Probation and Parole Violators in State Prison, 1991: Survey of State Prison Inmates, 1991. U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Special Report August 1995, NCJ-149076 PROBATION AND PAROLE VIOLATORS IN STATE PRISON, 1991: SURVEY OF STATE PRISON INMATES, 1991 (Note: This file does not contain data tables. The full text with tables is available from the Bureau of Justice Statistics Clearinghouse, 800-732-3277. Use title and NCJ number to order.) By Robyn L. Cohen BJS Statistician In 1991, 45% of State prisoners were persons who, at the time they committed their offense, were under conditional supervision in the community--either on probation or on parole. Based on the offense that brought them to prison, the 162,000 probation violators committed at least 6,400 murders, 7,400 rapes, 10,400 assaults, and 17,000 robberies, while under supervision in the community an average of 17 months. Based on the offense that brought parolees back to prison, these 156,000 offenders committed at least 6,800 murders, 5,500 rapes, 8,800 assaults, and 22,500 robberies, while under supervision in the community an average of 13 months. These are some of the results of a national survey in which personal interviews were conducted with 13,986 inmates in 277 State correctional facilities nationwide. Data were collected on individual characteristics of State prison inmates: their status at time of admission, past and current offenses, their victims, their criminal histories, and their prior drug and alcohol use. Highlights * Thirty-five percent of State prison inmates in 1991 were convicted of a new offense that they had committed while they were on probation or parole from a previous sentence; 10% of the inmates had been returned to prison for technically violating the conditions of their probation or parole. * Probation and parole violators comprised 30% of all the offenders in State prison for a violent crime. Probation and parole violators comprised 56% of inmates incarcerated for property offenses, 41% for drug offenses, and 85% of those in prison for public-order offenses. * Collectively, probation and parole violators committed 90,639 violent crimes while being supervised in the community (based on reported convictions). * More than 1 in 4 probation and parole violators were in prison for conviction for a violent crime. * Fifty-five percent of probationers and parolees reported that, in the month before their current offense, they were using drugs. An estimated 41% were using drugs daily. While drug testing of probationers and parolees is widespread, 2% of the drug-using probation and parole violators reported being revoked for failing a drug test. * Felony probationers and parolees are not permitted to possess a firearm while under supervision. Yet 21% of probation and parole violators imprisoned for a new offense reported possessing a firearm while under supervision. * Of probation and parole violators who owned or possessed a handgun in the month before arrest, almost 3 out of every 4 were armed when they committed their current offense. * Between 1975 and 1991, the number of parole and other conditional release violators entering State prisons increased from 18,000 to 142,000--twice the rate of growth of offenders newly committed from courts. * * * * * * * Criminal justice status defined In prison at any given time, inmates who were on probation or parole at the time of their offense are called violators, and those who were not on probation or parole are called no status. Probation is a sentence imposed by the court upon a convicted offender, requiring the offender to meet certain conditions of supervision in the community. Generally, probation is given directly, but it may be combined with a period of confinement (often called shock probation or a split sentence) or be imposed in lieu of a suspended jail or prison sentence. Violations (which include new offenses) of the conditions of supervised release may result in the court removing a person from probationary status and ordering incarceration under the suspended sentence of the prior conviction offense, a new offense, or both. Responsibility for enforcing the conditions of supervision in the community normally rests with a probation officer. The term parole refers to both the process for releasing offenders from prison prior to the expiration of their sentences and to the period of conditional supervision in the community following imprisonment. Parole release is an administrative procedure, generally carried out by an appointed commission or board, designed to assist the offender in the transition from confinement to full discharge into the community. Because release from prison to parole supervision occurs prior to the expiration of a sentence, violations of the conditions of release may result in the parolee being removed from parole status and returned to prison for the unexpended balance of the sentence. Reason for imprisonment: New conviction or technical violation Probation and parole violators were sent to prison either because they committed a new offense (for which they were convicted) while on probation or parole, or because they violated a technical condition of their release. A technical violator has failed to comply with the conditions that were imposed at the time he/she was placed on probation or released from prison on parole. Conditions frequently include, but are not limited to, abstaining from drug use, avoiding contact with other criminals, maintaining steady employment, and periodically reporting to a probation or parole officer. Of the 161,885 probation violators in prison, 74% had been convicted of a new offense and the remaining 26% had violated a technical condition of their community supervision. While technical violators had not been convicted of a new crime, a substantial number of them self-reported that they had been arrested for a new crime while on probation (87%). (An arrest for violating probation conditions was not counted as an arrest for a new crime.) Technical probation violators in prison who said they were not arrested for a new crime while on probation comprised about 1% of the prison population. Of the 155,874 parole violators in prison, 80% were in confinement following conviction for a new crime and the remaining 20% had been imprisoned for a technical violation. About 4 in 10 of these technical violators (43%), while not convicted of a new crime, had been arrested for a new crime while on parole supervision in the community. (An arrest for violating parole conditions was not counted as an arrest for a new crime.) Technical parole violators in prison who said they were not arrested for a new crime while on parole comprised about 3% of the prison population. Overall, for 77% of probation and parole violators the reason for their confinement was a new conviction for a crime committed while under community supervision. For the remaining 23% the reason was a technical violation during their period of supervision. Among these technical violators, 69% were arrested for a new crime. * * * * * Reasons for imprisoning technical violators As conditions of their release into the community, persons placed on probation or parole are routinely required to stay out of trouble, abstain from using drugs, avoid contacts with known offenders, maintain steady employment, and report to a probation or parole officer. In addition, such persons are also frequently required to obey special conditions, tailored to their specific circumstances. For example, an offender with a known alcohol problem may be required to participate in alcohol treatment and abstain from drinking. A man convicted of assaulting his ex-wife may be required to obey a restraining order that bars him from trying to see her again. A known drug addict may be required to undergo periodic drug testing. Although the prison inmate survey did not ask probation and parole violators what their specific conditions of release had been during the time they were on probation or parole, it did ask what conditions they had violated. Once a violation occurs, probation/ parole may be revoked. Sometimes revocation means the offender is removed from probation or parole and sent to prison. However, prison does not automatically follow revocation. In some jurisdictions, the offender's probation or parole is revoked but then immediately reinstated. New conditions may be added or the offender may receive a warning that continued infractions will result in imprisonment. Technical violators differ from violators who were sent to prison for committing a new offense while under supervision in that they were not convicted of a crime. While these violators are labeled technical in nature, they were not necessarily crime free. All of the technical violators in prison admitted to violating conditions of their release. Some simply reported being arrested for a new crime. Some reported failing to meet with their probation/parole officer. Others reported combinations of such violations. Which particular violation caused the person to be sent to prison was not determined in the survey. However, this much is known: * 87% of probation violators had at least one arrest for a new crime. However, only 25% had a revocation hearing for being arrested for a new crime. * 43% of parole violators had at least one arrest for a new crime. However, only 27% had a revocation hearing for being arrested for a new crime. * 10% had a revocation hearing for failing a drug test. (An estimated 2% of violators in prison for conviction of a new offense while under supervision in the community reported that in addition to a new conviction they had failed a drug test.) * 36% of probation/parole violators had a revocation hearing for failure to report to their probation/parole officer or absconded. (Approximately 10% of violators in prison for conviction of a new offense while under supervision in the community reported that in addition to a new conviction they had failed to report to their probation/parole officer or absconded.) * Parole violators were more likely than probation violators to have had a revocation hearing because they left the jurisdiction without permission from their probation/parole officer (14% versus 8%). * Probation violators were more likely than parole violators to report they had a revocation hearing for failure to pay fines, restitution, or other financial obligations (12% versus 3%). Technical violations committed while under supervision in the community Technical violators* -------------------- Type of violation Probation Parole ---------------------------------------------------- Arrest for new offense 87.0% 43.0% Positive test for drug use 9.6 10.2 Failure to report for drug testing/treatment 4.9 4.4 Failure to report for alcohol treatment 1.8 2.0 Failure to report to other counseling 3.8 2.3 Failure to report to probation/ parole officer/absconded 36.8 34.2 Left jurisdiction without permission 8.2 14.1 Failure to secure or maintain employment 1.5 2.8 Failure to pay fines, restitution, or other financial obligation 11.5 2.9 Maintained contact with other known offenders 1.7 2.5 Failure to report change of address .5 2.5 Alcohol or drug use 3.3 2.3 Weapons .7 2.5 Other reasons 12.0 17.0 Number of inmates 42,777 30,136 *Detail adds to more than 100% because some inmates had more than one type of violation. ---------------------------------------------------- Trends in probation and parole violators in State prisons In 1991 probation and parole violators comprised 45% of the State prisonpopulation, up from 17% in 1974, when the first national inmate survey was conducted. National surveys of State prison inmates Percent of total State prison population ------------------------------- Violators ------------------------ Year Total Probation Parole Other ------------------------------------- 1974 17% 12% 5% 83% 1979 35 19 16 65 1986 44 21 23 56 1991 45 23 22 55 ------------------------------------- Longer-trend data based solely on characteristics of persons entering State prisons show similar increases (appendix table 2). (Historically, two categories of offenders have comprised around 90% (sometimes more) of all persons entering prison: new court commitments and conditional release violators. New court commitments are sentenced directly to prison by the courts. Conditional release violators did not obey the conditions of their release from prison.) In 1926 "conditional release violators" (parolees and other such persons--released conditionally but returned to prison for violating the terms of their release) comprised 5% of all persons sent to prison (figure 2). Following decades of generally upward growth, the number reached 30% in 1992. Current commitment offense In 1991 nonstatus offenders (persons not on probation or parole when they committed their crime) outnumbered violators among State prison inmates. However, for certain offenses, violators outnumbered those who had no status. 85% of public-order offenders. One reason violators comprised such a large percentage of public-order offenders is that, by definition, public-order offenses include a type of offense--technical violation of probation and parole conditions--that only those on probation or parole can commit. Violators returned to prison as a result of a technical violation of the conditions of their supervision in the community comprised close to 70% of all public-order offenders admitted to prison. About 1 in 4 violators were serving a current sentence for a violent crime. More than half of nonstatus offenders (58%) were serving a current sentence for a violent offense. Property offenses were more characteristic of violators than of nonstatus offenders. Among those with no status, 17% were in prison for a property offense. Among probation violators it was 22%, and among parole violators it was 29%. Approximately 1 in 6 probation and parole violators had a drug offense as their commitment offense. An estimated 15% of parole violators had robbery as their commitment offense compared to 11% of probation violators. The commitment offense of burglary was also more characteristic of parole than probation violators: 15% of parole violators versus 10% of probation violators. By definition, the most serious offense of violators who entered prison for technically violating the conditions of their probation/parole (who were not convicted of a crime) was a probation/ parole violation. A probation/parole violation offense was more characteristic of probation than parole violators: 27% of probation violators versus 19% of parole violators. Demographic characteristics Demographic characteristics of probation violators, parole violators, and nonstatus offenders varied little. Probation violators were younger than parole violators (median age 27 versus 31) or nonstatus offenders (median age 31). Their younger age probably reflects the fact that judges are more likely to sentence a younger offender to probation than an older offender who has a more extensive criminal history. Being younger, probation violators were slightly less likely ever to have been married (36%) than parole violators (45%) or nonstatus offenders (49%). Both among violators and nonstatus offenders, male inmates far outnumbered female inmates. Males comprised over 90% of inmates, whether or not they were violators. Nonstatus offenders were more likely to have completed some college prior to their imprisonment (14.0%) than probation (9.7%) and parole (11.4%) violators. Among violators who had dropped out of school, 45% of parole violators compared to 33% of probation violators had obtained a GED. Overall, parole violators were more likely than probation violators to have attained a high school diploma or its equivalent as the highest level of education (51% versus 44%). Some parole violators may have obtained their high school equivalent degrees during their prior confinement(s). An estimated 4.6% of probation violators were not citizens of the United States, compared to 2.2% of parole violators. Criminal history State prison inmates reported the number of times they had been on probation or incarcerated in the past. Of those with no status at the time of admission, about half reported they had been on probation in the past. With respect to a history of incarceration, about half of nonstatus offenders reported they were serving their first confinement sentence. By definition, all parole violators had previously been incarcerated and all probation violators had previously been on probation. Substantial numbers of violators reported repeated prior incarceration and probation sentences. Probation violators were more likely than parole violators to have been on probation in the past. Among probation violators, 27% reported they had served time on probation 3 or more times. Parole violators were more likely than probation violators to have been incarcerated. Among parole violators, 43% reported they had been incarcerated in jail or prison 3 or more times. Time periods served defined Violators were depicted at three different time periods. The first period is the amount of time incarcerated for the offense that resulted in probation or parole--this is referred to as the prior conviction offense. This period is not relevant to most probation violators because most were not incarcerated. Approximately 20% of probation violators had served a short jail sentence as part of their probation sentence. The second time period is the amount of time spent under community supervision. For parole violators this period began when they were released from prison. The third period is the amount of time inmates expected to serve in prison. This is based on violators' estimates of when they expected to be released from prison for their current offense. Prior conviction offense that resulted in probation or parole Violent offenses are more likely to result in a prison sentence than nonviolent offenses. Not surprisingly, therefore, violent offenses comprised a relatively large percentage (33%) of the offenses that originally resulted in parole violators being placed on parole (appendix table 1). By comparison, violent offenses comprised 20% of the offenses that originally resulted in probation violators being placed on probation. Time on probation or parole prior to being sent to prison About half of probation violators were on probation less than a year before being removed from probation and placed in prison. (Approximately 32,104 probation violators reported that they had served some time in prison prior to probation. This combination of prison term and probation is often called "shock probation" or "split sentences.") The period of community supervision was shorter for parole violators. Half of them had been on parole for 8 months or less before being returned to prison. Before being released from prison and placed on parole, parole violators had served an average of 33 months in prison (tables 5 and 6). Time served in prison before parole release varied widely by offense. Probation and parole violators were placed on probation or parole as a result of their prior conviction offense. For each offense category, probation violators spent more time on average in the community than parole violators. For offenders who had a prior violent offense and were convicted of a new offense while in the community, probation violators, on average, spent 5 months more than parole violators; for property offenses, 5 months more; drug offenses, 4 months more; and public-order offenses, 5 months more. (The longer time in the community for probation violators as compared to parole violators does not take into account the fact that many of the probation violators may not have been free to commit new crimes during a portion of their probationary period because they were behind bars serving a short jail sentence. This combination jail-probation sentence, often called a "split sentence" or "shock probation," is widely imposed.) Expected time to serve in prison Probation and parole violators were asked to report how much time they expected to serve in prison before release. (Technical violators serve the suspended portion of the original sentences they received prior to probation/parole. It is not known if violators convicted of a new crime received a new sentence, were serving the suspended portion of the original sentence they received prior to probation/parole, or if additional time was added to the suspended portion of that sentence. This issue cannot be addressed by this survey.) Their "expected time to serve" is the inmate's own estimate of the time he or she will spend in confinement. Expected time to serve in prison before release was longer for parole violators than for probation violators. The average expected time from admission date to release date was 63 months for parole violators versus 49 months for probation violators. An estimated 29% of parole violators expect to serve 5 years or more in prison compared to 22% of probation violators. Of those convicted of a new offense, parole violators expect on average to serve 14 months more than probation violators (mean sentences of 63 and 49, respectively). Those who technically violated their probation and were currently in prison for a probation violation offense expected to serve on average 2 months more than those who were in prison for a parole violation offense (means of 29 and 27, respectively). For many offense categories, parole violators convicted of a new offense expected to serve more time on average in prison than probation violators convicted of a new offense. For violent offenses, parole violators expected to serve an average of 36 months longer than probation violators; and for property offenses, 5 months longer. Public-order was the only offense category in which probation violators expected to serve longer than parole violators. Probation violators expected to serve a mean of 42 months, 8 months longer than parole violators. * * * * * Criminal justice status at time of offense Surveys conducted among persons at the different stages of the justice system--at arrest, in pretrial detention, in jail or prison serving their sentence--all show that many of these persons were on probation, on parole, on pretrial release (out on bail, for example), or in some other criminal justice status at the time they committed (or were alleged to have committed) their crime. For example, of all persons who were arrested for the murder of a law enforcement officer from 1988 to 1992, 22% were on probation or parole at the time of the killing. Similarly, of arrested murderers who were convicted, acquitted, or otherwise adjudicated in 1992 in urban courts, 38% were on probation, on parole, on pretrial release, or in some other criminal justice status at the time of the murder. * * * * * Current conviction offense versus prior conviction offense that resulted in probation or parole The prior conviction offense was violent for 43% of probation violators with a current violent offense. Similarly, among prisoners who had been on probation after conviction for a drug offense, 41% had a new conviction for a drug crime which brought them to prison. The prior conviction offense was violent for 50% of parole violators with a current violent offense. Among those prisoners who had been on parole after conviction for a drug offense, 46% had a new conviction for a drug crime which brought them to prison. Probation violators were as likely as parole violators to currently be convicted for the same crime that originally resulted in a probation or parole status prior to imprisonment: Offense currently Prior conviction imprisoned for: offense for which ----------------- they received Probation Parole probation/parole violator violator --------------------------------------- Robbery 37.2% 42.0% Assault 18.8 20.5 Burglary 37.2 41.9 Larcency/theft 22.0 33.7 Fraud 31.0 37.8 Drug possession/ trafficking 58.1 59.9 --------------------------------------- Characteristics of offenses committed while on probation or parole Probation violators--During their average 17 months on probation, the nearly 162,000 imprisoned probation violators collectively committed 44,000 violent crimes for which they were convicted and sent to prison (derived from table 2). The number of probation violators convicted of property crimes totaled 35,000; drug offenses, 30,000; other offenses, nearly 10,000. These estimates are conservative because they are based only on crimes for which there was a conviction and sentence to prison, and many crimes do not result in conviction or imprisonment. Furthermore, these estimates do not take into account the number of persons victimized by a particular crime. In greater detail, over the course of an average 17 months in their communities, collectively the 162,000 probation violators: * murdered 6,400 people (more than half of them strangers) * raped or sexually assaulted 6,700 females and 700 males (a third were under the age of 12; 63% under 18) * robbed 17,000 people (86% of them strangers) * assaulted 10,400 people (9% of them wives, ex-wives, girlfriends and ex-girlfriends; half of them strangers) * burglarized 16,600 homes and businesses * stole 3,100 motor vehicles. Parole violators--During their average 13 months on parole, the 156,000 parole violators collectively committed a conservatively estimated: --46,000 violent crimes --45,000 property crimes --24,000 drug offenses --9,000 other offenses In greater detail, based on convictions of the 156,000 collectively: * murdered 6,800 people (half of them strangers) * raped or sexually assaulted 4,900 females and 600 males (21% were under the age of 12; 47% under 18) * robbed 22,500 people (86% of them strangers) * assaulted 8,800 people (14% of them wives, ex-wives, girlfriends and ex-girlfriends; more than half of them strangers) * burglarized 23,000 homes and businesses * stole 4,800 motor vehicles. Based on those with a current conviction for a violent crime, probation violators were more likely to victimize a minor compared to parole violators (13% versus 8%). Victim injury was also more prevalent among probation violators (60%) than parole violators (50%). Death, rape and sexual assault were more prevalent among nonstatus offenders at the time of admission in prison for a violent offense than among violators. About 17% of probation violators reported that they had sexually assaulted or raped their victim. About 12% of parole violators reported they had raped or sexually assaulted their victim. Probation and parole violators in prison for a violent crime were more likely to have victimized a stranger (59%), compared to 46% of nonstatus offenders in prison for a violent crime. Drug and alcohol use An estimated 55% of probation and parole violators who were convicted of a new offense while on probation or parole reported they had used drugs in the month before their arrest. However, 2% of violators who had used drugs in the month before their current offense reported a "positive test for drug use" as the reason for their revocation. (About equal percentages of probation and parole violators--1.7% and 2.6%--had used drugs in the month before their current offense and reported that they had a violation resulting from a positive test for drug use. There is no significant difference between probation and parole violators.) This suggests that much illegal drug use by probationers and parolees may go undetected. There was no significant difference in drug use between probation and parole violators. About 55% of violators had used drugs in the month before the current offense, compared to 46% among nonstatus offenders. Probation and parole violators were also more likely than nonstatus offenders to have used drugs daily in the month preceding their offense (41%), and to have been under the influence of drugs at the time of the offense (35%). Probation violators were more likely to have been under the influence of alcohol at the time of the current offense (34%) compared to parole violators (28%). Probation violators were also more likely to have been under the influence of both alcohol and drugs at the time of the current offense (17%) compared to parole violators (12%). Violators were more likely to have committed their offense to obtain money to buy drugs than nonstatus offenders. While under supervision, about 1 in 5 violators reported committing their offense to get money to buy drugs. More than half of probation and parole violators reported drug use in the month before their current offense and about a third were under the influence of drugs at the time of that crime. Heroin or opiate usage was greater among parole violators while marijuana use was greater among probation violators. Crack and cocaine use was similar for the two groups. At the time of the current offense, probation violators were just as likely to be using cocaine as they were marijuana. Parole violators were just as likely to be using heroin as they were marijuana. Weapons An estimated 24% of probation violators compared to 19% of parole violators convicted of a new offense owned or possessed a firearm in the month before their arrest, a period when they presumably were on probation or parole. Federal law prohibits felons from possessing a firearm, and in many cases, such possession is a violation of State law. Depending on the State, some or all probationers and parolees convicted of misdemeanors are forbidden by law or as a standard condition of probation or parole from possessing a firearm. Therefore, those probation and parole violators who possessed a firearm during their period of supervision were probably eligible to be revoked for a weapons violation. The data on type of violation suggest that of those who could have been revoked for possession of a firearm, less than 1% were actually revoked for a weapons violation. Approximately 21% of probation and parole violators owned or possessed a firearm in the month before arrest. Of those violators who possessed a firearm, an estimated 2 in 3 were armed with a firearm when they committed their current offense. Although 14% of violators (35,000) reported having a firearm when they committed their current offense, 1 out of every 5 of them (7,000) were actually convicted of a weapons offense. Probation violators are just as likely as parole violators to possess a handgun in the month before arrest. About 3 out of 4 violators who owned or possessed a handgun in the month before their arrest were armed with a handgun when committing their current offense. Of those violators armed with a handgun when committing their current offense,1 out of every 3 fired the handgun during the offense. Nonstatus offenders were more likely than violators to have been under the influence of drugs and/or alcohol and armed with a firearm at the time of the current offense. Of all violators who were armed with a firearm when committing their current offense (14%), more than half were under the influence of drugs and/or alcohol at that time. Armed with firearm when commiting Probation Parole No current offense violator violator status -------------------------------------------------- --and under the influence of drug and/or alcohol 8.0% 7.1% 9.6% ------------------------------------------------- Methodology The 1991 Survey of Inmates of State Correctional Facilities was administered for the Bureau of Justice Statistics by the U.S. Bureau of the Census. Personal interviews were conducted with 13,986 inmates to collect data on individual characteristics of State prison inmates, current offense and sentences, characteristics of victims of violent inmates, criminal histories, prior drug and alcohol use, and drug and alcohol treatment provided in prison. Similar surveys were conducted in 1974, 1979, and 1986. Sample design The sample for the 1991 survey was selected from a universe of 1,239 state prisons enumerated in the 1990 Census of State and Federal Adult Correctional Facilities or had been opened between completion of the census and February 29, 1991. The sample design was a stratified, two-stage selection. In the first stage correctional facilities were separated into two sampling frames; prisons with male inmates and prisons with female inmates. Prisons holding both sexes were included on both lists. Within each frame, prisons were separated into eight strata defined by census region (Northeast, Midwest, South, and West) and facility type (confinement and community-based). All prisons with 1,950 or more men were selected from the male frame, and all prisons with 380 or more women were selected from the female frame. The remaining prisons in the male frame were grouped into equal sized strata of approximately 2,600 males and then stratified by security level (maximum, medium, minimum, and classified). The remaining prisons in the female frame were grouped into strata of approximately 574 females. A systematic sample of prisons was then selected within strata on each frame, with probabilities proportional to the size of each prison. Overall, a total of 277 prisons were selected. In the second stage, interviewers visited each selected facility and systematically selected a sample of male and female inmates using predetermined procedures. As a result, approximately 1 in 52 males and 1 in 11 females were selected. A total of 13,986 interviews were completed, yielding an overall response rate of 93.7%. Based on the completed interviews, estimates for the entire population were developed using weighting factors derived from the original probability of selection in the sample. These factors were adjusted for variable rates of nonresponse across strata and inmate characteristics. Further adjustments were made to control the survey estimates to midyear 1991 custody counts projected from data obtained in the National Prisoner Statistics series (NPS-1). Accuracy of the estimates The accuracy of the estimates presented in this report depends on two types of errors: sampling and nonsampling. Sampling error is variation that may occur by chance because a sample rather than a complete enumeration of the population was conducted. Nonsampling error can be attributed to many sources, such as nonresponse, differences in the interpretation of questions among inmates, recall difficulties, and processing errors. In any survey the full extent of the nonsampling error is never known. The sampling error, as measured by an estimated standard error, varies by the size of the estimate and the size of the base population. Estimates of the standard errors have been calculated for the 1991 survey of prison inmates, (see appendix table 4). These standard errors may be used to construct confidence intervals around percentages in this report. For example, the 95-percent confidence interval around the percentage of probation violators who are between the ages of 18-29 is 61.2 plus or minus 1.96 times 1.3 (or 58.7% to 63.7%). These standard errors may also be used to test the statistical significance of the difference between two sample statistics by pooling the standard error of the two sample estimates. For example, the standard error of the difference in the percentage of probation violators in 1991 who were between the ages of 18 and 29 compared to parole violators between the ages of 18 and 29 would be 1.94 (or the square root of the sum of the squared standard errors in each category). The 95-percent confidence interval around the difference would be 1.96 times 1.91% (or 3.8%). Since the observed difference of 20% (61.2% minus 41.2%) is greater than 3.8%, the difference would be considered significant. Comparisons discussed in this report were determined to be statistically significant at the 95-percent confidence level. Statements of comparison qualified by language such as "slightly", "somewhat," or "marginal" indicate statistical significance at the 90-percent level (1.6 standard errors). Because of the sample design, State, local, or other subnational estimates cannot be made. Self-reported information Through a series of questions, inmates were asked if they were currently serving a sentence. Those inmates who responded they were not, but were pending trial, sentencing, or a probation or parole revocation hearing, were considered not sentenced and were not included in this report. The data presented here and the results will differ somewhat from those in other BJS reports. Criminal history data are based on self-reported information provided by each respondent. Through a series of questions, inmates were asked to report on past probation sentences as juveniles and as adults and on past sentences to incarceration up to 10 prior times. For each sentence, the inmates were asked the offenses for which they were sentenced, the type of institution in which they served time, the date of admission, and the length of time expected to serve. From this information, a criminal history profile was constructed. * * * * * The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. Jan M. Chaiken, Ph.D., is the director. BJS Special Reports address a specific topic in depth from one or more datasets that cover many topics. Robyn L. Cohen of the Bureau of Justice Statistics wrote this report. Substantial assistance in preparing this report was provided by Patrick A. Langan. Statistical review was provided by Tina Dorsey, Lisa D. Bastian, and Caroline Harlow. Graphical assistance and advice were provided by Marianne W. Zawitz and Rhonda C. Keith. Tom Hester edited the report. Marilyn Marbrook administered final report production, assisted by Jayne E. Robinson. August 1995, NCJ-149076 END OF FILE