Criminal justice involvement and incarceration among American military veterans, especially those who have served in combat, have long been of national concern. There was much expectation that Vietnam veterans would be at high risk of incarceration because of combat trauma, but recent research suggests that premilitary factors are more important than combat exposure in predicting antisocial behavior and incarceration (1,2).
More recently, stories have appeared in the press about criminal behavior and incarceration of veterans who have served in the Middle East in Operations Enduring Freedom (OEF), Iraqi Freedom (OIF), and New Dawn (OND) (3,4). However, the risk of incarceration among this new generation of veterans compared with the general population of veterans has yet to be studied, and so far there is no empirical research on the general characteristics, potential risk factors, and crimes of incarcerated OEF/OIF/OND veterans.
According to a recent report by the U.S. Department of Justice (5), about 10% of prison inmates are veterans. The incarceration rate of veterans (630 per 100,000) is less than half that of nonveterans (1,390 per 100,000), a circumstance that may be explained largely by age, given that in the general population, veterans are considerably older than nonveterans. The report also indicated that like the veteran population as a whole, a majority of incarcerated veterans served during a wartime period; however, only 20%–26% reported seeing combat duty. Yet data about the characteristics of incarcerated OEF/OIF/OND veterans, who are likely to be younger and more likely to have been exposed to war trauma, are not available. Examining these characteristics is important because they may have implications for prevention of incarceration, for treatment, and for delivery of prisoner reentry services among OEF/OIF/OND veterans.
Studies of earlier generations of veterans found that veterans with histories of incarceration have an increased risk of psychiatric problems, substance abuse and dependence, and homelessness compared with veterans with no incarceration history (6,7). Examination of the association of incarceration and homelessness among both veterans (6) and nonveterans (8) suggested that a cluster of nonmilitary-related factors, including sociodemographic characteristics, substance abuse and dependence, and mental illness, may be key risk factors for both conditions (5).
In 2007, the Veterans Health Administration (VHA) created a systemwide outreach program called Health Care for Reentry Veterans (HCRV) to facilitate connection with VHA services among incarcerated veterans upon their release from state and federal prisons. The HCRV program was launched under the umbrella of the Veterans Affairs (VA) homeless services with the intention of preventing both reincarceration and future homelessness.
This study represents the first examination of veterans served by the HCRV program. It compared the risk of incarceration of OEF/OIF/OND veterans and other veterans in the general population stratified by age, gender, and race. It also compared the characteristics of incarcerated OEF/OIF/OND veterans and other incarcerated veterans to identify factors that may have put the OEF/OIF/OND veterans at risk for incarceration.
The HCRV program was established to assist incarcerated veterans in accessing VHA services upon their release into the community with the ultimate goal of preventing both reincarceration and future homelessness. Each HCRV program consists of specialists who provide outreach to veterans in state and federal prisons. The specialists partner with state and federal correctional staff to meet with incarcerated veterans who are close to their release date to conduct prerelease assessments. They also facilitate postrelease linkages and provide short-term clinical management after release.
HCRV specialists, mostly social workers, help incarcerated veterans develop plans to connect to pension and compensation benefits as well as to medical and mental health services and to vocational and residential services. Contacts with veterans while they are incarcerated are limited to assessment and planning for postrelease treatment, and no formal VHA medical services are delivered in the incarceration setting. Thus HCRV specialists focus on initiating contact with incarcerated veterans to help them access VHA services after their release.
Because HCRV does not identify or target veterans with particular health problems or who served in a particular military service area or era, the data available from the program may be representative of the incarcerated veteran population, having no obvious tendency to under- or overrepresent certain groups. However, the data are from only one national, albeit geographically representative, program, so the extent to which they are completely representative of incarcerated veterans on all variables is unknown.
Administrative national data summarizing outreach assessments by the HCRV program of 30,968 veterans contacted from October 2007 to April 2011 were obtained for analysis.
Population data for a weighted sample of nonincarcerated veterans from the 2010 National Survey of Veterans (9) were used to compare the odds of being an OEF/OIF/OND veteran in age-race cohorts of the incarcerated sample with those of being an OEF/OIF/OND veteran among domiciled veterans nationally. Odds ratios comparing these samples were used to estimate and compare the risk of incarceration among various strata of the population of OEF/OIF/OND veterans and other veterans.
The 2010 National Survey of Veterans was a comprehensive nationwide mail survey that collected information about veteran characteristics and health service use. A total of 8,710 veterans participated, and the data were weighted to incorporate the probability of selection and survey nonresponse and were poststratified to known population totals to be representative of the entire noninstitutionalized veteran population.
Administrative data were based on information obtained by HCRV specialists in face-to-face interviews of incarcerated veterans. The specialists used a structured assessment form to collect information about sociodemographic characteristics, criminal history, clinical status, homelessness history, and service needs.
Information was collected about sociodemographic characteristics including age, gender, ethnicity, marital status, military history, and combat exposure. Service in OEF/OIF/OND was defined as serving in theaters of operations in Iraq or Afghanistan. Combat exposure was defined as receiving hostile or friendly fire in a combat zone.
The offense related to the current incarceration was classified into six categories: violent offenses (for example, murder, manslaughter, assault, or robbery), property offenses (for example, burglary, motor vehicle theft, stolen property, arson, or vandalism), drug offenses (for example, possession or trafficking), public order offenses (for example, weapons offense, prostitution, public intoxication, or driving under the influence), a probation or parole violation, or other or unspecified offenses. Clients were also asked whether they had been drinking alcohol or using drugs at the time of the offense.
The expected length of the client’s current incarceration, which included any time in prison, was calculated from information provided by the clients about their release date.
Clients were asked the age at which they were first arrested and how many times they had been arrested in their lifetime before their most current incarceration.
Clients were asked whether they had any serious medical problems. Psychiatric diagnoses were made by the HCRV specialist on the basis of observations and assessment and client’s self-reported history.
Clients were asked whether they had used any VA services in the past six months and if they were interested and were willing to participate in each of five different VA services: psychiatric or substance abuse treatment, medical services, residential treatment services (including the grant and per diem program), social-vocational assistance (including the compensated work therapy program and the incentive therapy program), and case management services.
Length of current homelessness was defined as the time between last being housed for 30 days or more, according to the client’s report, and the current incarceration. Homelessness history was assessed by asking clients how many separate episodes of homelessness they had experienced in the previous three years.
To examine the risk of incarceration among OEF/OIF/OND veterans compared with other veterans in the general population, the odds of being an OEF/OIF/OND veteran in the incarcerated sample were compared with the odds of being an OEF/OIF/OND veteran in the weighted population of nonincarcerated veterans according to the 2010 National Survey of Veterans. Odds ratios stratified by age groups were calculated. Supplementary analyses examining these odds ratios by gender and race were conducted.
Descriptive statistics were used to summarize the histories of homelessness of all veterans assessed by the HCRV program. Bivariate analyses, including t tests and chi square tests, were conducted to compare these characteristics among the incarcerated OEF/OID/OND and non–OEF/OID/OND veterans. Appropriate corrections were made for continuous variables that violated Levene’s test for equality of variances.
Multivariate analyses using backward stepwise logistic regressions were conducted to examine the independent associations of these characteristics to predict OEF/OIF/OND status. Only variables identified as significant in bivariate analyses were included in the multivariate analyses. To adjust for multiple comparisons and inflated type I error, all statistical tests described above were conducted at the .01 level with the statistical software SPSS, version 17.0 (10).
Of the total sample of 30,968 veterans assessed by the HCRV program, 1,201 (3.9%) were veterans of OEF/OIF/OND. Among the 29,767 (96.1%) other veterans, 3.0% (N=893) served before the Vietnam War, 27.9% (N=8,305) during the Vietnam War era, 47.8% (N=14,229) during the post-Vietnam era, and 18.7% (N=5,566) during the Persian Gulf War. A total of 2.6% (N=774) served after September 11, 2001, but were not deployed for OEF/OIF/OND.
National representative data from the 2010 National Survey of Veterans were used to compare the risk of incarceration among nonincarcerated veterans and OEF/OIF/OND veterans in HCRV. As Table 1 shows, across age groups, even the youngest, OEF/OIF/OND veterans were less than half as likely as other veterans to be incarcerated. This remained true for analyses of only males, only whites, only blacks, and only Hispanics (who may also have identified themselves as white or black).
Table 1Characteristics of veterans, by incarceration status, service era, and likelihood of incarceration among OEF/OIF/OND veteransa
| Add to My POL
|OEF/OIF/OND veterans||Other veterans|
| 20–29 years||501,974||638||329,942||1,395||.30||.24–.38|
| 30–39 years||470,083||375||806,872||3,661||.18||.14–.22|
| 40–49 years||477,246||149||2,497,942||10,368||.08||.06–.09|
| 50–59 years||204,319||33||3,598,472||10,569||.06||.04–.08|
| ≥60 years||135,549||4||11,733,941||3,754||.09||.03–.26|
| 20–29 years||377,495||625||236,262||1,332||.29||.22–.39|
| 30–39 years||304,255||367||514,936||3,530||.18||.14–.22|
| 40–49 years||336,309||143||1,737,108||10,166||.07||.06–.09|
| 50–59 years||146,305||33||2,733,655||10,482||.06||.04–.09|
| ≥60 years||118,749||4||10,548,167||3,741||.10||.03–.21|
| 20–29 years||339,476||390||225,880||862||.30||.22–.41|
| 30–39 years||368,927||224||627,977||1,947||.20||.15–.25|
| 40–49 years||335,687||89||1,820,921||4,954||.12||.07–.13|
| 50–59 years||156,555||14||2,687,107||5,100||.05||.03–.09|
| ≥60 years||105,491||0||10,520,700||2,507||.00||.00–.00|
| 20–29 years||74,698||152||75,073||375||.42||.23–.70|
| 30–39 years||66,096||93||121,422||1,291||.13||.08–.22|
| 40–49 years||104,069||43||461,586||4,463||.04||.03–.07|
| 50–59 years||34,247||15||613,676||4,570||.06||.03–.13|
| ≥60 years||17,690||4||666,721||935||.16||.05–.55|
| 20–29 years||416,447||71||293,084||106||.47||.33–4.15|
| 30–39 years||389,145||43||735,454||293||.28||.09–2.25|
| 40–49 years||419,643||9||2,262,004||680||.07||.02–.15|
| 50–59 years||185,006||4||3,190,619||608||.11||.04–3.24|
| ≥60 years||103,932||0||10,395,543||209||.00||.00–.00|
Table 2 shows the sociodemographic characteristics, criminal history, clinical status, and homelessness history of all veterans in the HCRV program. A majority of veterans were unmarried white men in their late forties who did not report combat exposure, were not working, and had little income in the past month. At the time of interview, veterans had been incarcerated for an average of over four years, and the most common current incarceration offense was a violent offense, property offense, or drug offense. Veterans reported an average age at their first arrest of 26 and an average of eight lifetime arrests. Nearly half reported using substances at the time of offense, but the response rate for that question was only 55.6%.
Table 2Characteristics of 30,968 incarcerated veterans, by service in Operations Enduring Freedom (OEF), Iraqi Freedom (OIF), or New Dawn (OND)a
| Add to My POL
(N=29,767)||Test of difference|
|Characteristic||N respondents||N||%||N respondents||N||%||N respondents||N||%||t||χ2||df|
| Widowed, separated, or divorced||17,694||57.2||479||39.9||17,215||57.9||152.51**|
| Never married||8,617||27.8||471||39.2||8,146||27.4|
|Days worked in past month||30,946||4.69||9.1||1,201||4.46||9.0||29,745||4.70||9.1||.88||1,301|
|Total income in past month||30,811||1,186||29,625|
| No income||20,091||65.2||733||61.8||19,358||65.3|
| Violent offense||28,472||9,995||35.1||1,124||422||37.5||27,348||9,573||35.0||3.06||1|
| Property offense||28,237||7,112||25.2||1,116||281||25.2||27,121||6,831||25.2||.00||1|
| Drug offense||28,222||6,851||24.3||1,115||242||21.7||27,107||6,609||24.4||4.18||1|
| Public order offense||27,919||4,326||15.5||1,106||189||17.1||26,813||4,137||15.4||2.23||1|
| Probation/parole violation||28,109||6,484||23.1||1,106||226||20.4||27,715||6,258||23.2||4.50||1|
|Length of current incarceration (M±SE months)||29,322||49.45±67.61||1,164||24.27±25.32||28,158||50.49±68.60||30.95**||1,968|
|Use of substances at time of offense||17,204||742||16,462|
| Alcohol and drugs||2,442||14.2||79||10.7||2,363||14.4|
|Age at first arrest (M±SD)||30,949||26.52±11.63||1,201||23.70±7.41||29,748||26.63±11.76||13.06**||1,456|
|Length of current homelessness||30,746||1,195||29,551|
| Not homeless||24,896||81.0||1,028||86.0||23,868||80.8|
| <6 months||2,226||7.2||79||6.6||2,147||7.3||29.35*||3|
| <2 years||1,779||5.8||55||4.6||1,724||5.8|
| ≥2 years||1,845||6.0||33||2.8||1,812||6.1|
|Episodes of homelessness||30,945|
| 1 or 2||6,584||21.3||213||17.7||6,371||21.4|
| 3 or 4||1,229||4.0||33||2.8||1,196||4.0|
|Currently homeless or any homeless history||30,948||9,201||30.0||1,201||280||23.3||29,767||8,921||29.7||24.63**||1|
|Any serious medical problem||30,226||18,050||59.7||1,181||633||53.6||29,045||17,417||60.0||19.12**||1|
| Alcohol abuse or dependence||30,840||13,848||44.9||1,196||516||43.1||29,644||13,332||45.0||1.56||1|
| Drug abuse or dependence||30,867||13,791||44.7||1,193||438||36.7||29,674||13,353||45.0||31.85**||1|
| Psychotic disorder||30,968||2,093||6.8||1,201||85||7.1||29,767||2,008||6.8||.20||1|
| Mood disorder||30,818||8,559||27.8||1,194||393||32.9||29,624||8,166||27.6||16.37**||1|
| Personality disorder||30,796||3,459||11.2||1,190||144||12.1||29,606||3,315||11.2||.94||1|
| Combat-related posttraumatic stress disorder||30,804||1,895||6.2||1,195||452||37.8||29,609||1,443||4.9||2160.17**||1|
| Adjustment disorder||30,804||6,263||20.3||1,195||285||23.9||29,649||5,978||20.2||9.65*||1|
| Other psychiatric disorder||30,757||3,324||10.81||1,190||190||16.0||29,567||3,134||10.6||34.18**||1|
|Use of any Veterans Affairs (VA) services in the past 6 months||30,686||1,791||5.8||1,189||125||10.5||29,497||1,666||5.7||49.22**||1|
|Willingness to use VA services|
| Mental health||28,606||16,845||58.9||1,132||738||65.2||27,474||16,107||58.6||19.37**||1|
| Medical care||28,733||23,476||81.7||1,134||935||82.5||27,599||22,541||81.7||.44||1|
| Residential treatment||28,244||10,390||36.8||1,108||345||31.1||27,136||10,045||37.0||15.83**||1|
| Social-vocational assistance||28,284||10,627||37.6||1,111||426||38.3||27,173||10,201||37.5||.29||1|
| Case management||28,391||11,855||41.8||1,116||465||41.7||27,275||11,390||41.8||.00||1|
Most veterans reported serious medical problems and had an alcohol or drug abuse or dependence diagnosis, but most were not currently homeless and reported no recent homelessness history. Most veterans had not used the VA in the past six months but reported being willing to use VA mental health and medical services.
Bivariate analyses revealed that compared with other veterans, OEF/OIF/OND veterans were significantly younger and were more likely to be white, be married, and have some income in the past month. OEF/OIF/OND veterans were expected to be incarcerated for a shorter duration of time, were more likely to have been using only alcohol (as opposed to alcohol and drugs) at the time of the offense, and were less likely to be homeless or have a history of recent homelessness. OEF/OIF veterans were less likely to have serious medical problems and to report drug abuse or dependence but were more likely to have a mood disorder, an adjustment disorder, and other mental health problems. Most notably, OEF/OIF/OND veterans were dramatically more likely to report combat exposure and combat-related posttraumatic stress disorder (PTSD). Compared with other veterans, OEF/OIF/OND veterans were also more likely to have used the VA in the past six months and were more willing to use VA mental health services but were less willing to use VA residential treatment.
Multivariate logistic regression analyses (Table 3) showed that among HCRV clients, younger age, being married, reporting combat exposure, having a shorter duration of incarceration, being first arrested at an older age, having combat-related PTSD, and not having a drug abuse or dependence diagnosis were all independently associated with OEF/OIF/OND service, with the largest effect sizes for combat exposure and combat-related PTSD.
Table 3Characteristics of veterans assessed by HCRV and their association with service in Operations Enduring Freedom, Iraqi Freedom, and New Dawna
| Add to My POL
|Race-ethnicity (reference: white)|
|Marital status (reference: married)|
| Widowed, separated, or divorced||–.30||.12||.74||.58–.94|
| Never married||–.50**||.13||.61||.47–.78|
|Duration of current incarceration||–.00*||.00||1.00||.99–1.00|
|Age at first arrest||.03**||.01||1.03||1.02–1.05|
|Drug abuse or dependence (reference: no drug abuse or dependence)||–.31*||.10||.74||.61–.89|
|Combat-related PTSDc (reference: no PTSD)||1.14**||.12||3.12||2.46–3.95|
|Use of any Veterans Affairs services in the past 6 months (reference: no use)||.46||.18||1.58||1.12–2.25|
National administrative data from the HCRV program revealed that OEF/OIF/OND veterans represented a small minority (3.9%) of the veteran population incarcerated in state and federal prisons. Moreover, even among the youngest age group (aged 20–29 years) and within different ethnic groups, OEF/OIF/OND veterans were less than half as likely to be incarcerated as other veterans, suggesting this finding is not simply an artifact of age or race. This finding may reflect the increased attention and service supports provided to OEF/OIF/OND veterans by the VA (11,12) and other organizations (13,14) and is consistent with reports that the proportion of prisoners who are veterans has declined steadily over the past three decades (5). It is also in accord with studies showing less risk of homelessness (15) and substance abuse (16) among OEF/OIF/OND veterans than among veterans of earlier service eras.
Of incarcerated OEF/OIF/OND veterans, a majority were unmarried white men in their early thirties who were not homeless, were not working, and had no income. Multivariate analyses showed that compared with other veterans, OEF/OIF/OND veterans were younger, were more likely to be married, and expected to have shorter incarceration periods. The most common current incarceration offense was a violent offense, reported by 38% of OEF/OIF/OND veterans and 35% of other veterans; these proportions are consistent with, although slightly lower than, the proportions contained in a previous report (5). After the data were controlled for age, OEF/OIF/OND veterans still had a shorter incarceration period and reported being arrested fewer times in their lifetime than other veterans.
Alcohol abuse or dependence was most common diagnosis for both OEF/OIF/OND veterans (43%) and other veterans (45%). Psychotic disorder was the least common diagnosis among OEF/OIF/OND veterans (7%), and combat-related PTSD was the least common diagnosis among other veterans (5%). Most incarcerated veterans had some mental health or substance abuse problems, which is consistent with previous studies (14,17). The high rates of substance use disorders among all incarcerated veterans suggest that the HCRV program should aggressively seek to address substance abuse problems. It is likely that many offenses leading to incarceration were either directly or indirectly related to substance abuse, as evidenced by the sizable proportion of veterans who endorsed using substances at the time of their offense and the number of veterans who were currently incarcerated for drug offenses.
Incarcerated OEF/OIF/OND veterans were less likely than other veterans to have a drug abuse or dependence diagnosis, which may be due to notable decreases over the past 28 years in the use of illegal drugs and cigarettes among military personnel, likely reflecting the military’s implementation in the1980s of a zero tolerance policy for drug abuse (18).
Perhaps the most salient clinical finding was that incarcerated OEF/OIF/OND veterans were three times more likely than other incarcerated veterans to have combat-related PTSD, probably reflecting their greater likelihood of combat exposure. This finding may have important clinical implications, given that PTSD has been associated with a host of negative health, psychosocial, and functional problems among OEF/OIF/OND veterans (19–21). Incarcerated adults often do not receive needed mental health treatment (22,23), and provision of these services after their release may be particularly important to prevent recidivism, homelessness, and other negative outcomes (24,25). Thus, programs like HCRV can hope to be instrumental in helping incarcerated veterans connect to needed mental health services upon their release and successfully integrate into the community.
Previous efforts to link incarcerated veterans to VA health care services have been challenging, but specialized outreach services have shown promise (26). Further research is needed to evaluate the success of prerelease outreach efforts in linking veterans to needed health and mental health services after discharge and to both preventing criminal justice system recidivism and homelessness and improving health.
Several limitations of this study deserve mention. Although the HCRV program did not formally target particular groups of veterans for outreach, the representativeness of the national sample in this study of the population of incarcerated veterans is undetermined. The HCRV program was directed at veterans soon to be discharged, so the sample may underrepresent younger veterans with longer sentences, although this possibility could not be empirically evaluated. However, one might expect that the VA’s focus on OEF/OIF/OND veterans and the extensive media attention paid to this group of veterans may lead to an oversampling of this group. Yet, this expectation only strengthens the credibility of the finding that incarcerated OEF/OIF/OND veterans are less likely to be incarcerated than other veterans.
Causal directionality of the associations between OEF/OIF/OND status and other characteristics is unknown, and causal risk factors cannot be conclusively determined from these data. The diagnoses of mental disorders, including PTSD, were not assessed with structured, standardized measures and relied on veteran self-report and the clinical skills and judgment of HCRV specialists. Moreover, PTSD from noncombat-related events was not assessed, which may be an area worthy of future study.
Data about whether veterans in the sample had filed VA disability claims were not available. However, given that combat-related PTSD is the most common VA psychiatric disability claim filed among veterans (27,28), it would have been useful to examine how pending disability claims or intent to apply for disability influenced the clinical presentation of incarcerated veterans (29) and the study results, especially because older veterans may be more likely than younger veterans to have had their disability claims settled.
HCRV data suggested that OEF/OIF/OND veterans are at lower risk of incarceration than veterans of other service eras. However, incarcerated OEF/OIF/OND veterans reported higher rates of combat exposure and PTSD than other incarcerated veterans, suggesting that some incarcerated OEF/OIF/OND veterans may especially benefit from PTSD treatment and VA services upon their release.