Rhetoric vs. Beds: Inside the 2026 Detention Gap That Keeps Catch-and-Release Alive
Photo: Vladimir Menkov, CC BY-SA 3.0, via Wikimedia Commons
There is a specific moment in every enforcement briefing when the arithmetic stops cooperating with the announcement. It usually comes in a question about beds.
The 2026 policy landscape is unusually aggressive on paper. Detention mandates attached to recent legislation require custody for certain charged individuals rather than discretionary review. Removal priorities have been narrowed and restated. Interior enforcement operations have been publicized in numbers not seen in years. And yet, in jurisdiction after jurisdiction, case outcomes are being shaped not by policy but by the availability of a mattress, a courtroom slot, and an airplane seat.
The Number That Matters
For most of the past decade, Immigration and Customs Enforcement maintained an average daily detained population in the range of roughly 30,000 to 40,000 people. Appropriations for 2025 and 2026 pushed funded capacity sharply higher, with published targets in the six figures in some planning documents. What matters is not the authorized number but the operational one: beds under contract, staffed, inspected, and available.
The gap between those two figures is where enforcement credibility goes to die. A facility that is funded but not yet built, contracted but not yet licensed, or licensed but not yet staffed houses exactly zero people.
The Mandate Meets the Mattress
Statutory detention mandates are the most consequential and least discussed capacity problem in the system. When Congress requires custody for a defined class of people, it removes discretion from the officer and transfers it to the facilities planner. If the bed is not there, the officer faces an impossible choice: violate the mandate or release someone the statute says should be held. In practice, agencies develop informal workarounds — prioritizing the most serious cases, releasing the rest on conditions, and describing the result in language that sounds like policy but functions as triage.
That is catch-and-release with better branding. It is not a failure of will. It is a failure of infrastructure.
Zoning Is Where Promises Die
Every detention bed begins with a local land-use hearing, and those hearings are brutal. County commissions from the Northeast to the Mountain West have rejected or delayed proposed facilities after organized opposition over traffic, water use, property values, and, less publicly, the politics of association. Some of that opposition is legitimate civic concern. Some of it is a convenient channel for the belief that if the facility does not exist, the enforcement will not happen.
The result is a severe siting problem in exactly the regions where interior enforcement is most needed. Contractors, both private and public, cannot build faster than the permit calendar allows, and no appropriations language accelerates a zoning board.
The Courtroom Chokepoint
Detention capacity is only half the pipeline. The other half is adjudication. The immigration court backlog now stands at several million pending cases, and the shortage of immigration judges, trial attorneys, and support staff is not a hiring problem that a single appropriations cycle can close. Cases that cannot be heard cannot be completed, and cases that cannot be completed keep people in custody longer than necessary at one end of the pipeline while reprioritizing what gets heard at the other.
There is a real cost here beyond dollars. Long detention without adjudication is a due process problem, and it is the strongest argument opponents of enforcement have. The way to fix it is not to release people into an unmonitored void; it is to fund the docket.
The Return of Catch-and-Release by Default
In several jurisdictions in 2026, agents have resumed releasing people with notices to appear and check-in requirements simply because no bed was available within the statutory custody period. Alternatives to detention — ankle monitors, telephonic check-ins, case management — have expanded, and the evidence on their appearance rates is better than critics admit. But an alternative is not a substitute for custody in cases that require it, and presenting a monitoring program as though it were detention is a category error that the public has begun to notice.
What Credible Enforcement Requires
If the goal is a system that does what it says, five things have to be true, and none of them can be waived:
- Contracted bed capacity, not appropriated bed capacity. Fund the space and the staff at the same time, and report on the difference quarterly.
- Siting reform. Federal facilities face state and local vetoes that no other federal law enforcement function tolerates. Congress can preempt, and it has chosen not to.
- Judicial hiring at scale. Judges, clerks, interpreters, and courtroom space — with performance metrics on case completion, not case openings.
- Repatriation agreements. Removal requires a destination. Diplomatic work on travel documents and charter capacity is unglamorous and indispensable.
- Honest reporting. Publish releases by reason code. If people are being released because no bed exists, say so in the data rather than in a footnote.
None of this is a secret. It is in the appropriations language, the inspector general reports, and the county commission minutes. The 2026 gap between announcement and outcome is not the product of insufficient conviction. It is the product of a system in which the announcement is free and the bed is not.
Until that changes, voters should treat every enforcement promise as a budget request in disguise, and ask the only question that matters: where are you going to put them?