In October 2016, the Pentagon’s health system, TRICARE, scrapped its caps on mental health visits and inpatient days, cut outpatient copays from $25 to $12 for some beneficiaries, expanded coverage for addiction treatment, and loosened the paperwork that kept civilian therapists out of its network. TRICARE is the insurer of record for 9.5 million service members, retirees, and their families
Nearly a decade later, Jangho Yoon of the Uniformed Services University of the Health Sciences in Bethesda has published the first system-wide test asking if that reform has changed the outcome that mattered most: whether fewer military beneficiaries have attempted suicide.
Writing in Health Economics Review, Yoon draws on more than 140 million person-quarters of claims data from the Military Health System Data Repository, covering 2.3 million active-duty troops, 1.5 million retirees under 65, and 5.7 million dependents between early 2015 and early 2020.
Because every beneficiary received the reform at once, there was no untreated group to compare against. Yoon instead uses an interrupted time-series design that looked for a break in the trend when parity took effect, with individual fixed effects to absorb differences in sex, race, and service branch.
The headline finding was that parity worked — but unevenly. All three groups saw an immediate drop in medically attended suicide attempts: roughly five fewer per 100,000 among active-duty service members, four fewer among retirees, and seven fewer among dependents, or declines of 11, 20, and 16% against the pre-reform baseline. The divergence came afterward.
Retirees kept improving by about 1.6 attempts per 100,000 each quarter, and dependents by 3.1. Over the 14 quarters Yoon tracked, the cumulative reductions reached 241 fewer attempts per 100,000 retirees and 435 fewer per 100,000 dependents, which the paper translates into roughly 3,600 fewer attempts among retirees and some 23,000 fewer among family members.
Active-duty troops tell a different story. Their initial dip did not hold. The quarter-by-quarter estimates hovered around zero, the cumulative effect lost statistical significance after about 21 months, and the post-reform slope was flat.
Junior enlisted personnel, the group under the heaviest occupational stress, showed the largest immediate decline and then a statistically significant reversal, a pattern Yoon reads as parity encouraging help-seeking only for the gains to be eroded by career pressures and stigma.
Yoon’s explanation turned on how the reform touched each group. Retirees and dependents paid copays and often relied on civilian providers, so cheaper care and a bigger network relaxed both financial and practical barriers at once. Active-duty members already paid nothing and got most of their care at military treatment facilities, so the reform reached them only through the supply side, and only partly.
Two-thirds of active-duty troops lived in areas designated as psychiatrist shortage zones, and 6% of all beneficiaries lived where there was no psychiatrist at all. What the reform could not touch were the barriers particular to military life: fears about confidentiality, career repercussions, and the cost of being seen to seek help inside a close-knit institution. For that population, Yoon concludes, parity should be seen as “a necessary but insufficient strategy for suicide prevention.”
The paper has arrived amid worsening trends. Suicide deaths among active-duty troops rose from 24.3 per 100,000 in 2021 to 28.2 in 2023, and the raw attempt rate in Yoon’s data climbed throughout the study period regardless of parity, peaking at about 73 per 100,000 in early 2019.
Diagnoses of anxiety and post-traumatic stress disorder nearly doubled among service members between 2019 and 2023, and mental disorders accounted for the largest share of inpatient bed days in the force that year.
The Pentagon’s own annual report on military suicide for 2023 called the long-term rise statistically significant and conceded that “it is not possible to identify a single root cause that might explain this trend.”
Yoon acknowledges these limits. Claims captured only attempts that reached a clinic or emergency room. The reform changed several things at once, so the study cannot say which lever mattered. Guard and Reserve members were excluded, and without a control group, concurrent events cannot be fully ruled out, though the results held up under alternative time trends, a logit model, and a trimming procedure for the linear probability model.
The study points to the Brandon Act, signed in late 2021 to give service members confidential access to mental health care, as the kind of cultural reform that might eventually let parity do for troops what it already did for their families.
