Texas Spent $1.35 Billion on Psychiatric Capacity. Its Forensic Waiting List Ended the Year Where It Started.

Texas Health and Human Services published its annual waiting-list report on November 6, and the number that traveled was 253. That is the average wait, in days, for a non-maximum-security forensic psychiatric bed at Montgomery County Hospital in the fourth quarter of fiscal 2026 — roughly eight and a half months between a judge's commitment order and a hospital admission.

The report runs 27 pages and answers a single legislative instruction: Rider 47 of the 2026-27 appropriations act requires the commission to tell the Legislative Budget Board and the governor's office how many people are waiting for mental health services, for how long, and what the state spent trying to shorten the lists. Read past the Montgomery figure and the document describes something more uncomfortable than a hospital shortage. It describes a system that spent $1.35 billion on forensic psychiatric capacity in one year and closed that year with 1,626 people still in the queue.

The list that ended the year where it began

At the end of the first quarter, 1,266 people were waiting for a non-maximum-security forensic bed and 468 for a maximum-security bed. By the fourth quarter the first figure was 1,149 and the second was 477. The non-MSU list fell 9 percent across the year. The MSU list rose 2 percent, and its fourth-quarter reading was the highest of the four snapshots.

Those are point-in-time counts, and the commission is careful to say so. What moves underneath them is volume. In the fourth quarter alone, 838 people came off the non-MSU list — admitted, resolved, found competent while sitting in jail, or otherwise disposed of — against a standing queue of 1,149. The queue is not frozen. It is being worked constantly and refilled at nearly the same rate.

The facility-level numbers make the pattern legible. Austin State Hospital's list went from 277 to 217 over the year. Terrell went from 245 to 238, San Antonio from 193 to 191, Rusk from 123 to 108. Two of the largest facilities barely moved at all, which is what a capacity ceiling looks like when demand is steady. And two others moved the wrong way: El Paso Psychiatric Center climbed from 74 to 85, and Big Spring State Hospital from 38 to 61.

Two facilities opened during the year — Permian Basin Behavioral Health Center began admissions on August 12, 2026, and the Texas Behavioral Health Center in Dallas on June 30. Both appear on the fourth-quarter table. Neither arrival bent the statewide line.

The same commitment, eight times the wait

A forensic commitment is a court order, not a preference. The person has been found incompetent to stand trial or not guilty by reason of insanity, and the state is obliged to provide the treatment that will restore competency or hold them under civil commitment. Where they live determines how long that takes.

In the fourth quarter, the average wait at El Paso Psychiatric Center was 337 days. At Terrell State Hospital it was 294, at Montgomery County 253, at San Antonio State Hospital 249, at Dallas Behavioral Hospital 240. Move down the same table and the numbers collapse: Big Spring 127 days, John S. Dunn Behavioral Sciences Center 101, Austin State Hospital 100, Wichita Falls 99, Harris County Psychiatric Center 93, Kerrville 42.

That spread — 42 days at one state facility and 337 at another, for the same category of order — is the most consequential figure in the report, and it is not a story about hospitals being slow. It is a story about catchment areas, special-population beds and the uneven distribution of contracted capacity. A person committed in El Paso waits eight times as long as a person committed in Kerrville.

The averages also wobble in ways that argue against reading any single quarter too hard. Rio Grande State Center, which does not maintain its own forensic list but accepts transfers, reported 471 average days in the second quarter, 378 in the third and 87 in the fourth. Small denominators and shifting case mix produce those swings. The direction of the year, not the quarter, is the usable signal.

What the money bought

The report's expenditure table totals $1.35 billion in forensic-related spending for fiscal 2026: $673.3 million for state hospital operations, $174.0 million for contract beds, $485.9 million for construction, $9.6 million for deferred maintenance, $4.2 million for emergency repairs, and $813,572 for the Forensic Support Team pilots in Dallas and Travis counties.

Nearly half of that — the construction line — is going into buildings that do not open this year. The Wichita Falls replacement, $452 million for 225 beds, is scheduled for winter 2027. Terrell's replacement, $573 million for 275 beds, for fall 2027. Fifty new maximum-security beds at Rio Grande, $120 million, fall 2027. A 75-bed hospital in Amarillo, $159 million, winter 2027. A 50-bed maximum-security hospital in Lubbock, $121 million, winter 2027. The conversion of a 40-bed hall at San Antonio State Hospital into maximum-security space, $15 million, is the only one expected before the year turns. Separately, $239 million in grants will build up to 332 beds, half of them forensic.

Then there is the maintenance line, which is where the arithmetic gets strange. The commission puts its deferred maintenance need at approximately $650 million. The Legislature appropriated $98 million for it in the 2025 supplemental budget. The state actually spent $9.6 million on deferred maintenance and $4.2 million on emergency repairs during the year — about 1.4 percent of the identified need. Beds come offline for exactly this: state hospitals must meet the Life Safety Code, and anti-ligature remediation removes beds from service while it happens.

Beds are not the binding constraint

A hospital bed is a slot in a flow, and the report describes three forces that govern the flow more directly than the bed count does.

The first is unscheduled arrivals. Across the four quarters, state hospitals logged 529 to 581 walk-ins each quarter — referrals from law enforcement, families or the patients themselves — and admitted 358 to 389 of them. Terrell alone fielded 323 walk-ins in the fourth quarter and admitted 251. Because hospitals operate under EMTALA, a person who needs crisis stabilization must be assessed and stabilized. Every walk-in admission is a bed not available for a planned forensic admission.

The second is discharge velocity. Discharge planning starts on admission, and the commission tracks it with a form completed at each 90-day clinical review, the Patient Needs Discharge Form. Some patients no longer need acute inpatient care but still need residential placement, supervision or medical support that does not exist for them, so they stay. Transition support specialists moved 192 people — 159 adults and 33 youth — to less restrictive settings as of July 31, 2026. Each of those is an admission slot created without pouring a yard of concrete.

The form that sets a floor on length of stay

The 90-day review cycle is not merely administrative. Because discharge readiness is assessed at those intervals, the reporting structure itself places a floor under how quickly a stabilized patient can be identified and moved. A system that measures discharge barriers every 90 days and then cannot place the patient in the community has converted a clinical decision into a housing problem, and the waiting list absorbs the difference.

The third force is the composition of the queue. A study published in April 2026 by the Texas Institute on Excellence in Mental Health followed 2,998 forensic patients and found that 623 of them had three or more separate waitlist entries, accounting for 2,058 records — 9 percent of all waitlist activity — with repeat cycling associated with lower-level offenses. A small population generates a disproportionate share of the workload, and those are the cases where an outpatient alternative should be easiest to justify.

The cheapest line items do the most work

Buried in the last pages of the report is the only direct comparison of cost per outcome that the state publishes, and it points away from hospitals.

Jail-based competency restoration served 707 people in fiscal 2025 and restored 372 of them, a 52 percent rate, with 53.6 percent restored within 60 days and only 169 needing admission to an inpatient facility. It cost $3.7 million in fiscal 2026, which works out to roughly $9,970 per restoration.

Outpatient competency restoration served 337 people, restored 149 — 44.2 percent — and discharged 324, with an average length of stay of 74 days for misdemeanor charges and 133 days for felonies. At $4.2 million, that is about $28,100 per restoration.

The Forensic Support Team pilot, launched in April 2025 in Dallas and Travis counties, monitors people on the waitlist, arranges re-evaluations, pursues diversion and coordinates with courts and jails. It removed 239 people from the forensic waitlists. Its entire fiscal 2026 budget was $813,572 — around $3,400 per removal.

Add the three programs together and they cost $19.0 million, about 1.4 percent of what the state spent on forensic psychiatric capacity. The report notes the constraint plainly: some counties have no access to outpatient or jail-based restoration at all, so people who could be restored in a less restrictive setting sit on a state hospital list instead. That is not a clinical judgment failing. It is a procurement gap.

Where substance use disappears from the data

Nowhere in 27 pages does the report disaggregate the forensic population by substance use. There is no count of how many people awaiting competency restoration have an opioid, alcohol or stimulant use disorder, no measure of whether treatment for it shortens restoration time, and no line item for addiction treatment inside the forensic budget.

That omission is not neutral. Competency restoration is, in practice, often a question of whether a person can be stabilized enough to participate in their own defense, and untreated substance use is one of the most common reasons that stabilization fails. The frequent-utilizer finding points at the same population: 623 people cycling through the system three or more times, mostly on lower-level offenses, is the profile where dual diagnosis care would be expected to change the trajectory. The state measures the queue without measuring the condition that fills it.

What the 2027 beds will and will not fix

When the six construction projects finish, Texas will have added roughly 715 beds, most of them forensic, most of them arriving within about eighteen months of each other. Set that against the current flow: 838 non-MSU removals in a single quarter. The new capacity is real and it is not trivial. It is also approximately two quarters of the system's existing throughput, delivered all at once, into a system whose admission rate is currently governed by walk-ins and discharge placement.

Which means the number to watch next November is not the bed count. It is the average days figure at El Paso, and the deferred-maintenance line, and whether the jail-based and outpatient restoration programs — the ones that cost $19 million and remove people from the queue at a few thousand dollars each — get funded at a scale that matches what the report shows they do. Families looking for levels of care below the state hospital will find the answer to their question in those three lines, not in the construction schedule.