,693 were rearrested for any new crime after being released from prison in 1994. The 1,693 recidivists consisted of 84.8% whose new arrest was in the same State that released them in 1994, and 15.2% whose alleged violation occurred in a different State. About half of all statutory rapists were not rearrested for any type of crime after their release. Of the 221 who were, 16.6% were rearrested outside the State where they were released. Rearrest and reconviction for a new sex crime Rearrest and reconviction All sex offenders Based on official arrest records, 517 of the 9,691 released sex offenders (5.3%) were rearrested for a new sex crime within the first 3 years following their release. The new sex crimes for which these 517 men were arrested were forcible rapes and sexual assaults. For virtually all of the 517, the most serious sex crime for which they were rearrested was a felony. Their victims were children and adults. The study cannot say what percentage were children and what percentage were adults because arrest files did not record the victim's age. Of the total 9,691 released sex, 3.5% (339 of the 9,691) were reconvicted for a sex crime (a forcible rape or a sexual assault) within 3 years. Sex offenders compared to non-sex offenders The 15 States in this study released a total of 272,111 prisoners in 1994. The 9,691 released sex offenders made up less than 4% of that total. Of the remaining 262,420 non-sex offenders, 3,328 (1.3%) were rearrested for a new sex crime within 3 years (not shown in table). By comparison, the 5.3% rearrest rate for the 9,691 released sex offenders was 4 times higher. Assuming that the 517 sex offenders who were rearrested for another sex crime each victimized no more than one victim, the number of sex crimes they committed after their prison release totaled 517. Assuming that the 3,328 non-sex offenders rearrested for a sex crime after their release also victimized one victim each, the number of sex crimes they committed was 3,328. The combined total number of sex crimes is 3,845 (517 plus 3,328 = 3,845). Released sex offenders accounted for 13% and released non-sex offenders accounted for 87% of the 3,845 sex crimes committed by all the prisoners released in 1994 (517 / 3,845 = 13% and 3,328 / 3,845 = 87%). Rapists and sexual assaulters Of the 3,115 rapists, 5.0% (155 men) had a new arrest for a sex crime (either a sexual assault or another forcible rape) after being released. Of the 6,576 released sexual assaulters, 5.5% (362 men) were rearrested for a new sex crime (either a forcible rape or another sexual assault). A total of 100 released rapists were reconvicted for a sex crime. The 100 men were 3.2% of the 3,115 rapists released in 1994. Among the 6,576 released sexual assaulters, 3.7% (243 men) were reconvicted for a sex crime. Child molesters and statutory rapists After their release, 5.1% (221 men) of the child molesters and 5.0% (22 men) of the statutory rapists were rearrested for a new sex crime. Not all of the new sex crimes were against children. The new sex crimes were forcible rapes and various types of sexual assaults. Following their release, 3.5% (150 men) of the 4,295 released child molesters were convicted for a new sex crime against a child or an adult. The sex crime reconviction rate for the 443 statutory rapists was 3.6% (16 reconvicted men). Time to rearrest All sex offenders Within 6 months following their release, 1.4% of the 9,691 men were rearrested for a new sex crime. Within 1 year the cumulative total grew to 2.1% rearrested. By the end of the 3-year followup period, altogether 5.3% had been rearrested for another sex crime. The first year was the period when 40% of the new sex crimes were committed (since 2.1% / 5.3% = 40%). Rapists and sexual assaulters The first year following release accounted for 40% of the new sex crimes committed by both released rapists (since 2.0% / 5.0% = 40%) and released sexual assaulters (since 2.2% / 5.5% = 40%). Child molesters and statutory rapists For child molesters and statutory rapists, the first year following their release was the period when the largest number of recidivists were rearrested. Similar to rapists and sexual assaulters, about 40% of the arrests for new sex crimes committed by child molesters and statutory rapists occurred during the first year. Demographic characteristics All sex offenders Race Among sex offenders released from prison in 1994, black men (5.6%) and white men (5.3%)were about equally likely to be rearrested for another sex crime. Hispanic origin Among released sex offenders, non-Hispanics were more likely to be rearrested for a new sex offense (6.4%) than Hispanics (4.1%). One reason for the lower rearrest rate for Hispanics may be that some were deported immediately following their release. Age Recidivism studies typically find that, the older the prisoner when released, the lower the rate of recidivism. Results reported here on released sex offenders did not follow the familiar pattern. While the lowest rate of rearrest for a sex crime (3.3%) did belong to the oldest sex offenders (those age 45 or older), other comparisons between older and younger prisoners did not consistently show older prisoners' having the lower rearrest rate. Time served before 1994 release All sex offenders The study compared recidivism rates among prisoners who served different lengths of time before being released from prison in 1994. No clear association was found between how long they were in prison and their recidivism rate. For example, those sex offenders who served from 7 to 12 months were rearrested for a new sex crime at a higher rate (5.2%) than those who served slightly less time (3.8%), which seemed to suggest that serving more time raised the recidivism rate. But other comparisons suggested the opposite. Compared to men who were confined for 7 to 12 months (5.2% rearrest rate), those who served more time (13 to 18 months) were less likely to be rearrested for any new sex crime (4.1%). Prior arrest for any type of crime All sex offenders Of the 9,691 released sex offenders, 21.5% (2,084 of the 9,691) had only 1 arrest in their criminal record up to the time they were released. That one arrest was the arrest for the sex crime that resulted in a prison term. The remaining 78.5% (7,607 men) had the arrest for their imprisonment offense in their record, and they also had at least 1 earlier arrest for some type of crime. For example, some had an earlier arrest for theft or a drug offense. Most of them did not have an earlier arrest for a sex crime. Compared to the 2,084 sex offenders with the 1 arrest in their criminal record, the 7,607 with a longer prior arrest record were more likely to be rearrested for another sex crime (5.9% compared to 3.3%). Rapists and sexual assaulters Of the 3,115 released rapists, the majority (83.1% of the 3,115, or 2,589 men) had more than 1 arrest (for any type of crime) prior to release from prison in 1994. Of these 2,589 released rapists, 5.4% (140) had a new arrest for a sex crime. The rate was lower (3.0%) for the 526 released rapists with no prior arrest. Results for sexual assaulters followed the same pattern: the 5,017 sexual assaulters with more than 1 prior arrest (76.3% of 6,576 is 5,017) were more likely to be rearrested for a new sex crime (6.2%) than the 1,559 with just the 1 prior arrest (23.7% of 6,576 is 1,559). Child molesters and statutory rapists Released child molesters with more than one prior arrest were more likely than those with only one arrest in their criminal record to be rearrested for a new sex crime (5.7% compared to 3.2%). The same was true of statutory rapists (5.3% compared to 3.5%). Number of prior arrests for any type of crime All sex offenders The more arrests (for any type of crime) the sex offender had in his criminal record, the more likely he was to be rearrested for another sex crime after his release from prison. Sex offenders with one prior arrest (the arrest for the sex crime for which they had been imprisoned) had the lowest rate, about 3%; those with 2 or 3 prior arrests for some type of crime, 4%; 4 to 6 prior arrests, 6%; 7 to 10 prior arrests, 7%; and 11 to 15 prior arrests, 8%. Prior arrest for a sex crime All sex offenders Prior to their release in 1994, 2,762 of the sex offenders (28.5% of the total 9,691) had 2 or more arrests for a sex offense in their criminal record: the arrest for the sex offense that resulted in their imprisonment, plus at least 1 earlier arrest for a sex crime. For the remaining 6,929 (71.5% of the total 9,691), their only prior arrest for a sex crime was the arrest that brought them into prison. (Any other prior arrests the 6,929 may have had were for non-sex crimes.) Following their release, the 2,762 with more than 1 sex crime in their criminal background were about twice as likely to be rearrested for another sex crime (8.3%) as the 6,929 with a single prior arrest (4.2%). Rapists and sexual assaulters Rapists (4.0%) and sexual assaulters (4.2%) with one prior arrest for a sex crime were less likely to be rearrested for another sex crime than rapists (7.4%) and sexual assaulters (8.7%) who had been arrested two or more times for a sex crime prior to release from prison in 1994. Child molesters and statutory rapists By definition, all 4,295 child molesters had been arrested for a sex offense at least once prior to their release in 1994 -- the sex offense that landed them in prison. For 3,049 of them (71% of 4,295), that arrest was their only prior arrest for a sex offense. The remaining 1,246 child molesters (29% of 4,295) had at least 2 prior arrests for a sex crime: the arrest for their imprisonment offense plus at least 1 other prior arrest for a sex offense (not necessarily one against a child). Of the 1,246 child molesters with multiple sex crimes in their past, 8.4% (105 of the 1,246) were rearrested for another sex crime (not necessarily another sex crime against a child), or more than double the 3.8% rate for the 3,049 released child molesters with just 1 prior arrest for a sex crime. Similar results were found for released statutory rapists. Those with a more extensive record of prior arrests for sex crimes were more likely to be rearrested for another sex crime (8.8%) than those with just one past arrest (2.6%). State where rearrested for a sex crime When sex offenders were arrested for new sex crimes after their release, the new arrest typically occurred in the same State that released them. Those arrests are referred to as "in-State" arrests. When released sex offenders left the State where they were incarcerated and were charged by police with new sex crimes, those arrests are referred to as "out-of-State" arrests. All sex offenders Of the 9,691 released sex offenders, 517 were rearrested for a new sex crime within 3 years. Most of those sex crime arrests (85.2% of the 517, or 440 men) were in the same State that released them. Seventy-seven of them (14.8% of the 517) were arrests in a different State. Sex offenders compared to non-sex offenders The 15 States in this study released 262,420 non-sex offenders in 1994, of whom 3,328 were rearrested for a new sex crime within 3 years (not shown in table). Of the 3,328 non-sex offenders arrested for a new sex crime, an estimated 10% were men rearrested outside the State that released them. The 15% figure for released sex offenders was high by comparison. Rapists and sexual assaulters A total of 155 released rapists and 362 released sexual assaulters were rearrested for a new sex crime within the 3-year followup period. In-State arrests for new sex crimes accounted for 85% of the rearrested rapists and 85% of the rearrested sexual assaulters. Out-of-State arrests accounted for the rest. Child molesters and statutory rapists A total of 221 child molesters were rearrested for a new sex crime (not necessarily against a child) after their release. Among the 221 were 191 (86.6%) whose new sex crime arrest was in the same State that released them in 1994. For the remaining 13.4%, the arrest was elsewhere. Of all statutory rapists, 5% (22) were rearrested for a new sex crime after their release. Of these 22, none had the new arrest outside the State that released them. Rearrest for a sex crime against a child Undercounts of sex crimes against children This section documents percentages of men who were arrested for a sex crime against a child after their release from prison in 1994. To some unknown extent, these recidivism rates undercount actual rearrest rates. That is because the arrest records that the study used to document sex crime arrests did not always contain enough information to identify those sex crime arrests in which the victim of the crime was a child. Some sense of the potential size of the undercount can be gained by comparing rearrests for any sex crime and rearrests for any sex crime against a child. Rates of rearrest for a sex crime are from 2 to 31/2 percentage points higher than rates of rearrest for a sex crime against a child, suggesting that rates of rearrest for a sex crime against a child could be, at most, a few percentage points below actual rates. No data on precise ages of molested children This section also documents the ages of the children that the men were alleged to have molested after their release from prison. Sex crime statutes contained in the arrest records of the released prisoners were used to obtain ages. The first step was to identify those sex crime statutes that were applicable just to children. Among those that were, some were found to apply just to children whose age fell within a certain range (for example, under 12, or 13 to 15, or 16 to 17). Those statutes applicable to children within specified age ranges became the source of information on the approximate ages of the allegedly molested children. Information on precise ages could not be determined because statutes applicable just to children of a specific age (for example, just to 12-year-olds, or just to age 15-year-olds) do not exist. Rearrest All sex offenders Following their release in 1994, 209 of the total 9,691 released sex offenders (2.2%) were rearrested for a sex offense against a child. For virtually all 209, the rearrest offense was a felony. For the reason given earlier, the 2.2% figure undercounts the percentage rearrested for a sex offense against a child. It seems unlikely that the correct figure could be as high as 5.3%, which is the percentage rearrested for a sex crime against a person of any age. The only way it could be that high is if none of the sex crime arrests after release were crimes in which the victim was an adult, an unlikely possibility. The more likely possibility is that the 2.2% figure undercounts the rate by a maximum of 1 or 2 percentage points. An estimated 76% of the children allegedly molested by the 209 men after their prison release were age 13 or younger, 12% were 14- or 15-years-old, and the remaining 12% were 16- or 17-years-old. Sex offenders compared to non-sex offenders Prisons in the 15 States in the study released 272,111 prisoners altogether in 1994, 9,691 of whom were the sex offenders in this report. As previously stated, 2.2% of the 9,691 sex offenders were rearrested for a child sex crime after their release. That rate is high compared to the rate for the remaining 262,420 non- sex offenders. Of the 262,420 non-sex offenders, less than half of 1 percent (1,042 of the 262,420) were rearrested for a sex offense against a child within the 3-year followup period (not shown in table). Since each of the 1,042 was charged at arrest with molesting at least 1 child, the total number they allegedly molested was conservatively estimated at 1,042. Of the conservatively estimated 1,042 children, 65% were age 13 or younger, 11% were 14- or 15-years-old, and 24% were 16- or 17-years-old (not shown in table). (These percentages were based on the 554 cases out of the 1,042 in which the approximate age of the child could be determined.) Assuming that the 209 sex offenders who were rearrested for a sex crime against a child each victimized no more than one child, the number of sex crimes they committed against children after their prison release totaled 209. Assuming that the 1,042 non-sex offenders rearrested for a sex crime against a child after their release also victimized only one child, the number of sex crimes against a child that they committed was 1,042. The combined total number of sex crimes is 1,251 (209 plus 1,042 = 1,251). Released sex offenders accounted for 17% and released non-sex offenders accounted for 83% of the 1,251 sex crimes against children committed by all the prisoners released in 1994 (209 / 1,251 = 17% and 1,042 / 1,251 = 83%). Rapists and sexual assaulters Following their 1994 release, 1.4% of the 3,115 rapists (44 men) and 2.5% of the 6,576 sexual assaulters (165 men) were rearrested for molesting a child. Child molesters and statutory rapists Within 3 years following their release from prison in 1994, 141 (3.3%) of the released 4,295 child molesters and 11 (2.5%) of the 443 released statutory rapists were rearrested for molesting another child. For the reasons outlined earlier, these percentages undercount actual rearrest rates by a few percentage points at most. Each of the 141 released molesters rearrested for repeating their crime represented at least 1 child victim. Of the conservatively estimated 141 children allegedly molested by released child molesters, 79% were age 13 or younger, 9% were 14 or 15 years of age, and 12% were ages 16 or 17. Prior arrest for a sex crime against a child All sex offenders After their 1994 release from prison, sex offenders with a prior arrest for child molesting were more likely to be arrested for child molesting (6.4%) than those who had no arrest record for sex with a child (1.7%). Rapists and sexual assaulters After being released in 1994, 4.0% of rapists with a prior arrest record for child molesting and 1.3% of those without were arrested for child molesting. The same pattern -- having a history of alleged child molesting was associated with a greater likelihood of arrest for child molesting -- was found for sexual assaulters. Those with a prior arrest had a 6.9% rate; those without, 1.9%. Child molesters and statutory rapists The 4,295 released child molesters fell into 2 categories: 1) 3,509 (81.7% of the 4,295) whose criminal record prior to their 1994 release contained no more than 1 arrest for a sex offense against a child (this was the offense for which they were imprisoned); and 2) 786 (18.3%) whose record showed the arrest for their imprisonment offense plus at least one earlier arrest for a sex offense against a child. After release, 7.3% of the 786 and 2.4% of the 3,509 were rearrested for molesting another child, indicating that child molesters with multiple arrests for child molesting in their record posed a greater risk of repeating their crime than their counterparts. Similarly, the 443 statutory rapists consisted of - * 356 (80.4%) whose first arrest for a sex offense against a child was the arrest that resulted in their current imprisonment * 87 (19.6%) with more than 1 prior arrest for a sex offense against a child. The 87 were more likely to be rearrested for child molesting (6.9%) than the 356 (1.4%). Molester's and child's ages at time of imprisonment offense Child molesters The released child molesters were all men who were arrested, convicted, and sentenced to prison for a sex crime against a child. At the time they committed their imprisonment offense, most (62.9%) were age 30 and older, and most (60.3%) molested a child who was age 13 or younger. Some of the victims were below age 7. Nearly half of the men (48.4%) were 20 years or more older than the child they were imprisoned for molesting. Among the men who were in prison for molesting a child age 13 or younger and who were released in 1994 for that crime, 2.8% were subsequently arrested for molesting another child. Of those whose imprisonment offense was against a 14- or 15-year-old, 3.7% had a new arrest for child molesting after their release. Of the men who were in prison for molesting a 16- or 17-year-old, 1.2% were arrested by police for molesting another child after leaving prison in 1994. Among the men who were 20 years or more older than the child they were imprisoned for molesting, 2.5% were rearrested for another sex offense against a child within the first 3 years following their release. That is a lower rate than the 3.2% rate for men who were 10 to 19 years older than the child victim in their imprisonment offense, and compared to the 3.6% for those 5 to 9 years older than the victim in their imprisonment offense. State where rearrested for a sex crime against a child When sex offenders were arrested for new sex crimes against children after their release, the new arrest typically occurred in the same State that released them. Those arrests are referred to as "in-State" arrests. When arrests occurred in a different State, they are referred to as "out-of-State." All sex offenders Of the 9,691 sex offenders, 209 were rearrested for child molesting after their release from prison in 1994. In 180 cases (86.3%), the alleged crime took place in the State that released him. In the 29 others (13.7%), it occurred elsewhere. Sex offenders compared to non-sex offenders The 15 States in this study released 262,420 non-sex offenders in 1994, of whom 1,042 were rearrested for a sex crime against a child (not shown in table). Of the 1,042 arrests, 11% were out-of-State rearrests. The comparable figure for released sex offenders was higher: 14%. Rapists and sexual assaulters Forty-four released rapists and 165 released sexual assaulters were rearrested for a sex crime against a child within 3 years. Out-of-State arrests for child molesting accounted for 13.5% of the 44 rearrested rapists and 13.7% of the 165 rearrested sexual assaulters. Child molesters and statutory rapists Police arrested 141 of the 4,295 released child molesters for repeating their crime. For 126 of them (89.2%), the new arrest for child molesting was in the same State that released them. For 15 (10.8%), the new charges for child molesting were filed in a different State. Of the 443 statutory rapists released from prison in 1994, 11 were rearrested for child molesting. All 11 of the arrests were in the same State that released the men. Rearrest for other types of crime All sex offenders Of the 9,691 male sex offenders released from prison in 1994 -- * 43% (4,163 men) were rearrested for a crime of any kind * 5.3% (517 men) were rearrested for a sex offense * 17.1% (1,658 men) were rearrested for a violent crime * 13.3% (1,285 men) were rearrested for a property crime of some kind. Of the 9,691 released men, 168 (1.7%) were rearrested for rape and 396 (4.1%) were rearrested for sexual assault. The 168 rearrested for rape plus the 396 rearrested for sexual assault totals 564, which is 47 greater than the total 517 who were rearrested for a sex crime. The reason is that 47 men were rearrested for both rape and sexual assault. The category of violent crime for which a prisoner was most likely to be rearrested was assault (8.8%, or 848 of the 9,691); the category least likely was homicide (0.5%, or 45 of the 9,691 men). Just over 1 in 5 sex offenders (2,045 out of 9,691)were rearrested for a public-order offense, such as a parole violation or traffic offense. Rapists and sexual assaulters Among the 3,115 released rapists -- * 46% (1,432) were rearrested for a crime of any kind * 18.7% (582) were rearrested for a violent crime * 0.7% (22) were rearrested for homicide * 14.7% (459) were rearrested for a property offense. A relatively small percentage of rapists (2.5%, or 78 of the 3,115) were charged with repeating the crime for which they were imprisoned. Among the 6,576 released sexual assaulters -- * 41.5% (2,731) were rearrested for a crime of any kind * 16.4% (1,076) were rearrested for a violent crime * 0.3% (23) were rearrested for killing someone * 12.6% (826) were rearrested for a property offense. Nearly 1 in 20 released sexual assaulters (4.7%, or 308 of the 6,576) were charged with committing the same type of crime for which had just served time in prison. Child molesters and statutory rapists Of the 4,295 child molesters released from prison in 1994 -- * 39.4% (1,693) were rearrested for a crime of any kind * 0.4% (17) were rearrested for intentionally or negligently killing someone. Child molesters were less likely to be rearrested for a property crime (10.6%, 456 of 4,295) than a violent crime (14.1%, 607 of 4,295). Of the 443 statutory rapists released in 1994 -- * 49.9% (221) were rearrested for some new crime * 0.7% (3) were rearrested for homicide * 22.6% (100) were rearrested for a property crime * 21.2% (94) were rearrested for a violent crime. ------------------------------------------------- Victims of sex crimes Survey of State inmates The 9,691 prisoners in this study were all men sentenced to prison for sex crimes. Characteristics of the victims of these sex crimes were largely unavailable for the study. For information on imprisoned sex offenders and their victims, data were drawn from a survey covering the approximately 73,000 male sex offenders in State prisons nationwide in 1997. Of the 73,000 victims of their sex crimes -- * about 90% were female * nearly 75% were white * 89% were non-Hispanic * 36% were below age 13 * altogether, 70% were under age 18. Child victims of sex crimes were more likely than adult victims to be male (11% versus 3%). Whites made up 76% of child victims and 66% of adult victims. The biggest difference between child victims and adult victims was their relationship to the man who committed the sex crime: Among cases where the victim was under 18, the boy or girl was the prisoner's own child (16%), stepchild (16%), sibling or stepsibling (2%), or other relative (13%) in nearly half of all child victim cases (46%). Among cases where the victim was an adult, the victim was a relative less often (11%). Among inmates who were in prison for a sex crime against a child, the child was the prisoner's own child or step- child in a third of the cases. Seven percent of the inmates reported their child victims to have been strangers. Among adult victims, 34% were strangers to their attacker. --------------------------------------------- Methodology 3-year followup period For analytic purposes, "3 years" was defined as 1,096 days from the day of release from prison. Any rearrest, reconviction, or reimprisonment occurring after 1,096 days from the 1994 release was not included. A conviction after 1,096 days was not counted even if it resulted from an arrest within the period. Separating sex offenders into four types The report gives statistics for four types of sex offenders. Separating sex offenders into the four types was done using information -- in particular, the statute number for the imprisonment offense, the literal version of the statute, a numeric FBI code (called the "NCIC" code, short for "National Crime Information Center") indicating what the imprisonment offense was, and miscellaneous other information -- available in the prison records on the 9,691 men. However, the prison records obtained for the study did not always contain all four pieces of information on the imprisonment offense. Moreover, the available offense information was not always detailed enough to reliably distinguish different types of sex offenders. The process of sorting sex offenders into different types involved first creating the study's definitions of the four types, and then determining which State statute numbers, which literal versions of those statutes, and which NCIC codes conformed to the definitions. Each inmate was next classified into one of the types (or possibly into more than one type, since the four are not mutually exclusive) depending on whether the imprisonment offense information available on him fit the study's definition. An obstacle to classifying sex offenders into types was that the labels "rape," "sexual assault," "child molestation," "statutory rape" were not widely used in State statutes, and when they were used they did not always conform to the study's definitions of them. In deciding which type of sex offender to classify the prisoner as, importance was attached not to the label the law gave to his conviction offense, but to how well the law's definition of the offense fit the study's definition of the type. Sex offenders compared to non-sex offenders In 1994, prisons in 15 States released 272,111 prisoners, representing two-thirds of all prisoners released in the United States that year. Among the 272,111 were 262,420 released prisoners whose imprisonment offense was not a sex offense. Non-sex offenders include inmates, both male and female, who were in prison for violent crimes (such as murder or robbery), property crimes (such as burglary or motor vehicle theft), drug crimes, and public order offenses. Like the 9,691 male sex offenders examined in this report, all non-sex offenders were serving prison terms of one year or more in State prison when they were released in 1994. At various places, this report compares 9,691 released male sex offenders to 262,420 released non-sex offenders. While labeled "non-sex offenders," the 262,420 actually includes a small number- 87- who are sex offenders. The 87 are all the female sex offenders released from prisons in the 15 States in 1994. Ages of molested and allegedly molested children Information on the ages of molested children was needed for two calculations: 1) age of the child the released sex offender was sent to prison for molesting, and 2) age of the child allegedly molested by the released sex offender during the 3-year follow-up period. The most frequent source of both was a sex statute: either the sex statute the offender was imprisoned for violating, or the statute the released prisoner was charged with violating when he was rearrested for a sex crime. The former was obtained from the prison records assembled for the study; the latter, from the assembled arrest records. None of the sex statutes was found to apply to a victim of a specific age; for example, just to 12-year-olds. But some were found to apply just to children in a certain age range; for example, under 12, or 13 to 15, or 16 to 17. While specific ages of children could not be obtained from statutes, the availability of information on age ranges at least made it possible to obtain approximate ages. The rule that was adopted was to record the victim's (or alleged victim's) age as the upper limit of a statute's age range. To illustrate, a statute might indicate that the complainant/victim be "at least 13 but less than 16 years of age." In that case, the age of the child was recorded as 15, since the statute indicated the upper limit of the age range as any age "less than 16." As another example, if a statute indicated the complainant/victim be "under 12 years of age," the child's age was recorded as 11, as the phrasing of the age range did not include 12-year-olds, only those "under 12." Because the victim (or alleged victim) was always assigned the age of the oldest person in the age range, the study made the victims (or alleged victims) appear older than they actually were. How missing data were handled in the report In many instances, the data needed to calculate a statistic were not available for all 9,691 released sex offenders. For example, the 9,691 were released in 15 States, but data needed to determine the number reconvicted were only available for the 9,085 released in 14 of the 15. Of the 9,085, 2,180 (24%) were reconvicted. When data were missing, the statistic was computed on those cases in which the data were available, but treated both in the tables and in the text as though it were based on the total population. For example, "24%" is the statistic that appears in all tables and text that give the percent reconvicted; and since 24% of 9,691 is 2,326, the text says that “2,326 of the 9,691 were reconvicted," despite the fact that the "24%" was actually obtained by dividing 2,180 by 9,085. The text could have been written to say “2,180 of the 9,085 were reconvicted," but that wasn't done because introducing a new denominator (9,085) into the text would have created confusion for the reader. Missing data on out-of-State rearrests Because of missing information, the study was unable to determine how many inmates released from New York prisons were rearrested outside of New York. The study was able to document how many prisoners released in the other 14 States were rearrested outside the State that released them. Because of incomplete New York data, the report's recidivism rates are somewhat deflated. Missing data on rearrest for a sex crime According to arrest records compiled in the study, 4,163 of the 9,691 released sex offenders were rearrested for a new crime of some kind. It was not always possible to determine from these records whether the new crime was a sex crime. For 202 rearrested prisoners, the arrest record did not identify the type of crime. For the rest the record did identify the type but the offense label was not always specific enough to distinguish sex crimes from other crimes. For example, if the label said "contributing to the delinquency of a minor," "indeceny," "morals offense," "family offense," or "child abuse," the offense was coded as a non-sex crime even though, in some unknown number of cases, it was actually a sex crime. According to arrest records, 5.3% of the 9,691 (517 out of 9,691) released sex offenders were rearrested for another sex crime. For the two reasons described immediately above, 5.3% was probably an undercount of how many were rearrested for a sex crime. How much of an undercount could not be firmly determined from the data assembled for the study. However, a conservative measure of the size of the undercount was obtained from the data. The study database included 121 rearrested sex offenders whose arrest record did not indicate they were rearrested for a sex crime (the rearrest was either for a non-sex crime or for an unknown type of crime) but whose court record did indicate they were charged with a sex crime. When the study calculated the percentage rearrested for a sex crime, the 121 were not included among the 517 with a rearrest for a sex crime. Had the 121 been included in the calculation of the rearrest rate, the total number rearrested for a sex crime would have been 638 rather than 517, and the percentage rearrested for a sex crime would have been 6.6% rather than 5.3%. This suggests an undercount of about 1 percentage point. Texas prisoners classified as "other type of release" Texas released 692 male sex offenders in 1994, of which 129 were classified as release category “17", defined as "other type of release." Numerous data quality checks were run on the 129 and the 64 of them who were rearrested. The rearrest rate for the 129 was about average for Texas releases. But numerous anomalies were found for the 64 who were rearrested: 1. The rearrest offense for the 64 was always missing from their arrest record 2. The date of rearrest for the 64 was always the same as their release date 3. Virtually all 64 were reconvicted for a sex crime 4. The sentence length imposed for their new sex crime was identical to the sentence they were serving when released in 1994. Because of these anomalies, the 129 were excluded from the calculation of "percent reconvicted for a sex crime." Counting rules In this report, rearrest was measured by counting the number of different persons who were rearrested at least once. A released prisoner who was rearrested several times or had multiple rearrest charges filed against him was counted as only one rearrested person. The same counting rule applied to reconviction and the other recidivism measures. If a released prisoner was rearrested several times, his earliest rearrest was used to calculate his time-to-rearrest. The same counting rule applied to reconviction and recidivism defined as a new prison sentence. If a released prisoner had both in-State and out-of-State rearrests, he was counted as having an out-of-State rearrest regardless of whether the out-of-State rearrest was his earliest rearrest. The same rule applied in cases where the released prisoner had both felony and misdemeanor rearrests, or both sex crime and non-sex crime rearrests. The person was counted as having a felony rearrest or a sex crime rearrest regardless of temporal sequence. The aim of these rules was to count people, not events. The only tables in the report that do not follow the rule are tables 41 and 42. First release All 15 States had first releases, but they could not be identified in 1 State (Ohio). They could be identified in Michigan, but Michigan data on sentence length did not fit the study's definition. Since sentence length was critical to several statistics calculated from data on first releases (for example, percent of sentence served), Michigan was excluded from all tables based on first releases. Analysis of statutory rape laws The publication's analysis of statutory rape laws in the United States benefitted greatly from the report "Sexual Relationships Between Adult Males and Young Teen Girls: Exploring the Legal and Social Responses," by Sharon G. Elstein and Noy Davis, American Bar Association, Center on Children and the Law, October 1997. Sampling error In 1994 State prisons in 15 States released 302,309 prisoners altogether. A total of 38,624 were sampled for a recidivism study. Results of that study and information regarding sampling and other methodological details are available in the BJS publication Recidivism of Prisoners Released in 1994, NCJ 193427, June 2002. The 302,309 total released consisted of 10,546 released sex offenders plus 291,763 released non-sex offenders. The 38,624 sample consisted of 10,546 released sex offenders plus 28,078 released non-sex offenders. The number of sex offenders in the sample was the same as the number in the 302,309 total because all sex offenders released in 1994 in the 15 States were selected for the study, not a sample of them. Because no sampling was used to select sex offenders, numbers and percentages in this report for sex offenders were not subject to sampling error. However, comparisons in the report between sex offenders and non-sex offenders were subject to sampling error because sampling was used to select non-sex offenders. Where sex offenders were compared to all non-sex offenders released in 1994, sampling error was taken into account. All differences discussed were statistically significant at the .05 level. Not all 10,546 sex offenders in the sample were used in the report. To be in the report, the sex offender had to be male and meet all 4 of the following criteria: 1. A RAP sheet on the prisoner was found in the State criminal history repository. 2. The released prisoner was alive throughout the entire 3-year followup period. (This requirement resulted in 21 sex offenders' being excluded.) 3. The prisoner's sentence was greater than 1 year (missing sentences were treated as greater than 1 year). 4. The State department of corrections that released the prisoner in 1994 did not designate him as any of the following release types: release to custody/detainer/warrant, absent without leave, escape, transfer, administrative release, or release on appeal. A total of 9,691 released male sex offenders met the selection criteria. The number of them released in each State is shown in the appendix table. Other methodological details To help the reader understand the percentages provided in the report, both the numerator and denominator were often given. In most cases, the reader could then reproduce the percentages. For example, the report indicates 38.6% (3,741) of the 9,691 sex offenders were returned to prison. Using the 3,741 and the 9,691, the reader could exactly reproduce the results. However, the reader should be aware that in a few places, the calculated percentages will differ slightly from the percentages found in the report. This is due to rounding. For example, 43.0%, or 4,163, of the 9,691 sex offenders were rearrested; however, 4,163 / 9,691 is 42.96%, which was rounded to 43.0%. Offense definitions and other methodological details are available in the BJS publication Recidivism of Prisoners Released in 1994, NCJ 193427, June 2002. End of file 10/22/03 ih