There are systems that, upon closer inspection, reveal something more unsettling: they are not failing, they are operating according to their incentives. This is beginning to happen with the migrant detention system in the United States. Each death in custody is usually explained as an exception: delayed medical care, a pre-existing condition, a specific omission, a human error. But when deaths accumulate in certain centers, under repeated conditions and within a policy of rapid expansion of incarceration, the interpretation can no longer be clinical or administrative. The result is a circuit that is difficult to dismantle: policy feeds the business, the business strengthens the infrastructure, and the infrastructure makes the policy more viable. This explains why the diplomatic response, although necessary, often falls short in the face of the magnitude of the phenomenon. Mexico has reacted with the tools it has: consular pressure, legal accompaniment, international activation, diplomatic notes, strategic litigation, political dialogue. Migrant detention not only fulfills an administrative function; it also fulfills a symbolic function. What is truly serious is that it occurs within a structure that had already been warned, documented, and questioned by civil organizations, strategic lawsuits, inspections, and accumulated testimony over the years. This forces an uncomfortable but necessary distinction: not all negligence is a failure. It must be structural. What we are seeing is not merely a humanitarian crisis. A significant part of the detention infrastructure is supported by private operators whose business, simply put, depends on managing bodies in custody. And there is the core of the problem. When incarceration is inserted into a logic of contracts, occupancy, operation, and profitability, the detained person ceases to be merely a rights-bearing subject and begins to become a cost unit. The problem is that the system remains useful for too many actors at the same time. Useful for those who need to show border control. Useful for those who turn detention into a contract. Useful for those who politically manage social fear of migration. Useful, even, for a public opinion trained to see the border as a threat and not as a mirror of its own democratic contradictions. And there is the real dilemma. Before the End The question is no longer just how many more people will die in custody, or how many more diplomatic notes will be sent. It is a policy that not only manages flows: it sends messages. And those messages have a political return. Harsh migration policy translates into electoral capital, into narrative discipline capacity, and at the same time, into growing contracts and budgets for an industry that has learned to prosper in the gray area between security, border, and the administration of the “other.” It produces the image of a strong state, capable of containing, classifying, and expelling. A symptom not addressed in time. Sometimes it appears as the most dangerous thing a bureaucracy can produce: the sum of normalized small omissions. But that expansion has not occurred only within the classic state apparatus. Dignity, within that design, is rarely efficient. No conspiracy is needed to cause harm. Sometimes, negligence is the predictable byproduct of a model designed to operate at the limit. The political dimension of the problem is also no minor matter. It is enough to have a system whose incentives reward containment, not care. That is why negligence does not always take the form of a major, visible violation. But it also exhibits a harsher truth: traditional diplomacy has a very narrow margin when it faces not an isolated excess, but an institutional ecosystem that functionally causes harm. Because the problem is not that Washington does not understand the seriousness of the matter. It is a protocol that is activated late. It is the exposure of a model. In recent months, migrant detention has grown significantly as a result of a policy aimed at hardening control, accelerating deportations, and expanding retention capacity. Mental health costs. A transfer that does not arrive. A file that reduces a life to a clinical note. The relevant thing is not only that this happens. There are systems that collapse. An occupied bed generates income; medical care, on the other hand, represents an expense. It is a civilizational issue. And at that point, continuing to call it “negligence” can be an elegant way of not naming what it really is. A model. The underlying question is another: how compatible can a democracy be with a system that has begun to accept death as a tolerable operational cost? Because that is the turning point. When life in custody ceases to be a material priority and becomes merely a rhetorical obligation, the problem is no longer administrative. Oversight costs. All of that is correct. All of that must be done.
