U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Special Report Veterans in Prison or Jail January 2000, NCJ 178888 Revised 9/29/00 By Christopher J. Mumola BJS Policy Analyst ----------------------------------------------------------------- This file is text only without graphics and many of the tables. A Zip archive of the tables in this report in spreadsheet format (.wk1) and the full report including tables and graphics in .pdf format are available from: http://www.ojp.usdoj.gov/bjs/abstract/vpj.htm ----------------------------------------------------------------- ----------------------------------------------------------------- Highlights Over 225,000 veterans held in Nation's prisons or jails in 1998 Estimated number of veteran inmates 1985 1998 Prison 100,200 156,400 Local jail 54,400 69,300 Percent of inmates reporting military service Prison 1986 1997 Federal 24.9% 14.5% State 20.2 12.5 1983 1996 Local jail 21.2% 11.7% * 1 in every 6 incarcerated veterans were not honorably discharged from the military. * About 20% of veterans in prison or jail reported seeing combat duty during their military service. * In 1998, an estimated 56,500 Vietnam War-era veterans and 18,500 Persian Gulf War-era veterans were held in State and Federal prisons. Veterans more likely than others to be in prison for a violent offense but less likely to be serving a sentence for drugs State prisoners, 1997 Offense Veterans Nonveterans Violent 55.1% 46.1% Property 20.3 22.2 Drug 14.4 21.5 Public-order 10.2 9.8 Average maximum sentence Violent 264 mo 214 mo Property 123 116 Drug 118 107 Public-order 86 79 * About 35% of veterans in State prison, compared to 20% of non- veterans, were convicted of homicide or sexual assault. * Veterans (30%) were more likely than other State prisoners (23%) to be first-time offenders. * Among violent State prisoners, the average sentence of veterans was 50 months longer than the average of nonveterans. Veterans in State prison reported higher levels of alcohol abuse, lower levels of drug abuse, than other prisoners State prisoners, 1997 Veterans Nonveterans Drug use in month before offense 45.4% 58.1% Alcohol dependent 30.6% 23.6% Identified as mentally ill 19.3% 15.8% Homeless at some time in 12 months before arrest 12.4% 10.3% * Veterans in State prison (26%) were less likely than other State prisoners (34%) to report having used drugs at the time of their offense. * Nearly 60% of veterans in State prison had driven drunk in the past, compared to 45% of other inmates. * About 70% of veterans, compared to 54% of other State prisoners, had been working full-time before arrest. ------------------------------------------------------------------- In 1998 there were an estimated 25,062,400 veterans in the United States population, including 225,700 veterans held in the Nation's prisons and jails. Males comprised 95% of all veterans and 99% of those veterans in prison and jail. Among adult males in 1998, there were 937 incarcerated veterans per 100,000 veteran residents, up from 571 per 100,000 in 1985. Despite the increase, male veterans were incarcerated at less than half the rate of adult male nonveterans (1,971 per 100,000) in 1998. Since 1985 the number of U.S. veterans has declined by nearly 3 million, an average decline of 1% per year. The number of male veterans dropped from 26,962,000 in 1985 to 23,834,100 in 1998, while the number of female veterans rose from 1,044,700 to 1,228,300. The number in prison or jail rose from 154,600 in 1985 to 225,700 in 1998, an increase of 46%. However, during this period the number of nonveteran inmates rose 172%. Veterans accounted for 12% of all inmates in 1998, down from 21% in 1985. These data are based on interviews with national samples of inmates. In responses to the 1997 Survey of Inmates in State and Federal Correctional Facilities and the 1996 Survey of Inmates in Local Jails, inmates provided information on their military service, as well as their criminal history and personal background. Veterans include all persons with prior military service In this BJS Special Report, the term "veteran" is defined as any person who has reported prior service in the United States Armed Forces, regardless of the type of military discharge. This definition differs from that used by the Department of Veterans Affairs (VA) to determine eligibility for various veteran benefit programs, as it includes persons not honorably discharged and a small number of persons awaiting discharge. However, VA counts of veterans in the U.S. resident population and BJS counts of those in prison or jail employ comparable definitions that include all types of discharges. The percentage of veterans in the U.S. population who had not been honorably discharged is unknown. Data from the Defense Manpower Data Center indicate that 9.5% of the 217,600 officers and enlisted persons discharged from the U.S. Armed Forces during fiscal year 1999 were not honorably discharged (including dishonorable, bad conduct, other than honorable, and general discharges). Not honorably discharged veterans are found in incarcerated populations in higher proportions -- 17% of veterans held in State prison, 15% of those in Federal prison, and 14% of those in local jails. Veterans in prison who had not been honorably discharged from military service have more serious criminal and substance abuse histories than other incarcerated veterans. Percent of veterans Percent of veterans in prison, 1997 in local jails, State Federal Type of discharge Honorable/a 83.0% 76.0% 80.7% Not honorable/b 13.8 17.4 14.9 Medical/c 3.2 5.4 3.3 Not discharged 0.0 1.3 1.2 a/Includes general discharges that have been converted to honorable status since leaving military service. b/Includes other general, dishonorable, undesirable, and bad conduct discharges. c/Includes Section 8 discharges. Number of U.S. veterans declining; number held in prison and jail rising In 1998 the veteran population in the United States totaled 25 million, but the number has been steadily declining. The Department of Veterans Affairs, using data from the decennial census, annual discharge data, and mortality rates, reports that the number of male veterans has dropped by more than 3 million since 1985. Veterans accounted for 1 in 3 adult male residents in 1985 but declined to 1 in 4 by 1998. During the same period, the number of female veterans has risen by over 180,000. Overall, the number of veterans has declined nearly 1% per year since 1985. In BJS surveys of jail and prison inmates conducted since the mid-1980's, a declining percentage reported prior military service. Veterans accounted for 13% of State prisoners in 1997 and 12% of jail inmates in 1996, down from 20% in 1986 and 21% in 1983, respectively. While these percentages declined, the overall size of the prison and jail populations rose sharply. Between 1985 and 1998, the number of prison and jail inmates grew by over 1 million. As a result, the number of incarcerated veterans also rose, from 154,600 in 1985 to 225,700 in 1998. Among adult males -- who comprise 95% of all veterans -- the incarceration rate of veterans has remained far lower than that of nonveterans. In 1985 there were 571 incarcerated male veterans per 100,000 male veteran residents, compared to a rate of 965 per 100,000 among male nonveterans. In 1998 the rate of male veterans had increased to 937 but was still less than half that of nonveterans (1,971). Number of veterans in U.S. resident population Year Total Male Female 1985 28,007,000 26,962,000 1,045,000 1990 27,183,700 26,089,700 1,094,000 1995 26,061,800 24,864,000 1,197,800 1998 25,062,400 23,834,100 1,228,300 Percent change, 1985-98 -10.5% -11.6% 17.5% Annual average change, 1985-98 -0.9% -0.9% 1.3% Source: K. Sorensen and T. Feild, Statistical Brief: Projections of the U.S. Veteran Population: 1990 to 2010, U.S. Department of Veterans Affairs, May 1994. Percent of inmates reporting military service in BJS surveys Prison 1986 1991 1997 Federal -- 20.2% 14.5% State 20.20 16.4 12.5 1983 1989 1996 Local jail 21.2% 15.5% 11.7% Estimated number of veteran inmates Year Prison Jail Total 1985 100,200 54,400 154,600 1990 124,200 62,800 187,000 1995 136,600 59,300 195,900 1998 156,400 69,300 225,700 Prison and jail inmates per 100,000 adult U.S. residents 1985 1998 Veteran Nonveteran Veteran Nonveteran Male 571 965 937 1,971 Female 57 45 187 138 Nearly 60% of incarcerated veterans had served in the Army Veterans in prison and jail had served in all branches of the U.S. Armed Forces, including the U.S. Coast Guard and National Guard. Although the U.S. Army has accounted for less than 40% of U.S. military personnel since 1971, Army veterans made up a majority of veterans incarcerated in both State (59%) and Federal (58%) prison, as well as local jails (57%). The Air Force accounted for an average of 29% of U.S. military personnel between 1960 and 1996, but Air Force veterans made up 9% of veterans in State prison, 12% of those in Federal prison, and 10% of veterans in local jails. (For details on military personnel data see Methodology.) Veterans' mean length of military service was about the same among jail inmates (44 months) and State (45 months) and Federal (48 months) prisoners. Veterans held in Federal prison were the least likely to report less than a year of military service (10%), followed by veterans in local jails (14%) and State prison (15%). 1 in 5 incarcerated veterans saw combat duty during their service About equal percentages of veterans in State and Federal prisons (20% of each) and local jails (21%) reported seeing combat duty during their military service. A larger percentage of the veterans in prison (50% of State and 61% of Federal) had served during a period of wartime. Percent of veterans in prison State Federal 1997 1991 1997 1991 Time of military service* Peacetime 50.3% 50.9% 39.0% 37.1% Wartime 49.7 49.1 61.0 62.9 World War II era (1941-45) 1.3 1.4 1.2 4.5 Korean Conflict era (1950-53 2.9 5.7 4.6 8.8 Vietnam War era (1964-73) 35.3 42.1 43.3 49.0 Persian Gulf War era (1990-91) 11.7 1.3 12.9 1.6 *Veterans may have served during more than one period of wartime. The reported wartime experience of incarcerated veterans changed after 1991, reflecting the end of the Persian Gulf War and the effects of aging among veterans from previous wars and military conflicts. In 1997 Vietnam-era veterans were still the most common wartime veterans, accounting for 35% of veterans in State prison and 43% of those in Federal prison. This represented a decline from 1991, when Vietnam-era veterans made up 42% and 49% of State and Federal prison veterans, respectively. In 1997 Persian Gulf War-era veterans made up 12% of State and 13% of Federal prison veterans. Veterans who served during the Korean Conflict era accounted for 3% of State and 5% of Federal prison veterans, down from 6% and 9%, respectively in 1991. One percent of both State and Federal prison veterans reported military service during World War II. A majority of veterans in State prison, but less than a third of nonveteran inmates, were white Because nearly 95% of veterans are men, virtually all incarcerated veterans in 1997 were men. Among nonveterans women accounted for 7% of State prisoners, 8% of Federal prisoners, and 11% of local jail inmates. In contrast, women accounted for 1% of the veterans in either State or Federal prison, and 2% of veterans in local jails. Veterans were also more likely than other inmates to be white non-Hispanic. Among State prisoners, over half (53%) of veterans were white non-Hispanics, compared to nearly a third (31%) of nonveterans. Among Federal prisoners the percentage of veterans who were white (50%) was nearly double that of nonveterans (26%). In both prisons and jails, black non- Hispanics made up the largest share of nonveteran inmates. Blacks made up 48% of the nonveterans in State prison, compared to 37% of veterans. The percentage of black inmates varied less by veteran status in Federal prison (38% of nonveterans, 35% of veterans) and local jails (41% of nonveterans and 37% of veterans). The percentage of Hispanics was sharply lower among veterans than nonveterans. In both State and Federal prison, as well as local jails, the percentage of Hispanics among veterans was less than half that of nonveterans. Federal prisoners displayed the largest difference, with Hispanics accounting for 30% of nonveterans and 12% of veterans. Among State prisoners, the median age of veterans was 10 years higher than that of nonveterans Veterans were much older than other prison and jail inmates. Among both State (41 years old) and Federal (43) prisoners, the median age of veterans was over age 40, while that of nonveterans was age 31 and 35, respectively. Nonveterans (63%) were almost 3 times as likely as veterans (23%) to be under age 35. In contrast, 32% of veterans in State prison were at least age 45, compared to 10% of nonveterans. Veterans were also more likely to have been married than nonveterans. Overall, 71% of veterans and 39% of nonveterans in State prison said they had been married at some time in their lives. However, the percentages of veterans (22%) and nonveterans (16%) who said they were currently married were much more similar. Veterans reported much higher levels of completed education than other inmates. Among State prisoners, veterans (32%) were about 3 times more likely than nonveterans (11%) to have attended college. Even higher levels of college attendance were reported by veterans in local jail (37%) and Federal prison (42%). In contrast, 44% of nonveterans in State prison reported that they had neither completed high school nor attained a GED, compared to 12% of veterans. Veteran inmates were more likely to be violent offenders, less likely to be drug offenders A majority (55%) of the veterans in State prison were sentenced for a violent offense, compared to 46% of nonveterans. Among Federal prisoners fewer than a quarter of veterans (22%) were violent offenders, but the percentage was still lower among nonveterans (13%). Veterans in both State (14%) and Federal (51%) prison were less likely to be drug offenders than nonveterans (22% and 65%, respectively). Among State prisoners veterans (18%) were over twice as likely as nonveterans (7%) to be sentenced for a sexual assault, including rape. While veterans also had a slightly higher percentage of homicide offenders (17%) than nonveterans (13%), they were no more likely than other State prisoners to be sentenced for other types of violent crime. In local jails, sexual assault was the only violent offense for which a greater share of veterans (6%) than nonveterans (3%) was held. (For further data on sexual assaults and military personnel, see box below.) About equal percentages of veterans and nonveterans in State prison were sentenced for property and public-order offenses. Among Federal prisoners, veterans had a higher percentage of fraud offenders (9%) than nonveterans (4%), but were otherwise equally likely to be convicted of property or public-order offenses. Among local jail inmates veterans had a similar percentage of both violent (27%) and property (25%) offenders as nonveterans (26% and 27%, respectively). Veterans (17%) in local jails were less likely to be drug offenders than nonveterans (23%), but more likely to be public-order offenders (31% compared to 23%). Sex offenders accounted for 1 in 3 prisoners held in military correctional facilities at yearend 1997 The high prevalence of sex offenders among incarcerated veterans is also reflected in data collected on persons still on active military duty. Since 1994, the U.S. Department of Defense (DoD) Corrections Council has annually reported data to the Bureau of Justice Statistics regarding persons incarcerated in military correctional facilities. These facilities, administered by the DoD, hold inmates whose offenses fell under the Uniform Code of Military Justice rather than civilian Federal, State, or local jurisdiction. (See Methodology.) In 1997 sexual assault was the most common offense for which inmates were held in military prisons, accounting for nearly a third (31%) of all military prisoners. Sexual assaults have also been the most common offense in each year since reporting began. All other violent offenses accounted for 21% of military prisoners, followed by drug (20%), property (17%), and public-order (2%) offenses. Another 8% of military prisoners were sentenced for military offenses, including AWOL, desertion, and insubordination. Percent of prisoners Type of offense in military facilities, 1997 Violent 52.2% Homicide 11.1 Sexual assault 30.6 Robbery 1.7 Assault 8.6 Other violent 0.3 Property 17.2 Drug 20.2 Public-order 1.7 Military offenses 7.6 Other 1.2 Number of prisoners 2,466 Note: For more data on military correctional facilities and inmates, see Correctional Populations in the United States, 1997, NCJ 177613, March 2000. Among violent State prisoners, over two-thirds of veterans knew their victims Among violent State prisoners, veterans (69%) were more likely than nonveterans (51%) to have known their victims. Veterans (22%) were twice as likely as other violent State prisoners (9%) to have victimized a relative (excluding spouses). Veterans (11%) were as likely as nonveterans (9%) to report victimizing an intimate, such as a spouse, girlfriend or boyfriend. A slightly higher percentage of veterans (32%) than nonveterans (27%) reported victimizing friends and acquaintances. About two-thirds of violent veterans, and half of other violent State prisoners reported victimizing a female. A majority (54%) of veterans, compared to 36% of nonveterans, said that females were the only victims of their violent offense. About a third (36%) of veterans, but a majority of nonveterans (53%) said that only men had been the victims of their violent crime. A tenth of both groups reported both male and female victims. Fewer veterans (34%) than nonveterans (44%) reported using a weapon during their violent offense. Veterans (36%) were more likely than nonveterans (20%) to have victimized a juvenile (under 18 years old). In particular, twice as many veterans (20%) as nonveterans (9%) reported that their youngest victim was a child, age 12 or younger. Combat veterans in prison were no more likely to be violent offenders than other veterans Among veterans in State prison, those who reported seeing combat duty during their military service were no more likely to be violent offenders than other veterans (55% of both). Overall, the offense distribution of combat veterans showed few differences from that of noncombat veterans. Combat veterans were also as likely to be property (19%) or drug (14%) offenders as noncombat veterans (21% and 15%, respectively). Veterans who had served in the military during the Vietnam War era (1964-1973) were more likely (59%) than other veterans in State prison (52%) to be violent offenders. In particular, higher percentages of Vietnam-era veterans were sentenced for homicide (20%) compared to other veterans (14%). Vietnam-era veterans (16%) were less likely than other veterans (23%) to be property offenders. About equal percentages of Vietnam-era veterans were held for drug (15%) or public-order (10%) offenses as other veterans (14% for drug and 10% for public-order offenses). Among State prisoners, veterans had less extensive criminal histories than nonveterans In State prisons veterans had less extensive criminal histories than other inmates. At the time of their arrest, veterans were less likely than other State prisoners to be on probation or parole supervision, bail, pretrial release or other forms of conditional release from custody. About a third of veterans (36%) and half of nonveterans (49%) reported some type of criminal justice status when arrested. Veterans in State prison were less likely than nonveterans to be recidivists (those inmates with a history of probation or incarceration sentences prior to their current confinement). Nearly a third of veterans were first-time offenders, while about a quarter of other prisoners did not report a prior sentence. Veterans in State prison were also less likely to have lengthy criminal records. Offenders with at least three prior probation or incarceration sentences accounted for 37% of veterans, and 44% of the nonveterans in State prison. In both local jails and Federal prisons, the criminal histories of veterans and nonveterans were more similar. An equal percentage (27%) of both veterans and other local jail inmates were first-time offenders, without prior sentences. Veterans (45%) were also as likely as other jail inmates (44%) to report three or more prior sentences. In Federal prison veterans (60%) were as likely as nonveterans (61%) to report a prior sentence. About 30% of both veterans and other Federal prisoners reported three or more prior probation or incarceration sentences. Criminal histories of combat veterans were less serious than other veterans in prison Veterans who had seen combat during their military service (29%) were less likely to have a criminal justice status when arrested than other veterans (38%). Forty percent of combat veterans were first-time offenders, compared to 27% of noncombat veterans. Combat veterans (39%) were also less likely to be violent recidivists than other veterans (46%). Overall, the criminal histories reported by Vietnam-era veterans in State prison were similar to those of other veterans. Similar percentages of Vietnam-era veterans (29%) and other veterans (31%) were first-time offenders. Vietnam-era veterans (29%) were less likely than other veterans in State prison (40%) to have a criminal justice status at the time of their arrest. Veterans' sentences averaged 3 years longer than other State prisoners On average veterans were serving longer prison sentences than other inmates. Among State prisoners, the mean maximum sentence length of veterans (191 months) was 3 years longer than that of nonveterans (152 months). Veterans (26%) were more likely to report sentences of 20 or more years than nonveterans (20%), as well as life or death sentences (11% of veterans, 7% of nonveterans). These differences may reflect the higher percentage of violent offenders among veterans. Since violent crimes carry longer sentences, veterans should be expected to have longer sentences than nonveterans. In addition, veterans reported longer sentences than other prisoners with the same offenses. Mean maximum sentence length, State prisoners, 1997 Type of offense Veterans Nonveterans Violent 264 mos. 214 mos. Homicide 406 328 Sexual assault 239 231 Robbery 267 193 Assault 151 141 Property 123 116 Drug 118 107 Public-order 86 79 Among violent offenders in State prison, the mean sentence length reported by veterans (264 months) was about 4 years longer than that of nonveterans (214 months). In particular, veterans convicted of homicide (406 months) and robbery (267 months) reported mean sentence lengths that were over 6 years longer than those of nonveterans (328 months and 193 months, respectively). Among nonviolent offenders the sentences of veterans and nonveterans were more similar. The mean drug offense sentence for veterans (118 months) was about a year longer than that of nonveterans (107 months). For both property and public-order offenders, the mean sentences reported by veterans (123 and 86 months, respectively) were 7 months longer than those of nonveterans (116 and 79 months, respectively). Among Federal prisoners, there was less variation in sentence lengths. On average the sentences reported by veterans (135 months) were less than a year longer than those reported by nonveterans (124 months). The percentages of veterans reporting sentences of 20 or more years (15%), as well as life or death sentences (5%), were similar to nonveterans (13% and 3%, respectively). The total time expected to be served by veterans (109 months) and nonveterans (103 months) varied by only 6 months. Jail sentences showed no variation by veteran status. The mean sentence lengths reported by veterans (13 months) and nonveterans (12 months) were nearly equal. Veterans (8 months) and nonveterans (9 months) also expected to serve equal amounts of jail time prior to release. Veterans were less likely than other inmates to report using drugs in the month before their current offense Among State prisoners, veterans (79%) were less likely than nonveterans (84%) to report past drug use. The difference in drug use in the month before the offense was greater (45% of veterans, 58% of nonveterans). In particular, fewer veterans (27%) than nonveterans (41%) reported marijuana use during this period. Veterans (26%) were also less likely than nonveterans (34%) to have been under the influence of drugs while committing their offense. In Federal prisons, veterans were as likely as other inmates (73% of both) to report past drug use. Veterans (40%) were less likely than nonveterans (46%) to have used drugs in the month before the offense. In local jails, veterans (81%) also reported levels of prior drug use similar to nonveterans (83%), but lower levels (44%) of drug use in the month prior to the offense than nonveterans (50%). At the time of the offense, veterans in both Federal prison (21%) and local jail (29%) were about as likely as nonveterans (23% Federal, 32% local jails) to have been under the influence of drugs. Among both State and Federal prisoners, as well as local jail inmates, the percentage of veterans who reported prior drug use was equal to, or lower than the percentage of nonveterans. However, in all three groups, the percentage of veterans who reported past use of specific drug types was often higher than that of nonveterans. For example, while veterans (81%) were about as likely as other jail inmates (83%) to report any prior drug use, veterans were more likely to have used heroin or other opiates (31%), depressants (38%), stimulants (42%), and hallucinogens (38%) than nonveterans (23%, 29%, 32%, and 31%, respectively). These findings suggest that among past drug users in prisons and jails, veterans were more likely than nonveterans to have used multiple types of drugs. Veterans were also more likely to have used intravenous (IV) drugs than nonveterans. Among both State and Federal prisoners, as well as jail inmates, a higher percentage of veterans than nonveterans reported prior IV drug use. Veterans in State prison (25%) or local jails (24%) were the most likely to report prior IV drug use, followed by veterans in Federal prison (17%). Among nonveterans, State prisoners (19%) reported the highest level of IV drug use, followed by local jail (16%) and Federal inmates (11%). Ever used a needle to inject drugs for non-medical reasons State prisoners Veterans 24.5% Nonveterans 19.4 Federal prisoners Veterans 17.1% Nonveterans 11.5 Local jail inmates Veterans 23.9% Nonveterans 16.2 Less than a third of incarcerated combat veterans used drugs in the month before their offense Veterans in State prison who saw combat during their military service were less likely to report prior drug use (69%) than noncombat veterans (82%). Fewer than a third (30%) of combat veterans reported drug use in the month before their offense, compared to half (49%) of other veterans. About a quarter of both combat (23%) and other veterans (25%) reported prior use of IV drugs. In their reports of prior drug abuse, Vietnam-era veterans varied little from other veterans in State prison. Vietnam-era veterans (77%) were about as likely as other veterans (80%) to report ever using drugs, and slightly less likely to report drug use in the month before the offense (41%, compared to 47% of other veterans). However, a third (32%) of Vietnam-era veterans reported prior IV drug use. Other veterans in State prison (20%) were no more likely than nonveterans (19%) to report prior IV drug use. Reports of prior alcohol abuse and dependence were higher among veteran inmates Veterans were more likely than other prison and jail inmates to report a variety of alcohol abuse indicators. According to the CAGE diagnostic instrument, 31% of veterans in State prison and 35% of those in local jails reported experiences that are consistent with a history of alcohol abuse or dependence. Nonveterans in both State prison (24%) and local jail (25%) were less likely to report these signs of alcohol dependence. Overall, fewer Federal prisoners were alcohol dependent, but reports were higher among veterans (21%) than nonveterans (15%). CAGE is an acronym for the four questions used by the diagnostic instrument to assess alcohol dependence and abuse. Respondents are asked if they have ever attempted to (C)ut back on drinking; ever been (A)nnoyed by others' criticism of their drinking; ever felt (G)uilty about their drinking; and ever needed a drink first thing in the morning as an (E)ye opener or to steady the nerves. The CAGE instrument determines a person's likelihood of alcohol abuse by the number of positive responses to these questions. (For additional information on the CAGE instrument, see Methodology.) A consistently higher percentage of veterans also reported engaging in troubled behaviors associated with past alcohol abuse. Veterans in both State (59%) and Federal (55%) prison were more likely to report driving drunk in the past than nonveterans (45% and 41%, respectively). About half (48%) of veterans, but less than a third (30%) of nonveterans in local jails reported a previous alcohol- related arrest. Veterans in local jails were also more likely to report alcohol-related physical fights (45%) than nonveterans (35%). Prior binge drinking incidents -- consuming as much as a fifth of liquor in 1 day -- were reported by 49% of veterans, compared to 39% of other jail inmates. According to the CAGE diagnostic questionnaire, similar percentages of Vietnam-era (33%) and other veterans (30%) reported signs of alcohol dependence. Vietnam-era veterans were also equally likely to report alcohol-related physical fights (41%) and binge-drinking incidents (46%) as other veterans (40% and 44%, respectively). Drunk driving reports were slightly higher among Vietnam-era veterans (63%) than other veterans (57%) in State prison. Similar percentages of combat (28%) and noncombat (31%) veterans reported signs of alcohol dependence. Combat veterans in State prison were about as likely to report a prior alcohol-related arrest (35%) or a binge-drinking experience (42%) as other veterans (38% and 45%, respectively). Incarcerated veterans and nonveterans reported equal participation in substance abuse treatment Veterans' substance abuse treatment histories were similiar to those of other inmates. Since their admission, veterans in both State (34%) and Federal (30%) prisons reported similar levels of participation in substance abuse treatment or programs as nonveterans (32% and 28%, respectively). Veterans in local jails ]also reported the same level of participation as other inmates (10% of both) in substance abuse treatment or programs since admission. Veterans received the same types of treatment as nonveterans. Among State prisoners 12% of both groups said they had received professional treatment, such as a residential facility or counseling, while similar percentages (29% of veterans, 27% of nonveterans) reported taking part in programs such as self-help groups and education programs. Veterans (59%) were also as likely as other State prisoners (56%) to report participation in substance abuse treatment or other programs at some time in their past. Incarcerated veterans were more likely to report a mental illness From responses to survey questions regarding their mental health, a higher percentage of veterans reported a mental illness than nonveteran inmates. Inmates were considered to be mentally ill if they met one of the following two criteria: they reported a current mental or emotional condition, or they reported an overnight stay in a mental hospital or treatment program. The varying rates of mental illness by veteran status were most evident in the local jail and Federal prison populations. Among jail inmates, 25% of veterans were identified as mentally ill, compared to 15% of nonveterans. Reports of mental illness were much lower in Federal prison, but the percentage of veterans reporting an indicator of mental illness (13%) was about twice that of nonveterans (6%). In State prison, the reported rates of mental illness among veterans (19%) and nonveterans (16%) were more similar. Among State prisoners, veterans who had seen combat duty during their service (22%) reported similar rates of mental illness as noncombat veterans (19%). Veterans who served during the Vietnam War era were also as likely as other veterans (19% of both) to report mental illness. Percent of veterans in State prison reporting mental illness, 1997 Vietnam-era service Yes 19.3% No 19.1 Served in combat duty Yes 21.9% No 18.6 Reported levels of employment and income were higher among veteran inmates Prior to their arrest, veteran inmates were more likely to be employed than nonveterans. Veterans in both State prisons (78%) and local jails (72%) reported higher levels of employment in the month before their arrest than nonveterans (67% and 63%, respectively). In particular, veterans (69% of State, 59% of local) were more likely to be working full-time prior to their arrest than nonveterans (54% of State, 48% of local). Among Federal prisoners, 72% of both groups had been employed prior to their arrest, but a greater share of veterans (65%) were employed full-time than nonveterans (59%). The personal income of veterans in the month before their arrest was also higher than that of other inmates. Among State prisoners, 53% of veterans reported an income of at least $1,000 in the month prior to arrest, compared to 41% of nonveterans. About half of veterans in local jails, compared to nearly a third of nonveterans, reported an income of at least $1,000. Veterans in both State prison (13%) and local jail (16%) were also less likely to report the lowest incomes (under $200) than nonveterans (20% and 25%, respectively). While overall veterans reported higher incomes, the percentage of inmates reporting the highest incomes (at least $5,000) was nearly equal for veterans (8% of State, 16% of Federal, and 4% of local) and nonveterans (8%, 14%, and 3%, respectively). Incarcerated veterans were as likely as nonveterans to have been homeless when arrested In both State and Federal prisons, as well as local jails, veterans and nonveterans reported living in similar types of housing prior to their arrests. Among State prisoners, about half of both veterans (52%) and nonveterans (53%) reported living in a house, and about a third of both (30% of veterans, 34% of nonveterans) said they had lived in an apartment. Similar percentages of veterans and other State prisoners reported living in each type of housing prior to their arrest, including other correctional facilities (2% of each) and group homes or institutions (1% of each). Veterans were no more likely to be homeless at the time of their arrest than other inmates. Among both State (2% of each) and Federal (1% of each) prisoners, equal percentages of veterans and nonveterans reported living on the street or in a homeless shelter prior to their arrest. The percentage of veterans in prison (12% of State and 6% of Federal) reporting homelessness in the year prior to arrest was also similar to that of nonveterans (10% of State and 4% of Federal). Veterans who had not been honorably discharged had more serious criminal and substance abuse histories In their current offenses, honorably and not honorably discharged veterans in State prison showed few differences. For example, nearly equal percentages of honorably (55%) and not honorably (56%) discharged veterans were serving a sentence for a violent offense. In addition, 35% of both honorable and not honorable discharges were on some type of criminal justice status at the time of their current arrest. Among State prisoners, veterans not honorably discharged reported more serious criminal histories than honorably discharged veterans. Violent recidivists accounted for 52% of veterans not honorably discharged, compared to 42% of honorable discharges. A higher percentage of veterans not honorably discharged (43%) also reported at least three prior sentences, compared to those with honorable discharges (36%). Incarcerated veterans not honorably discharged also reported higher levels of prior substance abuse. A majority of these veterans (57%) reported drug use in the month before their offense, compared to 43% of honorable discharges. About a third (32%) of veterans not honorably discharged, but nearly a quarter (23%) of honorable discharges, reported prior intravenous drug use. Veterans not honorably discharged also reported a higher incidence of alcohol dependence (36%) than honorable discharges (30%). While honorably discharged veterans reported lower levels of substance abuse, they did report a slightly higher incidence of mental illness (19%) than veterans not honorably discharged (16%). Percent of veterans in State prison, 1997 Discharge type Honorable Not honorable/a Current offense Violent 54.7% 56.5% Property 20.3 19.7 Drug 14.9 11.8 Public-order 10.1 11.6 Status at time of current arrest None 64.6% 64.7% Status 35.4 35.3 On parole 20.4 19.0 On probation 14.4 16.3 Other 0.6 0.0 Criminal history None 31.8% 25.4% Priors 68.2 74.6 Violent recidivists/b 42.3 52.1 Drug recidivists only 2.0 2.8 Other recidivists/c 24.0 19.7 Number of prior probation or incarceration sentences 0 31.8% 25.4% 1 17.2 18.3 2 15.1 13.8 3-5 21.4 25.0 6-10 9.2 12.4 11 or more 5.3 5.1 Prior drug use Ever used drugs 77.1% 86.4% Used IV drugs 22.5 31.6 Used drugs in month before offense 42.9 57.5 Prior alcohol abuse History of alcohol dependence 30.1% 35.9% Ever drove drunk 59.3 60.7 Ever engaged in binge drinking/d 43.8 48.9 Identified as mentally ille 18.8% 15.8% Number of prisoners 98,457 22,338 a/Excludes medical discharges and those veterans who have not yet been discharged. b/Recidivists with at least one current or past violent offense. c/Includes recidivists with unknown offenses. d/Binge drinking is defined as having consumed as much as a fifth of liquor in a single day. e/Reported either a mental or emotional condition or an overnight stay in a mental hospital or program. Methodology Data in this report are based on personal interviews conducted through the 1997 Survey of Inmates in State and Federal Correctional Facilities and the 1996 Survey of Inmates in Local Jails. Detailed descriptions of the methodology and sample design of each survey can be found in the following: Substance Abuse and Treatment, State and Federal Prisoners, 1997 (NCJ 172871); and Profile of Jail Inmates, 1996 (NCJ 164629). Accuracy of the estimates The accuracy of the estimates in this report depends on two types of error: sampling and nonsampling. Sampling error is the variation that may occur by chance because a sample rather a complete enumeration of the population was conducted. Nonsampling error can be attributed to many sources, such as nonresponses, differences in the interpretation of questions among inmates, recall difficulties, and processing errors. 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 for selected characteristics have been calculated for each survey (see appendix tables). These standard errors may be used to construct confidence intervals around percentages. For example, the 95% confidence interval around the percentage of veterans in State prison who were identified as mentally ill is approximately 19.3% plus or minus 1.96 times 1.22% (or 21.7% to 16.9%). These standard errors may also be used to test the statistical significance of the difference between two sample statistics by pooling the standard errors of the two sample estimates. For example, the standard error of the difference between veterans and other State prisoners who used intravenous (IV) drugs would be 1.40% (or the square root of the sum of the squared standard errors for each group). The 95% confidence interval around the difference would be 1.96 times 1.40 (or 2.74%). Since the difference of 5.1% (24.5% minus 19.4%) is greater than 2.74%, the difference would be considered statistically significant. Estimates of the veteran population This report includes incarceration rates for veterans and nonveterans in the U.S. population. These rates were calculated using projections of the veteran population that were generated by the U.S. Department of Veteran Affairs' National Center for Veteran Analysis and Statistics. These veteran population projections are based on 1980 and 1990 decennial census data on prior military service and discharge data from the U.S. Department of Defense (DoD). For more information on the veteran population, see K. Sorensen and T. Feild, Statistical Brief: Projections of the U.S. Veteran Population: 1990 to 2010, U.S. Department of Veterans Affairs, May 1994. Historical statistics on active duty military personnel The DoD annually reports the number of active duty military personnel in each branch of service, as well as the DoD total. For this report these historical data were used to calculate the percentage of active duty military personnel attributed to each branch of service at the end of the fiscal year, 1950 to 1998. These DoD data are available for the years 1950 to 1995 in Table 580, Statistical Abstract of the United States, 1998, p. 364. Data for later years were obtained from the following DoD website: http://web1. whs.osd.mil/mmid/military/history/ milhist.htm Military prison statistics In 1994 the DoD Corrections Council established an annual military confinement report. The council, comprised of representatives from each branch of military service, adopted a standardized questionnaire with a common set of definitions. With the Corrections Council, the Bureau of Justice Statistics produced a series of tables that provide a unified profile of persons in the custody of U.S. military authorities. (For more information, see Correctional Populations in the United States, 1995, NCJ 163916 and Correctional Populations in the United States, 1996, NCJ 170013.) CAGE diagnostic instrument for alcohol dependence The CAGE questionnaire is a diagnostic instrument for detecting a person's history of alcohol abuse or dependence. In a clinical test involving hospital admissions, three or more positive responses to the four CAGE questions carried a .99 predictive value for alcohol abuse or dependence. For data on the relationship of CAGE to other alcohol abuse indicators, see Substance Abuse and Treatment, State and Federal Prisoners, 1997, (NCJ 172871). For further information on the predictive values of the CAGE instrument, see the article "Screening for Alcohol Abuse Using the CAGE Questionnaire" by B. Bush and others, The American Journal of Medicine, Vol. 82, February 1987, pp.231-35. 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. Christopher J. Mumola wrote this report under the supervision of Allen J. Beck. Thomas P. Bonczar, Laura M. Maruschak and Todd D. Minton provided statistical assistance. Tom Hester edited the report. Jayne Robinson prepared the report for publication. January 2000, NCJ 178888 End of file ih 01/18/00 Revised 9/29/00 ih