The Unfinished Reform

The Department of War is changing how it handles future personnel actions. That is a good start, but it does little for veterans still living with questionable decisions already treated as final. One disputed finding can end a career, cost someone a retirement, hurt civilian employment, and follow a veteran for years. Congress does not need to create another court or reopen every military career. It needs a narrow process for credible cases, built around one principle: one government, one door, one accountable decision.

Secretary of War Pete Hegseth has argued that the individual warfighter remains the military’s decisive advantage and that the Department has a “sacred duty” to protect long-term performance, resilience, and health. That duty should not end at separation. When credible evidence shows that government action may have wrongly ended a career or continues to damage retirement, livelihood, or standing, the government should be willing to examine its own record.

The COVID-19 vaccine mandate offers a useful precedent. The point is not that every vaccine-related separation was unlawful. The point is that the government identified a defined group of completed personnel actions and went back to review them. On March 19, 2026, the Air Force reported identifying 377 people discharged solely for declining vaccination. It also reviewed 218 cases outside the initial criteria and corrected 595 records altogether. Congress can use the same basic approach for other documented categories of administrative injustice: set firm criteria, review the record, and correct what the evidence supports.

Personnel decisions are cumulative. One disputed finding can affect an evaluation, promotion, continued service, and retirement. When material records are missing or no authority examines the full chain, administrative finality merely freezes the government’s last decision in place. The family carries the lost income; the veteran carries the professional, financial, and personal consequences.

The problem is clearest in the National Guard, where state command, federal recognition, the National Guard Bureau, correction boards, and separate personnel and retirement systems may each control part of a case. A federal correction board may consider evidence concerning a state action, but its authority to correct state records, reverse state decisions, compel state participation, or implement every dependent federal consequence may be limited. A Reserve Component soldier can enter the 18-year retirement-protection window and lose a career through intertwined state and federal actions while no single authority owns the complete remedy.

While the offices sort out who has authority, the clock keeps running. Veterans chase records the government once held, explain the same case to one office after another, and receive partial answers that do not fix pay, retirement, or status. For someone already carrying the strain of military service, years of uncertainty, financial pressure, and institutional rejection can take a real toll. The government created these jurisdictions and controls most of the records. It should also provide one place responsible for moving a credible case to a complete answer.

Correction boards still have an important role, but this proposal is not simply another appeal. It is meant to deal with three problems the normal process does not handle well. First, boards decide applications; they generally do not rebuild every incomplete or disputed record. Second, no single office may control all the state, federal, personnel, pay, and retirement pieces. Third, even a favorable decision can stall until several offices act. The ABCMR also starts from the presumption that the Army acted correctly, does not serve as an investigative body, and leaves the veteran to prove the error or injustice.

·        One point of entry: Veterans file once, receive one case number, and work with one office responsible for intake, investigation, coordination, tracking, and verification.

·        Independent civilian review: Reviewers sit outside the commands and organizations being challenged, supported by investigators, records specialists, and legal advisors.

·        Authority to obtain and reconstruct records: The office coordinates state and federal records and determines the fair inference when government documents remain missing after good-faith efforts.

·        Firm deadlines and written explanations: Screening occurs within 60 days and accepted cases receive a merits decision within 180 days unless a documented extension is required.

·        Certified implementation: Every favorable decision identifies the records, agencies, deadlines, dependent actions, and one official responsible for confirming completion.

Officials or correction boards acting under 10 U.S.C. § 1552 would still make the final decision and explain in writing whether they adopted, changed, or rejected the panel’s recommendation. Congress should also be direct about any authority the current law does not provide, including investigation, access to records, state-federal coordination, tolling, enforcement, and implementation. The initiative should begin as a limited pilot with clear eligibility rules, a filing window, and a three-year review. A previously denied applicant should return only with material evidence that was never meaningfully considered or proof of a serious process failure that could have changed the outcome.

Screening should ask whether the applicant identified a credible defect that could reasonably have changed a finding, penalty, separation, promotion, or retirement result. Full review should then separate three questions: Was there error or injustice? Did it affect the outcome? What relief does the corrected record support? Passing the screen earns a full review; it does not presume the applicant wins.

Error and injustice are not always the same thing. Error asks whether the government followed the law, used the right standard, preserved the record, and provided the required process. Injustice asks whether the result is still fair after the full evidence and consequences are considered. Some cases involve both; others involve only one. This initiative should not reopen a case because someone dislikes the outcome. It should focus on defects serious enough to call that outcome into question.

The pilot should focus on sanctuary and 18-year cases, missing government records, omitted required hearings or boards, repeated allegations following a documented credibility defect, and claims delayed by movement between state and federal authorities. Congress should also address the three-year filing rule in § 1552(b): a veteran should not be timed out while the government decides who owns the case.

Whistleblower cases show why safeguards on paper require verification. In a 2015 report, GAO found that DoD missed the 180-day notification requirement in about half of military reprisal investigations closed in fiscal 2013. Cases closed in fiscal years 2013 and 2014 averaged 526 days, and the central system lacked records for at least 22 percent of service-conducted cases. DoD later improved tracking, oversight, milestones, and independence procedures. Congress should now determine whether similar gaps persist and whether substantiated claims produce timely, enforceable restoration.

A full review should reconstruct the decision rather than repeat the file’s conclusions. It should determine whether the evidence was credible; required notice, hearings, counsel, sanctuary, retirement, and medical protections were provided; state and federal authorities acted within their jurisdiction; records were accurate; and any failure could have changed the result. The panel should then map every later action materially dependent on the disputed decision.

A favorable decision must identify the records to correct, responsible agencies, deadlines, and official who will certify completion. If an erroneous finding influenced evaluation, promotion, continuation, retirement, pay, or benefits, each dependent action should be reviewed under the corrected record. Relief should restore only what that record supports. A correction that does not reach the pay and retirement systems is not a correction.

The strongest objection is cost and workload. That concern is real, but the alternative is not free. The government already answers repeat applications, congressional inquiries, inspector general complaints, litigation, and disputes circulated among agencies unable to resolve the whole case. A limited pilot with disciplined screening would measure demand, cost, recurring defects, relief granted, and whether corrections were actually implemented.

The limits should remain clear. The initiative would not create a court, retry courts-martial, disturb matters reserved for clemency, or treat an unsubstantiated allegation as false. Acceptance would not establish entitlement, complainants and witnesses would retain protection, and applicants outside the pilot would remain on the ordinary board track.

Congress should place the initiative in the next National Defense Authorization Act. The law should require authorized officials or correction boards to adopt, modify, or reject recommendations in writing; certify implementation of approved corrections; and identify where current law prevents a complete remedy. The Armed Services Committees should hold public oversight hearings, and veterans’ organizations should submit documented cases for the congressional record.

This is not only a veterans’ issue. People still serving watch what happens when someone challenges a record after separation. If a credible case can disappear between offices, the lesson is that accountability depends on whether a person can outlast the process. A narrow and transparent pathway would send a better message: service members remain accountable for misconduct, and the government remains accountable when it makes a material mistake.

The Secretary should have 180 days to design the initiative, identify the official responsible for implementation, and list every remedy § 1552 cannot reach. Annual reporting should show applications, screening decisions, decision times, relief granted, completed personnel, pay, and retirement corrections, recurring defects, and points where authority ran out. Those measures would show whether the initiative resolves cases rather than merely moving them.

Military service is built on a two-way obligation. The nation expects accountability, sacrifice, and respect for lawful authority from those who wear the uniform. The government should be held to a standard too. When it has the power to end a career, it should also have the courage to correct a consequential mistake. Finality matters, but it cannot become a substitute for justice. Separation should not erase the government’s responsibility for its own actions.

Congress need not presume every veteran was treated unjustly. It must ensure that credible claims receive complete, independent, and timely review. The immediate request is simple: authorize a limited Administrative Restoration Initiative in the next NDAA and require public reporting on whether approved corrections were fully implemented.

One government. One door. One accountable decision.

Mark W. Castillon served in the Army National Guard from 2004 to 2024, including deployments in support of Operations Iraqi Freedom and New Dawn and emergency responses from Hurricane Katrina through COVID-19. He writes on veteran mental health, whistleblower protection, leadership accountability, and due-process reform.