Invisible Warriors
Veterans in State Prisons
August 2025
By Angela Hawken, PhD, Sandy Felkey Mullins, JD, Lina Cook, MS, and Elena Pereira, MPH, MPA, New York University
Accurately identifying veterans in state prisons helps ensure they get the support they have earned through their service—from housing and mental health care behind bars to benefits through the Department of Veterans Affairs (VA) and reintegration services after release. Reliable counts of veterans allow prisons, policymakers, and veterans’ service organizations to connect the right people to the right resources at the right time. This reduces recidivism, protecting public safety, and improves employment, health, and other outcomes for veterans, their families, and communities.
The number of veterans currently in state prisons is unknown. From 2011 to 2012, the Justice Department identified 181,500 veterans in state and federal prisons as well as local jails, a number that represented about 8% of people incarcerated during that time.1 The most recent official government estimate identified 98,100 veterans (98% of them male) in state prisons in 2016,2 a figure that represented just under 8% of the total state prison population that year.3 There has been no official update to the federal count of veterans in prisons or jails since 2016.
Deriving an accurate count of veterans in state prisons is challenging because there is neither a common definition of veteran nor a uniform or best practice for whether and how state correctional agencies should identify veterans. Many states rely on veterans to self-identify their status during prison admission, but this method can be unreliable as veterans may hesitate to disclose military service due to stigma, shame, or fear of losing benefits.4 Conversely, some may falsely claim veteran status hoping to access veteran-specific programs or services during imprisonment. Some states identify veteran status through an external system that connects to VA records—the Veteran Re-Entry Search Service (VRSS)—but this method is not used consistently.
This study sought to understand the methods used by state prison systems to identify veterans in their custody. First, it presents each state’s approach to identifying veterans, using two states with available data as case studies to compare veteran identification methods (self-report and VRSS verification). Using these varied methods, the authors derived a total count of veterans known to be in state prisons, while acknowledging that this number is certainly an undercount of the true number of veterans.
Key Takeaways
- A lack of uniform definitions and methods for identifying veterans, as well as the absence of a recent federal estimate, means there is no accurate count of the number of veterans currently imprisoned in the U.S. This hampers efforts to shape policy and deliver appropriate programming to the nation’s veterans.
- The analysis conducted for this report indicated that in 2023, there were 49,000 veterans in prisons in the 46 states that provided a count, representing about 5% of the total prison population in those states.
- State correctional officials expressed reservations about the accuracy of the counts they provided to researchers for this study.
- The estimates reported here are almost certainly undercounts, but analysis of the best available information indicates that veterans make up a smaller share of the U.S. prison population than they did in the years following the Vietnam War.
- Better veteran identification in prisons is essential for connecting veterans to the services they have earned. The authors recommend the consistent use of VRSS, improvements to its accuracy, and more transparent reporting. These steps align with the CCJ Veterans Justice Commission’s broader call to enhance identification of veterans, expand in-custody VA access, and strengthen reentry support.
Veteran Identification Methods
Forty-three states provided information about their system for identifying veterans. These states reported using one of three methods (Figure 1).
- Self-report. Twenty states relied on veterans to self-report their status, information that was often collected during prison admission.
- VRSS verification. Sixteen states employed the Veterans Re-Entry Search Service (VRSS), a VA-based system that uses social security numbers to verify veteran status. VRSS verification was often conducted during prison admission, although some states regularly verified their entire prison populations.
- Hybrid. Seven states used both self-report and VRSS to identify veterans. This method is generally referred to as “self-reported and verified,” but the authors could not determine whether states verified the entire prison population through VRSS or only those who self-reported as veterans.
Additionally, three states provided incarcerated veteran counts without information on how they determined veteran status, and four states did not provide incarcerated veteran counts or identification methods.
Figure 1. Methods Used by State Prison Systems to Identify Veterans, 2023
Notes: Veteran identification methods were collected between June 2022 and May 2024. “Unavailable” refers to states that do not identify veterans, states that identify veterans but did not provide a method, and states that did not respond to the request for information. Nebraska uses VRSS in addition to self-report, but results are not entered into their records system; South Carolina uses self-report for internal purposes only and relies on VRSS for official verification. See Table S1 in the supplemental methodology report for underlying state-level data.
Comparing Veteran Identification Methods: Self-Report and VRSS
Data from South Carolina and Nebraska were used to determine whether self-report and VRSS verification yielded similar counts. These states provided sufficiently rich data for analysis and employed both methods of identifying veterans. Both states ask individuals to self-identify their veteran status and also submit their entire prison population to VRSS for verified identification.
Nebraska data also allowed for a comparison of self-reported veteran status among people with multiple prison admissions. Prior research suggests that some veterans may not self-report due to stigma, shame, or concerns about losing VA benefits.5 Conversely, others may falsely claim veteran status, hoping to access veteran-specific programs or services during incarceration. Using VRSS for verified identification should yield a more accurate count because it does not rely on self-reporting, although human error may still occur.
South Carolina
The South Carolina Department of Corrections provided data on 4,204 individuals in state prison custody between January 2015 and April 2024 who met at least one of two criteria: (1) they self-reported their veteran status at admission (including “non-veteran” responses), or (2) they were identified as veterans through VRSS during monthly full-population checks.
Among those who self-identified as veterans (n = 2,858), approximately 81% (2,302 people) were verified as veterans through VRSS. The remaining 19% (556 people) self-reported veteran status but were not verified through the federal system. This discrepancy may reflect data entry errors (e.g., incorrect Social Security Numbers), incomplete VA records, misclassifications by staff, or, in some cases, individuals falsely claiming veteran status to access veteran-specific services in prison. While these are all plausible explanations, it is not possible to determine with certainty why verification failed for nearly 20% of the people who self-identified as veterans.
On the other hand, about 34% of the 3,500 individuals verified as veterans by VRSS (roughly 1,200 people) either self-reported as non-veterans or did not self-report their veteran status at all. These individuals may have chosen not to disclose their status, may have been unsure whether they met the criteria to be considered veterans, or may have been affiliated with the military in other ways. For example, they may have been military dependents, civilian defense employees or contractors, ROTC participants, or military academy applicants, all groups that sometimes appear in federal records but do not qualify as veterans under most justice system definitions.
Figure 2. Self-Reported Versus VRSS-Verified Veteran Counts from South Carolina Department of Corrections, 2015 – 2024
Nebraska
The Nebraska Department of Correctional Services provided data on 90 individuals admitted between March 2023 and April 2024. All individuals were screened through VRSS. The data includes VRSS results, the individuals’ self-reported veteran status in the current incarceration period, and self-reported veteran status in a previous incarceration period (if any).
Veteran identification during the current incarceration
Of the 90 people in the dataset, 71 were identified as veterans through VRSS. Among that group, 34 individuals (or 48%) had also self-reported as veterans in the current incarceration period. The remainder either did not self-report (20% or 14 individuals) or self-reported that they were not veterans (32% or 23 individuals).6
The remaining 19 people in the larger 90-member dataset had self-reported as veterans and were not found to be veterans through VRSS. In other words, there were a total of 53 individuals who had self-reported as veterans in the current incarceration period, and 19 of the 53 (or 36%) who had self-reported as veterans were not verified as veterans through VRSS.
Veteran identification during a previous incarceration
About 40% of the 71 people identified above had a prior admission to the Nebraska Department of Correctional Services. Of the 12 individuals who self-reported their veteran status during their current admission, only half self-reported as veterans during a previous admission, underscoring the unreliability of this method.
The findings from these two states demonstrate the inconsistencies and limitations of relying solely on self-report or VRSS to identify veterans in correctional settings. Both South Carolina and Nebraska show substantial discrepancies between self-reported and verified veteran status, with some individuals claiming veteran status but not verifying through VRSS, and others being verified through VRSS but not disclosing their status. These mismatches suggest that neither method captures the full picture on its own.
Deriving a Count of Veterans in State Prisons
Although identification methods were available for 43 states, three additional states provided a count of veterans either through response to an inquiry or in published materials.
States varied widely in how many veterans they identified—in total numbers, as a share of the prison population, and even after accounting for the state’s veteran population in the community. These differences may have arisen from variation in identification methods, or they may be attributable to other factors that were not measured.
(Note: Figures 3-5 show the number, share, and ratio of veterans in state prisons. The count estimate spanned the period from June 2022 to May 2024, with most counts occurring in 2023. See Table S1 in the supplemental methodology report for underlying state-level data.)
Figure 3 shows the number of veterans identified in each state prison system. The number of veterans in state prison ranged from a low of 30 in Vermont to a high of 8,134 in Texas.
Figure 3. Number of Veterans Identified by State Prison Systems, 2023
Figure 4 shows the identified veterans as a percentage of the total state prison population. Factoring in each state’s overall prison population, New Jersey reported the smallest share of veterans in state prison, at just over 2%. Oregon reported the highest share of veterans at 15%. Texas, the state with the largest number of veterans, did not report the largest share of veterans in state prisons.
Figure 4. Identified Veterans as a Share of State Prison Populations, 2023
Figure 5 shows the share of veterans in state prisons compared to the share of veterans in each state’s general population to see whether veterans are over- or underrepresented in prison. A ratio above 1 means veterans are overrepresented in prison; a ratio below 1 means they are underrepresented.
Based on reported data from states, veterans were overrepresented in prison in 11 of 46 states, with Oregon showing the largest ratio (2.29). In contrast, veterans were underrepresented in 35 states, with Montana and Vermont reporting the smallest ratios (0.35). However, these findings may understate true overrepresentation if states are not accurately identifying all veterans in their custody.
Figure 5. Ratio of Imprisoned Veterans to State Veteran Population, 2023
The Challenge of Counting Veterans in Prison
Summing the individual state counts, this analysis identifies roughly 49,000 veterans in prisons in 46 states in or around June 2023. This number represents about 5% of the total state prison population in those 46 states. See Table S1 in the supplemental methodology report for full state-by-state data on identification methods, veteran counts, prison shares, and representation ratios.
Accurately identifying veterans in state prisons remains difficult. Twenty states relied solely on self-reporting, a method prone to undercounting. Even in states using a hybrid approach, the wording of questions regarding veteran status varied; some states asked about military service while others asked about veteran status. This may lead to confusion, especially for those with other-than-honorable discharges, which can limit access to VA benefits and make some former service members uncertain of their veteran status.7
Although verified systems such as VRSS are more reliable, the deployment of VRSS varied widely across states. Some states submitted names only of self-reported veterans to VRSS rather than screening the full prison population, limiting the tool’s effectiveness. In other states, VRSS verification results were not integrated into case management systems, and inconsistencies were reported, such as individuals verifying their status at one time and not another.
Additionally, many state corrections officials expressed little confidence in the accuracy of their veteran counts. Some lacked clear methodologies or were unable to describe how their numbers were generated.
Some of the decline in the share of veterans in state prisons between 2016 and 2023 likely reflects a real, long-term trend. This is consistent with historic declines that have been observed since 1986, when veterans made up about 20% of the state prison population.8 Examining state prison veteran counts as a percentage of the population over time shows that veterans made up 8% of the state prison population in 2016.9 The 49,000 figure generated by this analysis would put veterans at 5% of the state prison population in 2023.
Improving Veteran Identification in Prisons
The 49,000 incarcerated veterans identified in 46 states through this analysis is almost certainly an undercount, but it reflects the number states officially recognize. The drop from earlier estimates may be partly due to variation in identification methods, as well as overall declines in the number of incarcerated veterans and the greater prison population.
Regardless of what’s driving the trend, better data matter. When veterans aren’t identified, they may miss out on correctional programs such as veteran-specific housing units, specialized prison-based and community services, or access to VA benefits. Accurate counts are essential for shaping effective policy and ensuring that veterans receive the support they’ve earned through their military service.
This analysis highlights the fragmented and inconsistent nature of veteran identification in state prisons, reinforcing the CCJ Veterans Justice Commission’s call for a more uniform, data-driven approach. The commission recommends adopting a standardized definition of “military veteran” across government and expanding the use of tools like VRSS to better identify and serve veterans at all stages of the justice process. Stronger identification practices in corrections are also foundational to broader reforms proposed by the commission, including expanded diversion of veterans from prosecution and incarceration, targeted treatment, and reentry options tailored to veterans’ unique needs.
Recommendations
- States should consistently use VRSS. During the prison admission process, correctional staff should ask all people whether they have served in the U.S. Armed Forces or Reserves (not simply if they are “veterans”) and verify their information through VRSS. States should also regularly re-run their full prison population through VRSS to catch earlier errors.
- The VA should improve VRSS accuracy. Corrections officials report inconsistencies with the system. The VA should investigate how errors occur and identify ways to improve the performance of VRSS.
- States should report veteran counts publicly. Most states publish prison demographics but rarely include veteran status.
- The federal government should release regular data. The Bureau of Justice Statistics (BJS) has published only three reports on incarcerated veterans since 2000. The VA and BJS should collaborate to provide more frequent and accurate national data using VRSS.
About the Authors
Angela Hawken, PhD, is the former director of the Marron Institute of Urban Management at New York University, where she was also professor of public policy. She is now a member of the founding faculty and vice dean for research at the School of Government and Policy at Johns Hopkins University.
Sandy Felkey Mullins, JD, is a senior research scholar in the Litmus program at the Marron Institute of Urban Management at New York University, where she assists public safety-related agencies and organizations serving people who have been involved with the criminal justice system in collaboratively designing and implementing reforms.
Lina Cook, MS, is an assistant research scientist in the Litmus program at the Marron Institute of Urban Management at New York University. As a statistician, she primarily uses data to generate numbers and metrics to inform the work that is being done in the social justice realm.
Elena Pereira, MPH, MPA, is a health legislative research analyst at the New York City Department of Health and Mental Hygiene and a former participant in the Litmus program at the Marron Institute of Urban Management at New York University.
Acknowledgements
Stephanie Kennedy contributed to this report, along with former staffers Kathy Sanchez, Aaron Rosenthal, and Jim Seward, as well as other members of the Council team.
Support for the Veterans Justice Commission comes from Bank of America, The Arthur M. Blank Family Foundation, Craig Newmark Philanthropies, The Just Trust, LinkedIn, the National Football League, T. Denny Sanford, the Wilf Family Foundation, and the May & Stanley Smith Charitable Trust, as well as the John D. and Catherine T. MacArthur Foundation, Southern Company Foundation, #StartSmall, and other CCJ general operating contributors.
Suggested Citation
Hawken, A., Mullins, S. F., Cook, L., & Pereira, E. (2025). Invisible warriors: Veterans in state prisons. Council on Criminal Justice. https://counciloncj.org/invisible-warriors-veterans-in-state-prisons/
Endnotes
1 Bronson, J., Carson, E. A., & Noonan, M. (2015). Veterans in prison and jail, 2011–12 (NCJ 249144). Bureau of Justice Statistics. https://bjs.ojp.gov/content/pub/pdf/vpj1112.pdf
2 Maruschak, L. M., Bronson, J., & Alper, M. (2021). Veterans in prison (NCJ 252646). Bureau of Justice Statistics. https://bjs.ojp.gov/content/pub/pdf/vpspi16st.pdf
3 See Table 1: Carson, E. A. (2018). Prisoners in 2016 (NCJ 251149). Bureau of Justice Statistics. https://bjs.ojp.gov/content/pub/pdf/p16.pdf
4 Ahlin, E. M., & Douds, A. S. (2020). If you build it, will vets come? An identity theory approach to expanding veterans’ treatment court participation. Criminal Justice Review, 45(3), 319-336. https://doi.org/10.1177/0734016820914075
5 Ahlin & Douds, 2020.
6 People may opt not to self-report (marked “N/A” in the data) or may self-report that they are not veterans (marked “N”).
7 Department of Veterans Affairs. (2025). Veterans benefits administration: Applying for benefits and your character of discharge. https://www.benefits.va.gov/benefits/character_of_discharge.asp
8 See Figure 1: Noonan, M. E., & Mumola, C. J. (2007). Veterans in state and federal prison, 2004 (NCJ 217199). Bureau of Justice Statistics. https://bjs.ojp.gov/content/pub/pdf/vsfp04.pdf
9 See Table 2: Maruschak, Alper, & Bronson, 2021.
