Medicaid demonstrations are changing a narrow but consequential part of health care for people leaving jails and prisons. California received the first federal Section 1115 approval allowing a limited set of services before release, and other states have since pursued similar authority.
The policy does not make Medicaid the general payer for health care throughout a person's incarceration. It partially waives the longstanding inmate exclusion rule for eligible people during a defined pre-release period. The purpose is to begin assessment, treatment and care coordination before a person returns to the community, when a break in medication or follow-up can carry serious risk.
As of July 2026, KFF counted 19 state waivers approved under the Biden administration to facilitate reentry. The number shows that the model has spread beyond a single-state pilot. It does not establish that every approved state has implemented the same services, reached the same population or produced the same outcomes.
California Opened a Limited Pre-Release Window
CMS approved California's demonstration amendment in January 2023. The authority lets Medi-Cal cover specified pre-release services for eligible adults and young people in state prisons, county jails and youth correctional facilities. It can also connect them with community Medicaid providers as early as 90 days before release.
CMS described the benefit as a limited package rather than comprehensive Medicaid coverage behind bars. Its stated goals include identifying health needs, stabilizing certain serious physical and behavioral conditions, providing medications for addiction treatment where appropriate, and arranging continued care after release.
The distinction matters. Federal approval creates legal and financial authority for a demonstration; it does not mean that every facility can immediately deliver the full model. Correctional agencies, state Medicaid systems, managed-care plans and community providers still have to exchange eligibility and release information, build referral processes and maintain continuity when release dates change.
The Evidence Defines a High-Risk Transition
The urgency is grounded in more than administrative inconvenience. A landmark cohort study followed 30,237 people released from Washington State prisons from 1999 through 2003. During the first two weeks after release, their adjusted risk of death was 12.7 times that of other state residents. The relative risk of fatal drug overdose was 129, with a wide 95 percent confidence interval of 89 to 186.
Those figures are often repeated without their boundaries. They came from one state and an earlier drug environment, and they compare a recently released prison population with the general population. They demonstrate an acute transition risk; they are not a current national rate and do not mean that any individual leaving custody faces a predetermined outcome.
More recent observational evidence supports the importance of treatment continuity. A New England Journal of Medicine study of county-jail programs found that people with probable opioid use disorder who received medications for opioid use disorder in jail were more likely to receive such medication within 30 days after release and had lower observed rates of overdose, death and reincarceration than those who did not receive it. Because the study was observational, it cannot by itself prove that the medication programs caused every difference.
Implementation Will Decide What the Waivers Deliver
The demonstrations attempt to move work that often begins after release into the weeks before it. That can include screening, case management, prescriptions, a supply of medication at release and appointments with community providers. CMS also identified conditions such as diabetes, heart failure, hypertension, schizophrenia and substance-use disorders as areas where ambulatory treatment and continuity may matter.
Short and unpredictable jail stays complicate that design. A facility may not know a person's release date 90 days in advance. Medicaid eligibility records and correctional records may not match, and community clinics may have limited capacity. KFF's state data also show that processes for suspending and reinstating Medicaid coverage vary widely, from automated systems to largely manual ones.
Those operational details are not secondary. A benefit that exists in a waiver but does not produce a confirmed appointment, a usable prescription or an active coverage record at release has not closed the care gap for that person. Measuring enrollment alone would therefore give an incomplete picture of performance.
The Test Is Continuity, Not the Approval Count
CMS requires California to conduct a comprehensive and rigorous evaluation. The agency lists reduced emergency-department use, hospitalizations, overdoses and deaths among the demonstration's goals. Those are intended outcomes to be tested, not results already established by the approval document.
Recidivism claims require the same discipline. Better access to treatment may support a safer return to the community, but crime and reincarceration are influenced by housing, employment, supervision, local services and many other factors. A health waiver should not be credited with a public-safety effect until a well-designed evaluation separates those influences as far as possible.
The strongest case for these waivers is also the most measurable one: no one should lose a medically necessary treatment plan simply because responsibility changes at the prison gate. Nineteen approvals can show policy momentum. Only implementation data can show whether people actually received care before release, reached community providers afterward and avoided preventable breaks in treatment. If states publish those results with denominators, comparison methods and program differences intact, the demonstrations can be judged as health policy rather than sold as a promise.