U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Special Report Lifetime Likelihood of Going to State or Federal Prison March, 1997, NCJ-160092 The data in this report have been updated. See Prevalence of Imprisonment in the U.S. Population, 1974-2001 at: www.ojp.usdoj.gov/bjs/abstract/piusp01.htm The full text of this report is available through: * the BJS Clearinghouse, 1-800-732-3277 * on the Internet at http://www.ojp.usdoj.gov/bjs/ * on the BJS gopher: gopher://www.ojp.usdoj.gov:70/11/bjs/ * on the National Criminal Justice Reference Service Electronic Bulletin Board (set at 8-N-1, call 301-738-8895, select BJS). By Thomas P. Bonczar and Allen J. Beck, Ph.D. BJS Statisticians ------------------------- Highlights ------------------------------------------------------------------------- * If recent incarceration rates remain unchanged, an estimated 1 of every 20 persons (5.1%) will serve time in a prison during their lifetime. * Men (9.0%) are over 8 times more likely than women (1.1%) to be incarcerated in prison at least once during their life. * Among men, blacks (28.5%) are about twice as likely as Hispanics (16.0%) and 6 times more likely than whites (4.4%) to be admitted to prison during their life. * Among women, 3.6% of blacks, 1.5% of Hispanics, and 0.5% of whites will enter prison at least once. * The chance of going to prison for the first time declines with age-- Chance of going to prison during Not yet rest of life incarcerated All White Black by age persons men men Birth 5.1% 4.4% 28.5% 20 4.5 4.1 25.3 25 3.1 3.0 17.3 30 2.1 2.1 10.8 35 1.4 1.5 6.5 40 .9 1.1 3.6 45 .6 .8 2.1 * Nearly two-thirds of those admitted to prison for the first time will have been on probation and a third will have served a sentence to a local jail or juvenile facility. -------------------------------------------------------------------------------- An estimated 5.1% of all persons in the United States will be confined in a State or Federal prison during their lifetime, if incarceration rates recorded in 1991 remain unchanged in the future. The lifetime chances of a person going to prison are higher for men (9.0%) than for women (1.1%) and higher for blacks (16.2%) and Hispanics (9.4%) than for whites (2.5%). At current levels of incarceration a black male in the United States today has greater than a 1 in 4 chance of going to prison during his lifetime, while a Hispanic male has a 1 in 6 chance and a white male has a 1 in 23 chance of serving time. The lifetime likelihood of incarceration, the focus of this study, represents the percentage of all U.S. residents expected to be incarcerated in a prison at some time in life. Such a measure is hypothetical _ based on the assumption that recent rates of incarceration (and death rates) will not change in the future. Unlike a standard incarceration rate, which expresses the prevalence of incarceration on a single day, the lifetime likelihood measure summarizes the chances of a person being incarcerated at some time over an entire lifetime. ------------------------------------------------------ Summarizing life experiences based on annual rates ------------------------------------------------------ Estimates of the lifetime likelihood of going to prison are derived from standard life table techniques used by demographers and actuaries to summarize observed rates and to project them into the future. When applied to the study of the U.S. prison population, the lifetime estimates represent the combined experience of first admission to prison and mortality by age of the U.S. resident population during a 12-month period. The estimates do not reflect the experience of any actual group but rather of a hypothetical group that is subject to a fixed set of age-specific rates of first incarceration and mortality. As such, the estimates depict the projected mortality and incarceration experience of newborn U.S. residents, assuming that the mortality and incarceration rates observed during a 12-month period remain unchanged in the future. ( See Methodology.) Life table techniques have been applied in numerous fields to summarize the likelihood that an event may occur. The following four examples use lifetime rates to express statistics about familiar life events: * 5 out of 6 persons are expected to be a victim of an attempted or completed violent crime (rape, robbery, and assault, excluding murder) at least once during life, based on 1975-84 annual victimization rates. (See Lifetime Likelihood of Victimization, BJS, NCJ-10427, March 1987.) * 1 of every 100 men and of every 323 women are expected to be a victim of murder, based on 1978-80 annual murder rates and lifetime probabilities. (See Crime in the United States, 1981, FBI.) * 49.6% of all new marriages are expected to end in divorce, based on marriage and divorce statistics for 1976-77. (See National Estimates of Marriage Dissolution and Survivorship: United States, National Center for Health Statistics, November 1980.) * 1 of every 8 American women are projected to develop breast cancer in their lifetime, if 1987-88 rates remain unchanged. (See The Lifetime Risk of Developing Breast Cancer, Journal of the National Cancer Institute, June 1993.) ------------------------------------------------------------------------- --------------------------------------------- The chances of going to prison, by sex, race, and Hispanic origin ---------------------------------------------- Life table procedures have been repeated for males and females, whites, blacks, and Hispanics--subgroups with different rates of death and first incarceration. Estimates coming from these procedures summarize the combined impact of the differential rates. Overall, an estimated 5.1% of all persons in the United States will serve time in a State or Federal prison during their lifetime, if recent rates of first incarceration and mortality remain unchanged. ----------------------------------------------------------------------------------------------------------------- Table 1. Chances of going to State or Federal prison for the first time, by age, sex, race, and Hispanic origin Cumulative percent of resident population expected to go to State or Federal prison for the first time, by age-- 20 25 30 35 40 45 50 55 65 Lifetime Total 1.1 2.4 3.3 4 4.4 4.7 4.9 5 5.1 5.1 Sex Male 2.0 4.3 5.9 7.0 7.8 8.2 8.5 8.7 9.0 9.0 Female .1 .3 .6 .8 .9 1 1 1 1.1 1.1 Race/Hispanic origin White* .4 0.9 1.4 1.7 2 2.1 2.3 2.4 2.5 2.5 Male .7 1.7 2.5 3 3.5 3.7 4 4.1 4.3 4.4 Female 0 .2 .3 .4 .4 .5 .5 .5 .5 .5 Black* 4.1 8.4 11.6 13.6 14.9 15.4 15.7 15.8 16 16.2 Male 7.9 15.9 21.4 24.6 26.6 27.4 27.8 28 28.3 28.5 Female .3 1.1 2 2.7 3.1 3.3 3.4 3.5 3.6 3.6 Hispanic 1.5 3.6 5.2 6.3 7.5 8.2 8.8 9.1 9.4 9.4 Male 2.8 6.3 8.8 10.7 12.7 13.9 14.9 15.4 15.9 16 Female .1 .4 .7 .9 1.2 1.3 1.4 1.5 1.5 1.5 Note: The cumulative percents represent the chances of being admitted to State or Federal prison for the first time, by age. Estimates were obtained by sequentially applying age-specific first- incarceration rates and mortality rates for each group to a hypothetical population of 100,000 births. (See Methodology.) *Excludes persons of Hispanic origin. -------------------------------------------------------------------------------------------------------------------- Men are over 8 times more likely than women to be incarcerated in prison at least once during their lifetime. A male has a 9.0% (or 1 in 11) chance in his lifetime of going to prison, while a female has a 1.1% (or 1 in 91) chance. Blacks in the U.S. resident population (16.2%), regardless of their sex, are nearly twice as likely as Hispanics (9.4%) and 6 times more likely than whites (2.5%) to be admitted to prison during their lives. Among men, more than 1 in 4 blacks and 1 in 6 Hispanics, compared to 1 in 23 whites, will enter prison at least once if current first-incarceration rates do not change. An estimated 28.5% of black men, 16.0% of Hispanic men, and 4.4% of white men are expected to serve a State or Federal prison sentence. In general, women have lower lifetime chances of incarceration than men; however, black women (3.6%) have nearly the same chance as white men (4.4%) of serving time in prison. Among women, blacks are more than twice as likely as Hispanics (1.5%) and 7 times more likely than whites (0.5%) to be admitted to prison during their lives. --------------------------------------------- The likelihood of going to prison, by age --------------------------------------------- In addition to permitting calculation of the lifetime chances of incarceration, the same life table techniques produce estimates of the likelihood of incarceration for persons at specific ages. If the rates of first incarceration and mortality recorded in 1991 remain unchanged, an estimated 1.1% of all persons born today will go to prison by the time they are age 20; 3.3% by the time they are 30; and 4.4% by the time they are 40. At every age, men have higher chances of going to prison than women, and blacks and Hispanics have higher chances than whites. Based on current rates of first incarceration, an estimated 7.9% of black males will enter State or Federal prison by the time they are age 20, compared to 2.8% of Hispanic males and 0.7% of white males. The percentages double in the next 5 years of life and then begin to increase more slowly: By age 25, 15.9% of black males, 6.3% of Hispanic males, and 1.7% of white males are expected to have served some time in State or Federal prison. By age 30 -- 21.4% of black males, 8.8% of Hispanic males, and 2.5% of white males By age 35 -- 24.6% of black males, 10.7% of Hispanic males, and 3.0% of white males By age 40--26.6% of black males, 12.7% of Hispanic males, and 3.5% of white males will have served time. ------------------------------------------------------------ The chances of going to prison decline with advancing age ------------------------------------------------------------ The life table calculations also provide estimates of the likelihood of going to prison at some time during the rest of life among persons not previously incarcerated. Unlike the lifetime chances and the cumulative percents by age, these estimates represent the chances of going to prison in the future for persons not previously sentenced to prison. If age-specific first-incarceration rates remain unchanged, the future chances of going to prison among persons not previously incarcerated decline with age. For example, among U.S. residents who are age 30 and have not been previously incarcerated, an estimated 2.1% are expected to go to prison at some time before they die. -------------------------------------------------------------------------------------------------- Table 2. Chances of going to State or Federal prison at some time during the rest of life, by age, sex, race, and Hispanic origin Percent expected to go to State or Federal prison at some time during the rest of life among persons not previously incarcerated, by age-- Birth 20 25 30 35 40 45 Total 5.1% 4.5% 3.1% 2.1% 1.4% .9% .6% Sex Male 9.0 7.9 5.5 3.7 2.5 1.6 1.0 Female 1.1 1.0 .8 .6 .3 .2 .1 Race/Hispanic origin White* 2.5% 2.3% 1.7% 1.2% .9% .6% .4% Male 4.4 4.1 3.0 2.1 1.5 1.1 .8 Female .5 .5 .4 .3 .2 .1 .1 Black* 16.2% 14.1% 9.6% 6.0% 3.6% 2.0% 1.2% Male 28.5 25.3 17.3 10.8 6.5 3.6 2.1 Female 3.6 3.5 2.8 1.9 1.1 .6 .4 Hispanic 9.4 8.7 6.4 4.9 3.8 2.3 1.6 Male 16.0 14.8 11.1 8.6 6.8 4.3 3.0 Female 1.5 1.5 1.2 .9 .6 .4 .2 Note: Estimates were obtained by substracting the cumulative percent first incarcerated for each age from the lifetime likelihood of incarceration. (See Methodology.) *Excludes persons of Hispanic origin. --------------------------------------------------------------------------------------------- Among those 35 years old, an estimated 1.4% will go to prison; among those age 40, fewer than 1%. The chances of going to prison decline with age among all persons regardless of their sex, race, or Hispanic origin. The likelihood of going to prison is highest among black males and Hispanic males. Among men age 30 or younger, blacks have a greater chance of incarceration than Hispanics; among men age 35 or older, Hispanics have a greater chance of incarceration than blacks. Except for black and Hispanic males, persons who reach age 45 without having been sentenced to prison will have less than a 1% chance of going to prison during the rest of their life. After age 45, about 0.1% of all females and 0.8% of white males are expected to be incarcerated for the first time. -------------------------------------------------------------------- Why lifetime estimates exclude admissions to local jails Jails are locally operated correctional facilities that confine persons before and after adjudication. Unlike prisons, jails admit persons with sentences of a year or less. Jails also hold a wide variety of categories of inmates -- including those persons awaiting arraignment or trial; those with sentences of more than a year and awaiting transfer to State or Federal facilities; and those temporarily detained, under protective custody, or awaiting transfer to appropriate health facilities. Estimates of the lifetime likelihood of incarceration are limited to the chances of going to State or Federal prison. Data on first admissions to local jails, which are needed to incorporate the chances of going to a local jail in these calculations, do not exist: * In 1993, when the most recent Census of Local Jails was conducted, the annual number of new admissions to local jails totaled 9.8 million -- nearly 30 times the number of new court commitments to State and Federal prison during that year. (See Jails and Jail Inmates 1993-94, NCJ-151651, April 1995.) The census did not collect any data on the number of persons admitted to jail for the first time. * Most jail inmates serve only a few days before release. (See Pretrial Release of Felony Defendants, 1992, NCJ-148818, November 1994.) As a result, surveys of inmates based on persons held on a single day will not provide reliable statistics on those persons admitted during a 12-month period. ----------------------------------------------------------------------- ---------------------------------------------------------------------- The number of first admissions by age, sex, race, and Hispanic origin ---------------------------------------------------------------------- All of the life table calculations are derived from the number of first admissions to prison by age. Based on the 1991 surveys of State and Federal prisoners, a total of 169,773 persons in State and Federal prison at the time of the survey had been admitted to prison for the first time during the previous 12 months. An estimated 157,148 persons had been admitted to State prison and 12,625 to Federal prison. These estimates, however, exclude persons who had been admitted and then subsequently released during the 12 months prior to the surveys. To obtain the total number of first admissions, a series of adjustment factors (by sex, race, and Hispanic origin) were applied to the survey estimates. Based on data from the 1990 and 1991 National Corrections Reporting Program, these factors take into account the number of persons who entered and then were released prior to the surveys. Overall, nearly 20% of all persons admitted from court during the period (July 16, 1990, to July 15, 1991) were released prior to the surveys. As a result, an adjustment factor of 1.245 was applied to the survey estimate to produce a total of 211,400 first admissions. Separate factors were applied to each group by sex, race, and Hispanic origin. (See Methodology for details.) Number of first admissions 1991 Estimated survey total estimate number Total 169,773 211,400 Sex Men 154,002 190,100 Women 15,771 21,600 Race/Hispanic origin White 57,461 72,500 Black 73,066 92,100 Hispanic 35,414 40,600 Other 3,832 6,200 Age at admission 17 or younger 5,664 7,100 18-24 62,494 77,800 25-34 62,170 77,400 35-44 28,000 34,900 45-54 7,549 9,400 55 or older 3,896 4,900 Note. Detail may not add to total because categories were estimated separately. ------------------------------------------------- Characteristics of first admissions ------------------------------------------------- Among persons admitted to State or Federal prison for the first time in the 12 months prior to the 1991 inmate surveys, nearly a third had committed a violent offense. --------------------------------------------------------------------------------- Table 3. Characteristics of persons admitted to State or Federal prisons for the first time during the 12 months preceding July 1991 Percent of persons admitted Characteristic Total State Federal Number of persons admitted/a 169,773 157,148 12,625 Most serious offense/b Violent offenses 32.8% 34.9% 7.3% Homicide 5.7 6.1 .6 Sexual assault 7.5 8.1 .3 Robbery 9.9 10.3 4.5 Assault 8.1 8.7 1.2 Property offenses 24.5% 25.3% 14.8% Burglary 10.1 10.9 0 Larceny 5.4 5.7 1.7 Fraud 4.6 4.0 11.8 Drug offenses 33.4% 31.1% 61.6% Trafficking 21.4 19.3 46.6 Possession 11.3 11.2 13.2 Public-order offenses 8.5% 8.1% 12.9% Other offenses .8% .6% 3.3% Prior sentence to probation No 36.4% 34.7% 62.5% Yes 62.5 65.3 37.5 Number of times 1 36.2 37.5 25.5 2 15.6 16.4 7.3 3 or more 10.7 11.4 4.7 Prior sentence to local jail/c No 67.6% 66.9% 85.1% Yes 31.4 33.1 14.9 Number of times 1 14.6 15.1 10.0 2 6.1 6.5 2.0 3 or more 10.8 11.6 2.9 Current sentence length Meand 85 mo 86 mo 73 mo Median 60 60 51 Note: The 1991 surveys of State and Federal prisoners provided estimates of persons in prison at the time of each survey who had been admitted for the first time during 12 months preceding the surveys. Excluded from these estimates are persons admitted during the 12 months who had been released prior to the surveys. a/Data on offense were known for 98% of first admissions; on prior sentences to probation and jail for 99%; and on sentence length for 96%. b/Detail may not sum to total because only selected offenses are presented. c/Includes those inmates who had been admitted to juvenile facilities. d/Excludes persons with sentences of life, life without parole, life plus additional years, and death. ----------------------------------------------------------------------------------- Another third had committed a drug offense, and a quarter, a property offense. The most frequent offense among first time admissions was drug trafficking -- 19.3% of all State prisoners and 46.6% of Federal prisoners. The type of offense leading to the first admission to prison varied with the age of the offender. About 22% of first-time prisoners age 45 or older had committed a sexual assault, compared to fewer than 5% among those under age 25. --------------------------------------------------------------------------------------------- Table 4. Most serious offense of persons admitted to State and Federal prison for the first time, by age Most serious Percent of prison inmates, by age offense 45 ages All 19 or younger 20-24 25-34 35-44 or older Total 100 100 100 100 100 100 Violent offenses 32.8% 44.5% 32.2% 29.4% 29.6% 39.8% Homicide 5.7 6 4.8 5.8 4.9 9.9 Sexual assault 7.5 4.4 4.1 6.6 12.4 22.1 Robbery 9.9 19.7 12.3 8.8 3.8 1.1 Assault 8.1 12.6 8.6 6.8 7.2 6.4 Other violent 1.6 1.8 2.4 1.4 1.3 .3 Property offenses 24.5% 24.8% 28.7% 26.0% 18.7% 12.4% Burglary 10.1 13.4 12.8 11.0 4.9 .1 Larceny 5.4 4.5 5.9 5.9 5.3 2.5 Fraud 4.6 1.4 3.7 5.2 5.8 9.1 Other property 4.4 5.5 6.4 3.9 2.7 .7 Drug offenses 33.4% 26.5% 30.8% 35.7% 39.1% 30.5% Trafficking 21.4 18.6 19.2 23.6 23.3 18.8 Possession 11.3 7.6 10.8 11.3 15.2 11.1 Public-order offenses 8.5% 4.2% 7.7% 8.2% 11.1% 16.4% Other offenses .8% 0 .6% .8% 1.5 .8% Number of persons admitted 166,754 21,802 46,014 60,760 28,395 9,783 --------------------------------------------------------------------------------------------- Nearly 20% of first-time prisoners under age 20 had committed robbery, and 13% burglary, compared to 1% of those age 45 or older who had committed robbery and fewer than 1% burglary. The percentage of persons admitted for a public-order offense (such as driving under the influence, commercialized vice, weapons offenses, and Federal regulatory offenses) increased with age. Two-thirds of all persons entering prison for the first time had a prior sentence to probation and a third had been sentenced to a local jail or served time in a juvenile facility. Persons entering State prisons were more likely than those entering Federal prison to have had prior sentences to jail or probation. On average, persons entering State prison for the first time had received a total maximum sentence of 86 months; those entering Federal prison received a sentence of 73 months. Half of all first admissions to State and Federal prisons combined had received a sentence of 60 months or more. Such characteristics of first admissions to State and Federal prison may also be viewed prospectively -- if the patterns of offending and sentencing remain unchanged. For example, among the 5.1% of the hypothetical birth cohort who are expected to go to prison during their lifetime, nearly two-thirds will have been on probation and a third will have served a sentence to a local jail or juvenile facility before entering prison. In addition, about two-thirds of those going to prison will have been sentenced for either a violent or drug offense. -------------------------------------- Steady chances of going to prison since 1991 -------------------------------------- Although the life table calculations are derived from survey data collected in 1991, other data suggest that the chances of going to State or Federal prison have not since changed. Data collected annually in the National Prisoners Statistics (NPS) series, when combined with data from surveys of State prisoners conducted in 1974, 1979, 1986, and 1991, indicate that factors underlying calculations of the lifetime likelihood of incarceration have not changed since 1991. Between 1991 and 1995 the number of inmates in State and Federal prison increased 36%, from 792,535 to 1,078,545 prisoners. ---------------------------------------------------------------------------------------------------- Table 5. Other measures of the prevalence of incarceration in State or Federal prison, 1991-95 1991 1992 1993 1994 1995 Inmates in State or Federal prison at yearend/a Number 792,535 850,566 909,381 990,147 1,078,545 Rate per 100,000 adult resident 419 446 472 510 550 Total annual admissions to prison/c Number 502,464 520,340 518,562 544,034 562,724 Rate per 100,000 adult resident 266 273 269 280 287 New court commitments during the year/d Number 337,478 356,498 341,722 346,097 361,464 Rate per 100,000 adult resident 179 187 177 178 184 a/Data are based on the total number of inmates in custody of State and Federal adult correctional authorities. b/All rates are based on the U.S. resident population age 18 or older on July 1 of each year and have been adjusted for the 1990 Census undercount. c/Admissions are of prisoners sentenced to more than 1 year as reported annually in BJS' National Prisoners Statistics program (NPS-1). Previously published counts have been revised to include admissions to Federal prisons (reported in the NCRP for 1991 and 1992) and admissions to prisons in Alaska (estimated for 1994). d/New court commitments are those entering prison directly from a sentence by a court. Includes new court admissions and persons entering prison for the first time on any probation offenses. Excludes parole or other conditional release violators with new sentences, transfers, AWOL or escapee returns, and returns from appeal or bond. -------------------------------------------------------------------------------------------------------------- Relative to the number of adults in the U.S. population, the prison population rose from 419 per 100,000 adult residents in 1991 to 550 per 100,000 in 1995. During the same period the number of new court commitments rose from 337,478 admitted in 1991 to 361,464 in 1995. New court commitments exclude parole violators, other conditional release violators, and other offenders returned to prison. Relative to the adult population, however, the number of new court commitments remained almost unchanged -- 179 new court commitments per 100,000 adults in 1991 compared to 184 per 100,000 in 1995. First-incarceration rates, from which life table estimates are derived, are a subset of new court commitments. Based on data from past BJS surveys of inmates in State prison, about 50% of all new court commitments in any year are first admissions: First admissions as Survey a percent of new year court commitments* 1974 48.3% 1979 53.9 1986 48.6 1991 49.5 *Based on State inmates only, without adjusting for differing time periods and releases prior to each survey date. Other than persons entering a State or Federal prison for the first time, new court commitments include those who have been previously admitted on sentences for other offenses. In combination, these data suggest that the rate of first incarceration has remained relatively unchanged. While the Nation's prison population has risen by 36% since 1991, the number of new court commitments has risen by 7%. Assuming that the proportion of first admissions among new court commitments has stayed at 1974-91 levels, the number of first admissions may be expected to have increased by about 7%. When considered relative to the adult resident population, the first admission rate, like the new court commitment rate, is likely to have changed little since 1991. As a result, the overall lifetime chances of going to prison are projected to have remained stable through yearend 1995. -------------------------------------------------------------------- Comparing lifetime rates to other types of incarceration rates -------------------------------------------------------------------- Standard incarceration rates express the prevalence of incarceration on a single day or during a single year. At yearend 1995 about half of 1% of the Nation's adult population was confined in a State or Federal prison. During 1995 courts sent less than two-tenths of 1% of U.S. residents to a State or Federal prison. ------------------------------------------------------------------------------------------------ Table 6. Estimates of the prevalence of incarceration, by sex, race, and Hispanic origin Prisoners in State or New court commitments to State or Federal prison during 1995 Rate per Rate per Lifetime 100,000 adult 100,000 adult likelihood of Number/a residents/b Number/c residents/b incarceration All 1,078,545 550 361,464 184 5.1 Male 1,014,543 1,075 326,403 346 9.0 White, non-Hispanic 362,100 507 116,300 163 4.4 Black, non-Hispanic 448,200 4,424 158,700 1,566 28.5 Hispanic 179,700 1,957 47,000 512 16.0 Female 64,002 63 35,061 36 1.1 White, non-Hispanic 22,700 30 13,200 17 .5 Black, non-Hispanic 29,000 243 17,700 148 3.6 Hispanic 10,000 112 3,600 40 1.5 a/Total numbers of inmates in custody by sex are from NPS-1, 1995. Counts by race and Hispanic origin were estimated from NPS jurisdiction totals for 1995 and data on race by Hispanic origin from the 1991 surveys of State and Federal inmates. b/All rates are based on the U.S. resident population age 18 or older on July 1, 1995, and have been adjusted for the 1990 Census undercount. c/New court commitments are restricted to prisoners with a sentence of more than 1 year. Total number of prisoners admitted by sex are from NPS-1, 1995. Counts by race and Hispanic origin were estimated using data from the 1994 National Corrections Reporting Program and the Federal Justice Statistics Program. ----------------------------------------------------------------------------------------------------- These rates, though small when compared to the total adult population, represent large numbers of offenders who experience incarceration. At yearend 1995 nearly 1.1 million adults were held in a State or Federal prison. During 1995 more than 360,000 adults received a prison sentence of more than a year from State and Federal courts. Unlike 1-day prevalence rates or 1-year admission rates, the lifetime likelihood measures summarize the chances of going to prison over an entire lifetime. Overall, the lifetime chances (5.1%) based on the 1991 survey data were nearly 10 times the 1-day prevalence rates (0.55% or 550 per 100,000) observed at yearend 1995. All of the measures of incarceration reveal large differences by sex, race, and Hispanic origin; however, differences in lifetime rates are typically smaller than differences in 1-day prevalence rates. Over a lifetime men are about 9 times more likely than women to serve a sentence in a State or Federal prison. At yearend 1995 men were 17 times more likely than women to be in prison. (See the figures below.) Among non-Hispanic men, blacks are 6.5 times more likely than whites to serve some time in prison during their life, but 8.7 times more likely to be in prison on any 1 day (table 6). Among non-Hispanic women, blacks are 7.2 times more likely than whites to go to prison during their life, but 8.1 times more likely to be in prison on any 1 day. Underlying these differences are variations between men and women and among blacks, whites, and Hispanics in rates of recidivism. (See Recidivism of Prisoners Released in 1983, NCJ-11940, April 1989.) Measures of the lifetime chances of going to prison, based on age-specific first-incarceration rates, are unaffected by differences in recidivism among groups. One-day prevalence rates, however, include persons who have been to prison before. ------------------ Methodology ------------------ ************************** Life table techniques ************************** Life table techniques were used to incorporate the effects of mortality and incarceration into a single statistical model. In generating estimates of the lifetime chances of going to prison, the model assumes a constant set of age-specific mortality and incarceration rates. Since the computations involve two forms of exit from an initial cohort, they represent an application of what is known to demographers as double-decrement life table techniques. (For other applications, see Shryock and Siegel, The Methods and Materials of Demography, U.S. Department of Commerce, October 1971.) Conceptually, a life table traces a cohort of 100,000 newborn babies through their entire lives under the assumption that they will be subject to currently observed age-specific mortality and incarceration rates. Lifetime estimates are computed by successively reducing this hypothetical population of 100,000 births by the number of persons expected to die and the number of persons incarcerated for the first time at each year of age. The cumulative sum of the projected number of persons incarcerated at each age represents the number of persons per 100,000 expected to be incarcerated over the course of a lifetime. Dividing the total by 1,000 yields an estimate of the percentage of persons likely to be incarcerated during their lifetimes. All of the life table calculations are based on mortality and incarceration rates for single years of age through age 100. Calculations for age groups 0 to 13 and ages 56 or older were based on data for each single year of age and then grouped for presentation. Calculations were closed out at age 100. ******************************* Age-specific mortality rates ******************************* The Population Division, U.S. Bureau of the Census, provided age-specific mortality data for 12 different demographic groups, defined by sex, race, and Hispanic origin. The rates for each demographic group were estimated for each year of age based on death data reported to the National Center for Health Statistics for the 12-month period ending June 30, 1992. Unlike crude death rates (based on the number of deaths divided by the number of persons of the same age), age-specific mortality rates represent the proportion of persons alive at the beginning of an age interval who are expected to die during the age interval. For each year of age, these mortality rates may be interpreted as the probability that persons at their x-th birthday will die before reaching their next birthday. For ease of reporting, these probabilities have been expressed as rates per 100,000. (Appendix table 1.) ---------------------------------------------------------------------------------- Appendix table 1. Age-specific mortality rates, July 1, 1991, to June 30, 1992 Number Number dying per dying per Age 100,000 Age 100,000 0-13 ... 36 189 14 48 37 201 15 67 38 211 16 84 39 220 17 97 40 230 18 104 41 242 19 107 42 255 20 109 43 270 21 111 44 286 22 113 45 305 23 114 46 326 24 114 47 352 25 113 48 383 26 113 49 418 27 114 50 458 28 118 51 502 29 122 52 552 30 128 53 608 31 134 54 670 32 141 55 737 33 151 56-60 ... 34 163 61-70 ... 35 175 71 or older ... Note: These age-specific mortality rates represent the proportion of persons alive at the beginning of the age interval dying during the interval (times 100,000). For basic assumptions and estimation procedures, see Population Projections of the United States, by Age, Sex, Race, and Hispanic Origin: 1993 to 2050, Current Population Reports, P25-1104, November 1993. ...Not shown, because all calculations were based on rates for single years of age. ------------------------------------------------------------------------------------------- *********************************************************************** Estimating the number of persons at risk to first incarceration *********************************************************************** The population at risk to first incarceration is the number of persons in the U.S. resident population by exact year of age who have not been previously incarcerated in a State or Federal prison (Appendix table 2). ----------------------------------------------------------------------- Appendix table 2. Estimates of the number of U.S. residents at risk to first incarceration, by age, January 1, 1991 Number of persons in Number of persons U. S. resident prison or at risk to first population, on parole, incarceration, 1/1/91/a 12/31/90/b 1/1/91/c Age (1) (2) (3) 0-13 53,027,564 0 53,027,564 14 3,404,925 116 3,404,809 15 3,401,313 117 3,401,196 16 3,436,075 1,536 3,434,539 17 3,475,353 2,665 3,472,688 18 3,567,430 8,520 3,558,910 19 3,994,822 16,911 3,977,911 20 4,280,930 27,097 4,253,833 21 3,984,084 37,062 3,947,022 22 3,802,151 45,201 3,756,950 23 3,767,250 48,969 3,718,281 24 3,921,965 52,667 3,869,298 25 4,088,954 58,505 4,030,449 26 4,259,741 61,782 4,197,959 27 4,363,213 61,696 4,301,517 28 4,257,819 63,465 4,194,354 29 4,539,502 62,230 4,477,272 30 4,736,119 61,758 4,674,361 31 4,409,419 61,695 4,347,724 32 4,416,914 58,519 4,358,395 33 4,454,119 52,088 4,402,031 34 4,465,642 51,317 4,414,325 35 4,393,016 45,608 4,347,408 36 4,189,629 43,912 4,145,717 37 4,080,089 41,183 4,038,906 38 3,874,636 34,604 3,840,032 39 3,948,668 30,362 3,918,306 40 3,975,783 31,065 3,944,718 41 3,659,296 26,700 3,632,596 42 3,669,784 22,640 3,647,144 43 3,789,348 20,441 3,768,907 44 3,459,696 18,114 3,441,582 45 2,920,338 16,132 2,904,206 46 2,824,184 13,152 2,811,032 47 2,967,838 10,540 2,957,298 48 2,775,615 10,699 2,764,916 49 2,647,185 9,020 2,638,165 50 2,480,456 7,818 2,472,638 51 2,284,098 6,901 2,277,197 52 2,271,908 5,648 2,266,260 53 2,214,263 5,208 2,209,055 54 2,160,156 4,697 2,155,459 55 2,130,246 4,580 2,125,666 56-60 10,352,262 15,699 10,336,563 61-70 20,174,425 13,638 20,160,787 71 or older 19,429,758 2,513 19,427,245 a/Estimates of the number of U.S. residents by age on 1/1/91 were obtained from the U.S. Bureau of the Census (U.S. Population Estimates, by Age, Sex, Race, and Hispanic Origin: 1990 to 1995, PPL-41.) Using Post Enumeration Survey (PES) estimates, the published data were adjusted for the undercount in the 1990 decennial census by adding an adjustment factor for each year of age. b/The number of persons in State and Federal prison or on parole, by age, at yearend 1990 were estimated. Yearend totals were obtained from BJS' National Prisoners Statistics (NPS-1) and Annual Parole Data Survey. Counts by each year of age were then estimated by multiplying the age distribution of prisoners (from the 1991 inmate surveys) or by the age distribution of new parolees (from the 1991 National Corrections Reporting Program). c/Estimates were calculated by subtracting column (2) from column (1). This procedure may result in a slight overestimate of the population at risk and thereby lead to an underestimate of the percent entering prison. (See Limitations of the life table estimates, page 13.) ----------------------------------------------------------------------------------------------------- To calculate the population at risk, estimates of the U.S. resident population on January 1, 1991, were obtained from the U.S. Bureau of the Census. (See U.S. Population Estimates, by Age, Sex, Race, and Hispanic Origin: 1990 to 1995, PPL-41, February 1996.) These published data were adjusted for the undercount in the 1990 decennial census, by age, sex, race, and Hispanic origin, using the 1990 Post Enumeration Survey. (National adjustment data were released by the Population Division, U.S. Bureau of the Census, in May 1996.) The number of persons in State and Federal prison or on parole at yearend 1990 were then subtracted from these adjusted population estimates to obtain the population at risk to first incarceration. Excluded from the population estimates by age, sex, race, and Hispanic origin were those persons known to be incarcerated or previously incarcerated (on parole). ********************************************************* Estimating the number of first admissions to prison ********************************************************* The number of persons admitted to a State or Federal prison for the first time by age were estimated by combining information from the 1991 Survey of Inmates in State Correctional Facilities and the 1991 Survey of Inmates in Federal Correctional Facilities with data from the 1990 and 1991 National Corrections Reporting Program (NCRP). The surveys provide an estimate of the number of persons still in prison who first entered prison during the 12-month period before the surveys were conducted. The NCRP files provide information with which to estimate the number of persons who entered during the 12-month period ending July 15, 1991, and who had already been released prior to the survey dates. The State and Federal inmate surveys were conducted in June, July, and August of 1991. Included in these surveys were detailed questions about the current and past prison sentences of State and Federal inmates. A total of 13,986 inmates in State prisons and 6,572 inmates in Federal prisons were interviewed. (For a description of the sample designs and accuracy of the survey estimates, see Comparing Federal and State Prison Inmates, 1991, NCJ-145864.) These surveys permit identification of those inmates who had entered prison for the first time on any sentence during the 12-month period prior to the conduct of the survey. Based on the midpoint of the survey period, the surveys cover those inmates who entered prison after July 15, 1990, and who were still in prison 12 months later. Estimates by age, sex, race, and Hispanic origin, were then developed using weighting factors derived from the probability of selection in each sample. For each demographic group, the number of first admissions by age were obtained by adding estimates from the State and Federal surveys (Appendix table 3). -------------------------------------------------------------------------------- Appendix table 3. Estimates of the number and rate of first admissions to State or Federal prison, by age, July 16, 1990, to July 15, 1991 First admissions to State or Federal prison during 12 months preceding the 1991 surveys Number in prison Total number Age at first at time admitted during Rate per 100,000 admission of survey/a 12-month period U.S. residents/c (1) (2) (3) 0-13 0 0 0 14 60 75 2 15 415 517 15 16 1717 2138 62 17 3472 4323 124 18 6530 8131 228 19 9638 12001 302 20 12025 14973 352 21 8425 10490 266 22 9035 11250 299 23 7598 9461 254 24 9243 11509 297 25 7386 9197 228 26 7839 9761 233 27 7507 9347 217 28 7388 9199 219 29 6774 8435 188 30 5831 7260 155 31 6560 8168 188 32 5019 6249 143 33 3876 4826 110 34 3990 4968 113 35 4692 5842 134 36 4089 5091 123 37 3799 4730 117 38 3018 3758 98 39 2840 3536 90 40 2690 3349 85 41 2083 2594 71 42 1957 2437 67 43 1425 1774 47 44 1407 1752 51 45 1024 1275 44 46 940 1170 42 47 1063 1324 45 48 1009 1256 45 49 964 1200 45 50 746 929 38 51 367 457 20 52 422 525 23 53 524 652 30 54 490 610 28 55 538 670 32 56-60 1813 2257 ... 61-70 1134 1412 ... 71 or older 411 512 ... ...Not shown, because subsequent calculations were based on rates for single years of age. a/The number of persons admitted to a State or Federal prison for the first time were estimated from information in the 1991 inmate surveys. These surveys, conducted in June, July, and August 1991, provide an estimate by age of the number of inmates still in prison who had been admitted for the first time in the previous 12 months. Not included in the surveys are those who had been admitted during the 12-month period and had already been released. b/ Ratio adjustment factors derived from the 1990-91 NCRP, when multiplied by column (1) provide an estimate of the total number of first admissions. c/Calculated by dividing column (2) by the number of persons at risk to first incarceration (Appendix table 2, column 3) and then multiplying by 100,000. ------------------------------------------------------------------------------------------ The survey estimates, however, understate the total number of first admissions to prisons -- excluded from the surveys are those persons admitted during the 12-month period who had already been released. To correct for this incomplete coverage, adjustment factors were calculated based on data reported in the 1990 and 1991 NCRP collections. (For a description of NCRP reporting criteria, see National Corrections Reporting Program, 1991, NCJ-145861.) The NCRP data provide counts by sex, race, and Hispanic origin of the number of new court commitments to State or Federal prison during the 12-month period ending July 15, 1991. The NCRP data also provide counts of those released before July 15, 1991, among those who had been admitted during the period. For each group, an adjustment factor was computed by dividing the number of new court commitments by the number of new court commitments minus the number of releases (Appendix table 4). -------------------------------------------------------------------------------------------- Appendix table 4. NCRP adjustmentfactors for estimating the total number of first admissions to prison, from July 16, 1990, to July 15, 1991 Total Male Female Total 1.245 1.235 1.37 Non-Hispanic White 1.262 1.251 1.391 Black 1.261 1.250 1.385 Hispanic 1.146 1.141 1.228 Note: Adjustment factors based on new court commitments and first releases as reported in the 1990 and 1991 NCRP for the period July 16, 1990, to July 15, 1991. Data were restricted to 28 States and the Federal Bureau of Prisons, which provided information on admission and release records. Data were further restricted to inmates with a total maximum sentence of more than a year. Each adjustment factor represents the ratio of the total number of new court commitments during the 12-month period to the number of new cour commitments minus the number of releases during the same period. For presentation, the adjustment factors were rounded to 3 decimal places. ----------------------------------------------------------------------------------------------- These factors were then multiplied by the survey estimates to generate the total number of first admissions for each age group. To illustrate the calculation and use of the NCRP adjustment factors, consider the estimate of the number of persons admitted to prison for the first time at age 20. Based on the combined estimates from the State and Federal surveys, a total of 12,025 persons age 20 were first admitted to prison between July 16, 1990, and July 15, 1991. The NCRP data reveal that 19.7% of all new court commitments during the period had been released before the end of the period. By multiplying 12,025 by the adjustment factor 1.245 ( or 1/.803), the actual total number of first admissions was estimated at 14,973. Estimates of the rate of first admissions to State or Federal prison at each age (Appendix table 3, column 3) were calculated by dividing the number of persons admitted to State or Federal prison during the 12 months prior to the 1991 surveys (column 2) by the estimated number at risk to first incarceration in each age group (Appendix table 2, column 3) and then multiplying by 100,000. For example, approximately 14,973 of the estimated 4,253,833 persons at risk were admitted to prison for the first time at age 20 in the 12 months prior to the 1991 surveys -- the equivalent of 352 per 100,000 persons age 20 in the U.S. resident population. ********************************************* Calculating the lifetime chances of going to prison ********************************************* To illustrate the application of life table techniques to the lifetime chances of going to prison, consider the following computations for the hypothetical cohort of 100,000 births: 1. Estimates of the expected number of persons in the original population of 100,000 who will die during each age interval were obtained by multiplying the age-specifc mortality rates (from Appendix table 1) by the number of persons alive and not previously incarcerated at each year of age (Appendix table 5, columns 1 and 2). Appendix table 5. Estimating the lifetime likelihood of going to State or Federal prison Hypothetical population of 100,000 births reduced by mortality and incarceration in each successive year of age Likelihood of going to prison Number alive during rest of life and not Expected Cumulative Number incarcerated Expected number number of expected Age at first at beginning number first first to go to admissionage interval of deaths/a admission admission prison Percent/d (1) (2) (3) (4) (5) (6) 0-13 100,000 1,217 0 0 5,147 5.15 14 98,783 48 2 2 5,147 5.21 15 98,733 66 15 17 5,145 5.21 16 98,652 82 62 79 5,130 5.20 17 98,508 96 122 201 5,068 5.15 18 98,290 103 224 425 4,946 5.03 19 97,963 104 296 721 4,722 4.82 20 97,563 106 343 1,064 4,426 4.54 21 97,114 109 257 1,321 4,083 4.20 22 96,748 109 290 1,611 3,825 3.95 23 96,349 109 245 1,856 3,536 3.67 24 95,995 110 285 2,141 3,291 3.43 25 95,600 108 218 2,359 3,006 3.14 26 95,274 108 221 2,580 2,788 2.93 27 94,945 108 206 2,786 2,567 2.70 28 94,631 112 207 2,993 2,361 2.49 29 94,312 115 178 3,171 2,153 2.28 30 94,019 120 146 3,317 1,976 2.10 31 93,753 125 176 3,493 1,830 1.95 32 93,452 133 133 3,626 1,654 1.77 33 93,186 141 102 3,728 1,520 1.63 34 92,943 151 105 3,833 1,418 1.53 35 92,687 162 124 3,957 1,314 1.42 36 92,401 175 113 4,070 1,190 1.29 37 92,113 185 108 4,178 1,076 1.17 38 91,820 193 90 4,268 969 1.06 39 91,537 202 82 4,350 879 0.96 40 91,253 209 78 4,428 797 0.87 41 90,966 221 64 4,492 719 0.79 42 90,681 231 61 4,553 655 0.72 43 90,389 244 42 4,595 594 0.66 44 90,103 258 46 4,641 552 0.61 45 89,799 274 39 4,680 506 0.56 46 89,486 292 37 4,717 467 0.52 47 89,157 313 40 4,757 429 0.48 48 88,804 341 40 4,797 390 0.44 49 88,423 369 40 4,837 350 0.40 50 88,014 403 33 4,870 309 0.35 51 87,578 439 18 4,888 277 0.32 52 87,121 481 20 4,908 259 0.30 53 86,620 528 25 4,933 239 0.28 54 86,067 575 25 4,958 214 0.25 55 85,467 631 26 4,984 189 0.22 56-60 84,810 4,100 90 5,074 163 0.10 61-70 80,620 14,058 52 5,126 73 0.04 71 or older 66,510 ... 21 5,147 21 0.00 Note: Calculations for ages 0 to 13 and ages 56 or older are based on data for each single year of age and then grouped for presentation. Also for presentation, data in columns 2 and 3 were rounded to produce the totals in column 4. ...Not shown. a/To estimate the number expected to die at each year of age, age-specific mortality rates (from Appendix table 1) were multiplied by the number of persons alive and not previously incarcerated (column 1). b/To estimate the number expected to go to prison at each year of age, age-specific first admission rates (from Appendix table 3, column 3) were multiplied by the number of persons surviving (column 1 minus column 2, above). c/The number expected to go to prison at each age and remaining ages was estimated by subtracting from 5,147 the cumulative number previously incarcerated (column 4, previous age interval). d/Percents were calculated by dividing the number expected to go to prison (column 5) by the number alive and at risk to first incarceration (column 1) and then multiplying by 100%. * For example, among those persons who reach age 20 without having been previously incarcerated, a total of 106 are expected to die before reaching age 21 (that is, 97,563 times the rate of .00109). 2. The number of persons in this hypothetical cohort who are at risk to incarceration during an age interval was then calculated by subtracting the number dying from the number alive and not previously incarcerated at the beginning of the age interval (column 1 minus column 2). * 97,457 of the persons who survived to age 20 without being incarcerated are expected to reach age 21 (not shown). 3. The projected number of persons in the original 100,000 population who will be admitted to prison at each age (column 3) was then obtained by applying the age-specific first admission rates (from Appendix table 3) to the number of persons alive and not previously incarcerated. * Among the 20 years old in the original population who had not previously been incarcerated and were expected to reach their next birthday, 343 are expected to be incarcerated before their 21st birthday (.00352 times 97,457). 4. The number of persons in the original 100,000 population who are expected to be incarcerated by the time they reach a specific age was then calculated by summing the number incarcerated at each age for all younger age groups (column 4). * The cumulative sum of persons age 20 or younger, for instance, is 1,064. This number indicates that of the original 100,000 population, 1,064 (or 1.06%) will be incarcerated before their 21st birthday. 5. An estimate of the lifetime likelihood of incarceration is obtained by summing the number incarcerated for all age groups. As a result, 5,147 persons in this hypothetical population of 100,000 (or 5.15%) are expected to go to prison at least once during their lifetime. 6. These procedures also provide estimates of the likelihood of going to prison in the future for persons not yet incarcerated at any given age (column 6). For each age interval, the percent expected to go to prison in the future was estimated by subtracting from 5,147 the cumulative number incarcerated before reaching the age interval (column 5); then dividing by the number of persons alive and not yet incarcerated at the beginning of that age interval (column 1); and multiplying by 100%. * Among persons who live to age 20 without being incarcerated, the likelihood of going to prison in the future (4.54%) was calculated by subtracting 721 (the cumulative number incarcerated before age 20) from 5,147; dividing by 97,563; and multiplying by 100%. ************************************************ Limitations of the life table estimates ************************************************ 1. The age-specific incarceration rates are based upon the experiences of prisoners interviewed at one point in time, and do not incorporate a forecast of future rates of imprisonment, which may be affected by changes in criminal behavior, changes in law enforcement, and changes in sentencing policies. Life table techniques assume stable mortality and first-incarceration rates. Recent data suggest that rates of first incarceration have remained relatively constant since 1991. 2. All of the calculations are based only on incarceration in State or Federal prison. Prior incarcerations in juvenile facilities and local jails and prior probation supervision were excluded because of the lack of data permitting estimation of the number of first admissions to these forms of correctional supervision. The estimates in this report generally approximate the likelihood of being incarcerated as an adult for a felony. 3. Estimates of the number of first admissions to prison by age, sex, race, and Hispanic origin are subject to sampling and nonsampling errors. Because a sample rather than a complete enumeration of the State and Federal prison population was conducted, the estimated number of first admissions may vary depending on the size of the estimate and the size of the base population for each demographic group. (For estimates of the sampling errors for the 1991 surveys, see Comparing Federal and State Prison Inmates, 1991, NCJ-145864.) Nonsampling error can be attributed to many sources, such as nonresponse, differences in interpretation of questions, recall difficulties, and processing errors. Among inmates, the number of first admissions may be slightly overestimated because of underreporting of criminal histories. The full extent of nonsampling error is unknown. 4. Estimates of the population at risk to first incarceration by age may be overestimated. Although the population was adjusted for those persons in prison or on parole, the total number of previously incarcerated persons in the U.S. resident population is unknown. As a result, the age-specific rates of first incarceration may also be underestimated, leading to an underestimate of the lifetime likelihood of incarceration. (Calculations for population subgroups that experience high rates of incarceration are likely to be most affected.) ----------------------------------------------------------------------- The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. Jan M. Chaiken, Ph.D., is director. BJS Special Reports address a specific topic in depth from one or more datasets that cover many topics. Thomas P. Bonczar and Allen J. Beck, Ph.D., wrote this report. Tom Hester and Priscilla Middleton edited the report. Marilyn Marbrook, assisted by Yvonne Boston, administered final production. March 1997, NCJ-160092This report and many of its data, as well as other reports and statistics, may be found at the Bureau of Justice Statistics Internet World Wide Web site: http://www.ojp.usdoj.gov/bjs/ ----------------------------------------------------------------------- End of file.