Legal rights do not disappear simply because they become difficult to exercise. In American immigration detention, the state can move people, expand facilities and exercise custody faster than lawyers, courts and inspectors can respond. The danger is not always power operating outside the law, but power moving beyond the practical reach of those meant to hold it accountable.
On June 26, 2026, a federal judge sentenced David Courvelle, a former contract detention officer at the South Louisiana ICE Processing Center, to three years in prison. Courvelle had pleaded guilty to sexually abusing a woman held in federal immigration custody; according to the Justice Department, he had also used other detainees to help coordinate and conceal his conduct. The case offers an obvious story about abuse of authority. It also offers a less obvious lesson about accountability: Courvelle was investigated, prosecuted, convicted and punished. The law did not disappear.
That matters because the most dramatic description of immigration detention is not necessarily the most accurate one. Calling detention centres “lawless zones” may capture the experience of someone who cannot reach a lawyer, challenge a transfer or obtain a timely remedy, but it obscures the institutional reality. Immigration detention in the United States is surrounded by statutes, regulations, detention standards, inspectors, contract monitors, immigration judges, federal judges, congressional oversight and internal complaint systems. The Department of Homeland Security itself has mapped 17 headquarters and component entities with detention-oversight responsibilities, including bodies charged with investigating complaints, monitoring contracts, reviewing conditions and investigating misconduct.
The harder question is therefore not whether law exists inside the system. It is what happens when the state can arrest, confine, transfer and expand detention capacity faster than the institutions designed to make rights usable can respond. A person can possess a right to representation without having a lawyer, a right to seek judicial review while being transferred hundreds of miles from potential counsel, and the protection of detention standards inside a facility that began housing people before a required inspection had occurred. Oversight may exist in several places at once and still fail to produce correction at the moment correction matters most.
This is the accountability gap at the centre of American immigration detention. It is not best understood as a space outside law, but as a system in which the distance between a formal right and a usable remedy can become an instrumentally important fact. In such a system, geography matters, but so do time, institutional capacity, contract design, information, legal representation and the speed with which the machinery of enforcement can move. The constitutional question is not exhausted by asking whether a remedy eventually exists. It also includes a more practical question: can the person reach it while it is still capable of changing what happens next?
The Case That Exposed the Wrong Question
Courvelle’s conviction is useful precisely because it resists the easiest argument. If a contract officer commits a federal crime against a detainee and is later prosecuted by the federal government, the episode cannot honestly be used as proof that detention personnel are categorically immune from law. The Justice Department’s case demonstrates the opposite: criminal accountability can reach inside a contracted immigration facility, even when the perpetrator is not a direct ICE employee.
But retrospective punishment and preventive accountability are not the same thing. A prosecution answers what happens once misconduct has been discovered, investigated and converted into an evidentiary case. It does not by itself answer how quickly a detainee can report abuse, whether confidential communication is available, whether staff supervision detects patterns early, whether complaints reach an independent body, or whether contract and facility structures make warning signs visible before a criminal case is necessary.
That distinction changes the analytical centre of the story. A system can punish some offenders and still contain conditions in which detecting misconduct is difficult. It can maintain formal grievance procedures while requiring the person who needs them to navigate language barriers, confinement, institutional hierarchy and dependence on the same facility for communication, medical care and daily necessities. None of those facts proves that a particular complaint will be ignored. Together, however, they determine how costly it is for a right to move from paper into action.
The more revealing measure of accountability is therefore not simply whether someone can be punished at the end. It is how many institutional steps separate the vulnerable person from an actor capable of stopping the problem earlier. Courvelle’s sentence demonstrates that law can eventually arrive. The question for the detention system is what has to happen before it does.
The State Is Present. The Remedy Can Still Be Distant
The image of an unregulated detention network is difficult to reconcile with the amount of formal oversight surrounding ICE custody. DHS’s own review identified 17 departmental and component entities with detention-oversight roles. Some inspect facilities, some receive complaints, some investigate professional misconduct, some monitor health care or contractual compliance, and some examine broader civil-rights or inspector-general concerns. Four of those entities, DHS said, had congressional mandates to receive and investigate specified complaints or allegations concerning detention conditions and treatment.
Institutional abundance, however, is not the same thing as remedial speed. A complaint system can exist without being the mechanism capable of preventing tomorrow’s transfer. An inspector can document a failure after a site visit without possessing authority to decide an individual bond question. A contract monitor can identify a performance deficiency but cannot function as the detainee’s lawyer. A federal judge can order relief, but only once a dispute has reached a court in a form the court can hear.
The distinction is between oversight of the institution and reachability of a remedy by the individual. The first asks whether the system contains bodies assigned to detect and correct failures. The second asks what a detained person must do, and how quickly, to activate a right relevant to his or her particular situation. A detention system may perform better on the first measure than on the second, or the reverse. Treating the two as interchangeable conceals the central problem.
This is why the number of oversight bodies can be misleading in both directions. Critics who say there is no oversight erase real institutions, investigations and corrective mechanisms. Officials who point to the existence of those institutions do not thereby prove that remedies are accessible at the speed required by confinement. The constitutional significance lies in the gap between capacity on paper and capacity in time.
Civil Detention, Carceral Power
Immigration proceedings occupy an unusual place in American law. Removal proceedings are civil rather than criminal, and immigration detention is generally justified as regulatory rather than punitive: to ensure appearance at proceedings, facilitate removal where authorised, or serve other immigration purposes established by statute.
The Supreme Court has repeatedly recognised broad congressional authority over immigration detention while also recognising that noncitizens within the United States are protected by the Fifth Amendment’s Due Process Clause. In Zadvydas v. Davis, the Court stressed that freedom from physical restraint lies at the core of protected liberty and interpreted federal law to avoid authorising potentially indefinite post-removal detention.
That combination creates a tension easily lost in political debate. The system is civil in legal classification, but the power being exercised is physical confinement. The doors lock regardless of whether the proceeding is labelled criminal or administrative. A person may be detained while litigating whether removal is legally permissible, while seeking asylum or another form of relief, or while disputing whether the government may continue detention without a bond hearing.
The Supreme Court has not treated every form of immigration detention as constitutionally suspect. In Demore v. Kim, it upheld mandatory detention for a defined category of removable noncitizens with specified criminal convictions during the limited period of their removal proceedings. Yet the boundaries around prolonged detention continue to generate litigation.
In June 2026, the Supreme Court agreed to review an administration appeal involving prolonged immigration detention and whether due process required a bond hearing. That particular case, Genalo v. G.M., was dismissed on September 11, 2026, after the parties jointly requested its dismissal following a pardon that removed the basis for the respondent’s detention. The Supreme Court therefore issued no decision on the merits, leaving the underlying Second Circuit ruling in place. The broader constitutional dispute over prolonged immigration detention and access to bond hearings remains a subject of litigation, but that particular appeal can no longer be described as awaiting a Supreme Court ruling.
The relevant point is not that civil detention is illegitimate because it resembles incarceration, nor that its civil character makes its procedural safeguards unimportant. It is that a system exercising one of the state’s most consequential powers—the power to confine a person—operates through a legal framework different from criminal prosecution. That difference affects who receives appointed counsel, how custody is reviewed, which courts have jurisdiction over particular claims and how quickly legal challenges can be assembled. The classification is therefore not a semantic detail. It structures the routes through which liberty can be recovered.
A Right to Counsel Is Not the Same as Having Counsel
American immigration law gives people in removal proceedings the privilege of being represented by counsel of their choosing, but at no expense to the government. The Justice Department states the distinction plainly: unlike criminal defendants who qualify for appointed counsel, respondents in immigration proceedings generally must represent themselves, retain a lawyer, find an accredited representative or secure pro bono assistance. Immigration courts provide lists of pro bono providers, but inclusion on such a list does not mean an organisation is able to accept every case.
This may be the cleanest example of rights without reach because nothing has to be formally denied for the gap to emerge. The government can accurately say that the person is allowed to have a lawyer. The person can accurately say that he has no lawyer. Both statements can be true simultaneously. Detention increases the importance of everything that lies between them. A non-detained person can search for counsel, travel to an office, gather documents, make repeated calls, obtain help from family and remain in one location while the relationship with a lawyer is established. A detainee may depend on facility phone systems, legal-visitation schedules, remote communications, family intermediaries and the availability of lawyers willing and able to represent someone held in a particular jurisdiction.
ICE’s 2025 National Detention Standards require facilities covered by those standards to permit private meetings with legal representatives and provide extensive legal-visitation hours, demonstrating that legal access is recognised as an operational requirement of detention itself. Technology complicates rather than simply worsens this picture. A July 2026 GAO study found that about 73 percent of immigration hearings involving detained respondents from fiscal years 2022 through 2025 were conducted remotely.
The judges and attorneys GAO interviewed generally described remote hearings favourably, citing flexibility, lower costs and, in some circumstances, improved access to private attorneys, while also identifying technology and interpretation challenges and circumstances in which in-person proceedings could be preferable. That evidence matters because it prevents a lazy conclusion that digital justice is inherently inferior. Remote technology can reduce geographic barriers and make legal participation easier. Yet technology cannot substitute for representation itself, and a functioning video connection does not solve the problem of finding counsel, preparing a case or communicating confidentially before the hearing. The difference between “access to a legal process” and “capacity to use that process effectively” remains the distinction that matters.
The State Can Move Faster Than the Right
In February 2026, a federal judge in Minnesota found that ICE practices during a large enforcement operation had unconstitutionally obstructed detainees’ access to lawyers. The dispute involved inadequate telephone access and rapid transfers out of state, among other conditions. The court ordered measures designed to secure private attorney communication and temporarily restrict rapid transfers so detainees would have an opportunity to obtain legal assistance. ICE disputed allegations that it had intentionally violated detainees’ rights, but the ruling exposed a structural problem independent of intent: custody can move faster than representation.
Transfer is administratively ordinary. A national detention network must move people for available bed space, court schedules, medical needs, operational requirements and eventual removal. Yet a transfer does more than change a bed assignment. It can change proximity to family, the practical reach of a lawyer, the organisations capable of offering assistance and, in some circumstances, the federal judicial district in which a habeas petition concerning detention is pursued.
The resulting friction does not require a conspiracy to exist. An enforcement bureaucracy is designed to process custody efficiently; legal representation is organised around human relationships, document gathering and jurisdiction. If one side of the system can move a person overnight while the other needs days to locate, consult and represent that person, speed itself creates an asymmetry. The right has not vanished, but its route has become longer while the state’s route has become shorter.
This is why geography should be understood institutionally rather than literally. A detention centre may be physically close to a city and still be difficult to reach through overloaded phone systems, restricted visiting windows or a shortage of available lawyers. Conversely, remote hearings and secure legal communication can sometimes make a distant facility more legally accessible than geography suggests. What matters is the effective distance between the detainee and the person or institution capable of changing the legal situation.
The temporal dimension is even more important. A successful legal challenge after a transfer, prolonged detention or removal may establish an important principle, compensate a wrong in some circumstances or prevent recurrence. It cannot always restore the weeks or months during which the person remained confined, nor can judicial review easily reconstruct opportunities lost because counsel entered the case too late. In detention, time is not merely the background against which rights operate. It is part of the value of the right itself.
Oversight Exists—But What Does It Correct?
The federal government’s own inspection data reveal why counting inspections cannot settle the oversight question. GAO examined four DHS entities that inspected immigration detention facilities and found that they approached facilities with different purposes and methods. From fiscal years 2022 through 2024, ICE’s Office of Detention Oversight rated facilities acceptable or better in 238 of 241 inspections, even while identifying deficiencies involving matters such as sanitation, food service and environmental health.
The Office of the Immigration Detention Ombudsman produced a very different-looking record. Its inspections were often triggered by complaints or specific concerns rather than being general compliance reviews, so the numbers are not directly comparable. In 31 inspections during the same period, OIDO found compliance with the specific standards under review in only two and identified 174 deficiencies across the remaining 29, with medical care representing the largest category. The difference does not demonstrate that one office was truthful and another was not; it demonstrates that what an inspection finds depends partly on what it is designed to look for.
That is the analytically interesting part. “The facility passed inspection” sounds like a comprehensive judgment when it may instead mean that a particular inspector applied a particular set of standards through a particular methodology at a particular time. A complaint-driven inspection asks a different question from a recurring compliance review. A medical inspection sees something different from a contract audit. The word compliance can therefore conceal a surprisingly complex measurement problem.
GAO’s criticism focused precisely on that institutional layer. It found that inspection bodies lacked sufficiently defined performance goals and measures to assess whether their inspection programmes were achieving their intended outcomes. The problem was not that inspectors were doing nothing; it was that government lacked a robust way to determine the effectiveness of the inspection architecture itself.
The situation became still more revealing in 2026. DHS told GAO that an appropriations law signed on April 30 had eliminated OIDO and that the department therefore did not plan to implement one of GAO’s recommendations concerning the office. GAO responded that the Ombudsman for Immigration Detention remained a position established by statute under 6 U.S.C. § 205 and kept its recommendation open. That is almost a miniature version of the article’s thesis. An oversight function can remain embedded in law while its operational capacity becomes uncertain or diminished. The legal institution and the practical institution are related, but they are not identical. Accountability depends on both.
Fragmented Accountability
ICE detention is not one type of building run through one administrative chain. Facilities can be operated directly by federal authorities, private companies, state or local governments, or through contractual arrangements connecting several of those actors. Health care may be provided by ICE Health Service Corps personnel in one facility and by local or contracted providers in another, with federal oversight layered over the delivery system. GAO’s inspection work explicitly distinguishes among those arrangements because responsibility for providing and monitoring services changes from facility to facility.
Fragmentation is not inherently evidence of failure. Governments contract for services across the public sector, and local partners may possess infrastructure or expertise the federal government does not want to reproduce everywhere. Private provision can sometimes add flexibility, and a distributed network can respond to rapidly changing detention needs more easily than a wholly federal estate.
The accountability problem emerges when responsibility is divided faster than authority can be traced. A detainee experiencing a medical failure may encounter facility staff, a private health provider, ICE officials and contract requirements that assign different duties to different actors. A congressional investigator may need to determine whether a failure arose from the standard itself, the contract, the contractor’s performance or federal monitoring. A corrective action that reaches only one layer can leave the underlying relationship unchanged.
DHS’s 17-entity oversight map illustrates the same complexity from the government side. Specialisation can improve accountability by ensuring that criminal misconduct, civil rights, medical care, contracting and detention standards are examined by offices with relevant expertise. It can also produce boundaries: one office can identify a problem whose remedy belongs to another. The existence of many watchers does not necessarily produce one institution that owns the full outcome.
This is the difference between fragmented responsibility and absent responsibility. The former is more subtle because every participant may be able to point to a real duty being performed somewhere in the system. The question becomes whether the chain is clear enough that a failure can be followed from complaint to correction without disappearing into the spaces between mandates.
Congress Can Enter. Why Does Access Still Become a Conflict?
Congressional oversight provides an unusually clear example of the gap between legal authority and operational friction. Section 527 of the 2024 DHS appropriations law prohibited the use of department funds to prevent members of Congress, or designated congressional employees, from entering DHS facilities used to detain or house noncitizens for oversight. The provision also specified that members of Congress themselves could not be required to provide advance notice, while DHS could require up to 24 hours’ notice from designated congressional staff.
The legal protection did not prevent a prolonged dispute over how congressional access would work in practice. ICE’s public congressional-relations page described a requirement that visit requests be submitted at least seven calendar days in advance, with shorter notice requiring approval from the DHS secretary. Members of Congress reported denied or delayed visits, and the administration sought to enforce advance-notice requirements that lawmakers challenged in federal court.
The conflict subsequently produced an important judicial correction. In December 2025, a federal judge temporarily blocked the earlier advance-notice policy. After the administration issued a renewed requirement in January 2026, Judge Jia Cobb again intervened, and on March 2 stayed the revised policy while litigation continued. The court found that the challenging lawmakers had shown a substantial likelihood of success in arguing that the requirement conflicted with federal law and exceeded the department’s authority.
ICE’s publicly accessible congressional-relations guidance nevertheless continued to display the seven-day scheduling language. That should not be mistaken for an enforceable requirement overriding the court’s decision. It reveals a different problem: public-facing administrative instructions, statutory protections and judicial restrictions can send conflicting messages even when the formal legal position has been clarified. The right to inspect may exist, and a court may protect it, while the process of obtaining access remains contested at the facility gate.
The significance is not that Congress lacks oversight authority; the statutory text is evidence that it possesses it. Nor does every scheduling rule automatically establish unlawful obstruction, since particular facilities, categories of site and security circumstances can generate legal and factual disputes. What the conflict shows is that even a comparatively explicit oversight right must still pass through operational rules, institutional interpretation and people responsible for admitting visitors.
That distinction matters because democratic oversight is often discussed as though authority automatically becomes access. In practice, the space between them can determine what oversight sees. An announced visit can still reveal substantial information, but unannounced access serves a different purpose when Congress wants to observe ordinary conditions rather than conditions prepared for a visit. The law itself recognises that distinction by addressing advance notice.
The same tension applies beyond Congress. Public accountability depends on records, court filings, journalists, lawyers, inspectors and civil society being able to observe enough of the detention system to test official descriptions against lived conditions. A detention network can comply with formal reporting requirements and remain difficult to understand in real time simply because information moves more slowly than enforcement. Transparency, like representation, has a temporal dimension.
When Enforcement Capacity Outruns Accountability Capacity
The scale of American immigration detention changed rapidly after January 2025. According to ICE data reported by GAO, the average daily detained population rose from 39,314 on January 20, 2025 to 67,204 by April 1, 2026, an increase of 71 percent. Over the same period, the number of authorised ICE detention facilities increased from 134 to 239, a rise of 78 percent. A 2025 law provided ICE about $75 billion through fiscal year 2029, including $45 billion specifically for detention capacity.
The expansion continued alongside a sharp rise in enforcement activity. Reuters reported that ICE made nearly 51,000 arrests in August 2026, the third consecutive monthly record, while removals remained around 1,200 per day. More than three-quarters of the recent arrests reviewed by Reuters involved people without criminal records, and many cases involved individuals with pending immigration claims or no immediately executable removal order.
Those numbers do not by themselves demonstrate an accountability failure. Congress and an administration can lawfully choose to devote substantially more resources to immigration enforcement, and an increase in detention is not evidence that detention standards have been violated. The institutional question is comparative: what happens when enforcement capacity can be funded and expanded faster than legal, medical, inspection and oversight capacity grows around it?
Camp East Montana at Fort Bliss offers unusually concrete evidence of what such a mismatch can look like. The Army and ICE accelerated the procurement and construction of what became ICE’s largest detention facility, operating under a $1.3 billion contract. GAO concluded that the expedited schedule damaged planning and acquisition, that the facility opened without a pre-occupancy inspection required by ICE policy, and that it initially lacked features including perimeter security cameras, outdoor recreation space, and space for attorney and family visits. Subsequent problems included medical-service gaps, unsanitary conditions and the loss of a loaded firearm.
The lesson is more precise than “rapid expansion causes abuse.” GAO did not establish such a universal rule, and many facilities can expand without reproducing Camp East Montana’s failures. What the case demonstrates is that physical custody capacity can be created before every safeguard designed to accompany custody is fully operational. The state can acquire the ability to confine first and then force the systems around confinement to catch up.
That sequencing is the institutional risk worth watching. A government can hire enforcement personnel, lease buildings or convert warehouses through large appropriations. Training inspectors, creating attorney-access infrastructure, establishing reliable medical systems, staffing courts and building durable complaint mechanisms involve different budgets, agencies and timelines. If those curves do not rise together, the result is not a lawless system. It is a system in which the capacity to exercise power temporarily outruns the capacity to check it.
The Contractor Question Is About Chains of Responsibility
Private detention is often debated through a single moral claim: that profit creates an incentive to detain more people or reduce conditions. That argument can be politically powerful, but by itself it is too broad to explain how accountability actually succeeds or fails. A government-run facility can perform badly; a contractor can meet demanding standards. The analytically useful questions concern contract design, monitoring independence, penalties, information and who is able to compel correction.
GAO’s long-running work on ICE contracts shows why those details matter. In a 2021 review, it found that ICE had increasingly used guaranteed minimum payments under which operators were paid for a fixed number of detention beds whether or not the beds were occupied. ICE spent $20.5 million in May 2020 on more than 12,000 unused beds per day, on average. GAO also found that officers responsible for contract oversight in eight of 12 field offices it interviewed reported concerns that management interfered with their ability to exercise contract-enforcement responsibilities independently.
The picture has changed since then, which is precisely why old criticisms should not be repeated mechanically. ICE told GAO in 2024 that it had decided to phase out guaranteed minimums and move toward a facility-operating-charge structure that could allow costs to be reduced or renegotiated when actual use fell below expectations. As of June 2026, however, GAO said ICE had not yet provided enough documentation to demonstrate full implementation of the new approach, and GAO’s recommendation concerning the independence of contract oversight remained open.
Camp East Montana illustrates the same issue through a newer contract. GAO found that the arrangement required the government to pay for meals and operating services based on capacity for 5,000 people even when the detained population was far lower—about 1,600 at the end of February 2026. The agency estimated that different pricing arrangements could save tens of millions of dollars, while also documenting the broader operational problems associated with the facility’s accelerated opening.
Contract economics therefore belong in an accountability article not because every dollar of profit proves mistreatment, but because contracts distribute incentives and authority. If the government pays for fixed capacity, specifies standards, delegates daily operations and depends on its own contract monitors to identify breaches, then the quality of public power partly depends on the architecture of a private agreement. The detention decision remains governmental. The day-to-day conditions through which that decision is experienced may not be.
A Zone of Exception Does Not Need to Be Lawless
There is still something useful in the language of “exception,” provided it is used carefully. A genuine constitutional state of exception implies extraordinary legal powers or suspension of ordinary rules, and applying that label casually to the entire immigration detention system would obscure too much. ICE detention is authorised by statutes, constrained by courts, subject to standards and inspected by government bodies. Detainees regularly litigate, sometimes successfully, against the government.
Yet exception can also describe a practical condition rather than a formal constitutional category. A person may possess rights that become unusually difficult to activate because confinement concentrates control over movement, communication, location and time in the same institution whose decisions the person may need to challenge. What is exceptional is not the disappearance of law but the dependence of the person on institutional routes for reaching it.
This is why immigration detention can generate accountability gaps without requiring anyone to declare an area outside the Constitution. The gap can emerge from ordinary administrative acts: a transfer completed before counsel is secured, a facility opened before required inspection, a complaint divided among several offices, a congressional visit stalled by conflicting rules, or an oversight body that remains recognised in statute while its organisational capacity changes. Each action can be bureaucratically legible on its own. Their combination determines whether legal protection operates at the speed of detention.
The strongest evidence against describing the system as simply lawless is also evidence for taking the accountability gap seriously. Courts have intervened. Inspectors have found deficiencies. GAO has issued recommendations. DOJ prosecuted Courvelle. These are not meaningless exceptions to an otherwise unregulated machine; they are parts of the constitutional system functioning as checks. The uncomfortable question is whether those checks operate early enough and consistently enough to shape the exercise of power rather than merely document its consequences. A right that can be vindicated only after the decisive event is still a right in law. It is not necessarily the same protection in life.
Accountability That Arrives Too Late Is a Different Kind of Accountability
Immigration politics encourages a false choice between two descriptions of detention. One depicts an authoritarian machinery operating beyond law; the other points to statutes, judges, inspections and appeal rights and concludes that constitutional safeguards are therefore intact. The actual institutional landscape is more difficult. The United States has a dense legal architecture around immigration detention, and that architecture can produce genuine accountability. It can also leave substantial gaps between the moment state power acts and the moment another institution is able to review what it has done.
Those gaps become more important as the system expands. From January 2025 to April 2026, detention capacity and the detained population rose at rates that would be significant for any large public institution. The important question is not simply whether Congress supplied money for beds, but whether legal access, medical capacity, inspection quality, contract supervision, courts and independent oversight can scale alongside them. Camp East Montana provides a warning about what can happen when construction and custody are accelerated while parts of the surrounding safeguard architecture lag behind.
The same principle explains why counsel matters more than the abstract existence of the right to counsel, why transfer speed matters to habeas review, and why inspection numbers require interpretation rather than celebration or dismissal. Rights become effective through institutions, and institutions operate at different speeds. The capacity to write a right into law can be separated by years from the capacity to build the lawyers, procedures, data systems and oversight bodies that make it usable at scale.
A constitutional system should be judged partly by what happens after power is abused. Prosecution, judicial review, damages, inspector-general investigations and legislative scrutiny all matter. But a system of liberty cannot rely exclusively on retrospective correction, because confinement creates injuries that later victories cannot always reverse. Time spent unlawfully detained is not returned by a favourable ruling; access to a lawyer after a decisive transfer is not equivalent to access before it; a facility corrected after inspection is not the same facility experienced by those who were held there before the correction.
That does not make every delay a constitutional violation, every transfer an obstruction or every detention illegitimate. It does establish a more demanding standard for evaluating institutional power. The question is not merely whether the state can eventually explain what it did or whether a court can eventually decide whether it was lawful. It is whether the channels of accountability remain close enough to the exercise of power that they can still alter its course.
The danger inside a heavily regulated detention system is therefore not necessarily that law disappears. It is that law remains visible while becoming progressively harder to reach from the position of the person who needs it. The state can preserve the language of rights, maintain inspection offices, publish detention standards and still create a practical hierarchy between the speed of enforcement and the speed of remedy.
Nor is the underlying problem confined to immigration detention. Wherever a government can carry out a coercive decision more quickly than the affected person can obtain independent review, the practical force of a right depends partly on the timing of the remedy. A constitutional government does not prove its commitment to rights only by prosecuting an officer after a crime has been uncovered or by releasing a detainee after a judge rules that custody was unlawful. It proves that commitment more fully when the person subjected to its coercive power can reach a lawyer, an inspector, Congress or a court while those institutions still have time to matter.
The critical line is crossed not only when a right is abolished, but when the machinery of the state learns to move so much faster than the machinery of accountability that the right arrives after the event it was meant to govern.
