The United States Has Built a Detention System Without a Name. This Is Its Blueprint.
Nobody expects the Spanish Inquisition, or the concentration camp.
The archipelago does not announce itself. It is assembled from procurement documents, court orders, sheriff’s agreements, and airline manifests. It is not a single facility, a single statute, or a single agency. It is a pipeline—a set of modular legal channels and converted infrastructure that can receive any population the enforcement apparatus selects, without requiring a new law, a new category, or a new public debate. Its domestic capacity is measured in hundreds of thousands of beds. Its foreign capacity is measured in dozens of third countries that have agreed to hold people the United States cannot legally send home. Its legal form is retrospective: the government acts, and the person challenges afterward, if they are still within reach of a court.
What follows is not a moral claim. It is a reading of the machinery, and the machinery has a grammar. A group is excluded from the positions that carry status. The exclusion is written into law or administration. The representation contracts. The apparatus is built. Then the question is whether the apparatus is turned on the target. The apparatus is built before it is aimed. That is the point of building it.
The distinction that matters is between machinery that is still reversible and machinery that has become a project. Everything before the final step is reversible, because the capacity exists and the law has not yet been arranged to permit it. The final step is the point of no easy return, because the apparatus is now doing what it was built to do. The United States has moved further along that sequence than any other liberal democracy, and the reason is not intention. It is the form of its statutes.
PART ONE: THE DOMESTIC PIPELINE
I. Capacity, Conversion, and the WEXMAC Bypass
The domestic apparatus is not a network of purpose-built immigration prisons. It is a conversion pipeline that takes existing infrastructure—warehouses, state prisons, military bases, federal prisons—and reclassifies it as detention capacity, while a parallel procurement vehicle bypasses the ordinary acquisition rules that would otherwise slow the conversion.
The domestic apparatus is not a network of purpose-built immigration prisons. It is a conversion pipeline that takes existing infrastructure—warehouses, state prisons, military bases, federal prisons—and reclassifies it as detention capacity. It runs on five distinct channels, and the channels do not share a single procurement route, a single legal authority, or a single agency.
The capacity target is explicit. A February 2026 plan, the Detention Reengineering Initiative, sets a goal of 92,600 beds by the end of 2026, at a cost of $38.3 billion** (Boston Globe, 2026; Deccan Chronicle, 2026). The plan involves purchasing warehouses and converting them into “immigration jails,” with the largest sites holding between 1,500 and 10,000 people each. ICE purchased **11 warehouses for $1.07 billion between January and April 2026, and by June had decided to sell seven of them, after spending more than $20 million on nonrecoverable costs (Washington Post, 2026).
The military-base channel was partially realized and largely abandoned. A $1.3 billion soft-sided tent complex at Fort Bliss in Texas became the nation’s largest immigration detention center. Plans for Joint Base McGuire-Dix-Lakehurst in New Jersey and Camp Atterbury in Indiana were announced but did not move forward (CBS News, 2026).
The WEXMAC channel is the structural reason the pipeline can move faster than ordinary procurement rules would allow. The Worldwide Expeditionary Multiple Award Contract, a Navy-administered vehicle designed for expeditionary logistics, was expanded from a $10 billion** ceiling to **$55 billion, and then to $65 billion in January 2026 (Migrant Insider, 2026; Real News Network, 2026). It has 154 contract holders, though only 87 have any spending obligations to date (WYPR, 2026). A senator described the arrangement as the military’s warfighter contract vehicles being “misused to facilitate the largest immigrant detention camp system since World War II’s detention of persons of Japanese ancestry” (Real News Network, 2026).
The Bureau of Prisons is a second conversion channel. On 6 February 2025, ICE and the BOP entered a six-month agreement allowing ICE to detain up to 800 people in five BOP prisons: FDC Miami, FDC Philadelphia, FCI Atlanta, FCI Leavenworth, and FCI Berlin (CBS News, 2026). A federal watchdog later found that staffing gaps and care delays at FCI Berlin were “exacerbated in some ways by the BOP receiving a new custody population from ICE” (CBS News, 2026).
State prisons are a third channel. A minimum-security state prison in McCook, Nebraska—the Work Ethic Camp—was reconfigured as a federal immigration detention center, receiving its first detainees in late 2025, with a capacity of 200 expanding to 300 (CBS News, 2026). The Angola prison in Louisiana was converted into an immigration detention center called the “Louisiana Lockup,” used to indefinitely detain people as punishment for past crimes for which they have already served their time (Boston Globe, 2026).
FEMA is a fourth channel, and it operates on a different logic from the others. FEMA allocated more than $600 million** to state and local agencies to detain migrants, and the **Detention Support Grant Program** was created in 2025 specifically to reimburse Florida **$608.4 million for “Alligator Alcatraz” and other state-run detention centers; the first payment of $58 million was made in May 2026 (Newsfromthestates, 2026). FEMA headquarters has been used to host immigration hearings, a move that congressional critics described as dragging FEMA “into border security” (Congressional correspondence, 2026).
Family detention is the fifth channel, and it is the one that was revived rather than converted. It was ended by the Biden administration in 2021 and reopened in 2025. ICE reopened family detention centers at Karnes City and Dilley, Texas. By early 2026, more than 5,600 people had been imprisoned at Dilley between April 2025 and February 2026, and the number of children in ICE detention climbed to about 175 children a day in 2025 (Marshall Project, 2025). From January to October 2025, at least 3,800 children under the age of 18, including 20 infants, were arrested and detained by U.S. immigration authorities (Marshall Project, 2025).
The warehouses are the machinery made of steel. They are built before anyone says what they are for. That is the pattern the whole essay describes.
II. The Legal Channels: How the Pipeline Can Receive Citizens
The domestic pipeline does not require a new statute to receive U.S. citizens. It runs on a set of existing channels, and each is being tested.
18 U.S.C. § 4001(a) states that “No citizen shall be imprisoned or otherwise detained by the United States except pursuant to an Act of Congress.” This is the primary legal barrier, and it is not a formality. The statute was passed in 1971 explicitly to prevent the kind of domestic detention camps that the 1950 Internal Security Act had authorized.
Enter the NDAA
The NDAA’s detention provisions were drafted to be ambiguous by design. Section 1021 of the 2012 NDAA affirmed the President’s authority to detain individuals “under the law of war” (the 2001 Authorization for Use of Military Force), and it did not explicitly exempt U.S. citizens (Human Rights Watch, 2012). When Senator Feinstein attempted to amend the provision to explicitly exclude citizens, the amendment failed; the fall-back amendment that was adopted instead states that the section “should not be construed as taking a position on the US citizen question one way or the other” (Human Rights Watch, 2012; Lawfare, 2023). The administration at the time asked the Senate Armed Services Committee to remove language that would have specifically exempted Americans from indefinite detention (Politico, 2011).
The result is a statute that is silent on the citizen question, and silence in a detention statute is not a prohibition. It is a delegation. The NDAA does not authorize the detention of citizens, and it does not forbid it. It leaves the question to whoever operates the machinery.
The NDAA’s authority is not, on its face, directed at Black and Latino citizens. It is directed at “covered persons” under the AUMF—those who were part of or substantially supported al-Qaeda, the Taliban, or associated forces. The claim here is structural, not empirical: the detention authority is status-neutral with respect to citizenship, and the enforcement discretion that fills it is documented as racialized. The category “covered person” is not a racial category; the filling is empirical, and the empirical filling is produced by discretion. The NDAA is the statutory door through which such detention would pass if it were authorized, and the door is ajar because the citizen question was deliberately left open.
The Insurrection Act is the most direct pathway for domestic military detention. The Act empowers the President to deploy the military on U.S. soil for law enforcement functions, giving troops the power to “search, arrest and detain U.S. citizens” (ACLU of Maine, 2026). In June 2026, reporting revealed that senior officials, including Stephen Miller and Vice President JD Vance, advocated invoking the Act in Minneapolis amid protests (New Republic, 2026; House Oversight, 2026). During the same period, Miller reportedly called for suspending habeas corpus rights in deportation cases, a move that would remove the primary legal tool citizens use to challenge unlawful detention (House Oversight, 2026). The proposals were considered and, at the time, not executed. The legal channel remains open.
For naturalized citizens, denaturalization is the mechanism that severs the constitutional membrane protecting them, converting a citizen into a deportable subject. At the end of 2025, USCIS directed its field offices to supply the DOJ with 100 to 200 denaturalization cases per month in fiscal year 2026 (NPR, 2025; NYT, 2025). The historical rate was approximately 11 cases per year; the new quota represents an increase of two orders of magnitude (Democracy Forward, 2026). The procedure is civil, not criminal, and the burden is on the government—but the volume of the drive is the structural signal.
Two further channels are documented but should be treated with calibration. 18 U.S.C. § 111—assaulting, resisting, or impeding federal officers—has been used against citizens who question or interfere with enforcement operations. It produces detention, but it is a criminal charge rather than a detention authority, so it belongs to the criminal cage rather than the civil pipeline. The material witness statute (18 U.S.C. § 3144) has been criticized for being improperly used to preventively detain citizens, but the claim is contested and the empirical record is thin. It is a channel that exists, and it has been used in ways that critics describe as improper, but it has not been shown to be a systematic instrument.
III. The Citizen Pipeline: 506 Cases and the Retrospective Proof
The pipeline is not hypothetical. ProPublica documented 506 U.S. citizens detained by immigration agents between January 2025 and September 2026 (ProPublica, 2026; Democracy Now!, 2026). The findings are the record of the pipeline operating at scale.
More than 100 Americans were held after agents questioned their citizenship; nearly all were people of color (ProPublica, 2026). At least 70 American children and teens were detained, some handcuffed (ProPublica, 2026). 36 Americans were held for at least a day without being able to contact a lawyer or their family (ProPublica, 2026). Roughly a dozen citizens were deported, mostly children (ProPublica, 2026).
The legal form is retrospective. The person is taken into custody, and the burden is on them to prove citizenship from behind detention walls. Dulce Consuelo Diaz Morales, a Maryland-born U.S. citizen, was held for 25 days across five facilities in four states before being released, after her lawyers filed additional documents with the immigration court (ProPublica, 2026). The government “effectively shifted the burden onto a U.S. citizen to prove her citizenship while she was incarcerated,” her lawyer stated (ProPublica, 2026). Brian José Morales García, a Denver-born U.S. citizen, was deported to Mexico in April 2026 after a traffic stop, despite telling officers he had a birth certificate and Social Security card at home. He returned to Texas in September 2026 after a federal lawsuit, nearly five months after his deportation (ProPublica, 2026). Chanthila Souvannarath was deported to Laos in October 2025 despite a federal court order blocking his removal and while his citizenship status was legally unresolved (ProPublica, 2026).
The apparatus is not designed to exclude citizens. It is designed to process presence and documentation, and the enforcement of documentation is racialized in its application because the discretion is racialized. The DHS position is that “ICE does not arrest or deport U.S. citizens” (ProPublica, 2026). The 506 cases are the record against that position.
The pattern of the cases is the pattern of machinery that has been built and has not been aimed. The citizen detentions are not a designated intake. They are incidental capture, produced by an apparatus that is running and has not been committed. The distinction matters because incidental capture is reversible and designated intake is not.
IV. The Bridge: How the Status-Neutral Apparatus Finds Racial Minorities
The legal form is neutral. The enforcement is not. The bridge between the two is the documented practice of racial profiling, and it is the mechanism through which the detention apparatus reaches Black and Latino citizens.
The court finding. In a 111-page decision, a federal judge appointed by Donald Trump found “compelling and troubling evidence” that immigration officers were racially profiling immigrants and legal residents alike in violation of their constitutional rights (Independent, 2026). The judge found that federal immigration officials “adopted a policy” that allows agents to stop, search and arrest people “based solely on their race or ethnicity” without reasonable suspicion or probable cause (Independent, 2026). The lead plaintiff, Mubashir Khalif Hussen, a 20-year-old U.S. citizen, was walking to lunch in Minneapolis when masked ICE agents stopped him. He repeatedly told them “I’m a citizen,” but agents refused to look at his ID. “That don’t matter,” they told him (Independent, 2026). His employer showed up with a copy of his passport card through the windshield; the officers ignored it. He was held for several hours and shackled before an agent finally looked at a photo of his passport (Independent, 2026).
The human rights finding. Human Rights Watch documented “widespread and routine racial profiling” during Operation Metro Surge, with agents stopping, arresting, or detaining individuals “on the basis of their race or ethnicity” (HRW, 2026). The organization interviewed 12 individuals who had been stopped, arrested, and detained in Minnesota, reviewed legal filings and dozens of habeas petitions, and examined the sworn declarations of 29 individuals, including 20 U.S. citizens (HRW, 2026). Minnesota residents of Somali and Latin American descent were “notably targeted, despite the fact that the overwhelming majority of these communities are US citizens or have green cards” (HRW, 2026).
The legal authorization. In September 2025, the Supreme Court, in Noem v. Vasquez Perdomo, lifted a lower court’s temporary order blocking immigration arrests based solely on certain factors, including apparent race, ethnicity, or language (HRW, 2026). The decision is described by critics as “the Kavanaugh stop”—a reference to Justice Kavanaugh’s concurrence, which argued that officers must be able to consider “ethnicity, language, location, and type of work when assessing reasonable suspicion” (Above the Law, 2026). The practical effect, as HRW notes, is that “immigration agents may feel that there are no clear judicial constraints on their ability to stop and arrest people on the basis of their race or ethnicity” (HRW, 2026).
The data. The government’s own arrest data shows the racialized output of the discretion. A UCLA analysis found that Latinos accounted for 96% of Trump-era ICE arrestees who were subsequently deported (UCLA, 2026). LULAC, citing the UCLA study, reported that from January to October, federal immigration agents arrested more than 187,000 Latinos and deported more than 126,000, and that Latino arrests jumped from under 3,900 to nearly 6,000 after Stephen Miller set the quota of 3,000 arrests per day (LA Times, 2026). The NAACP Legal Defense Fund documented over 400,000 ICE arrests across all fifty states in the first 14 months of the administration, with the enforcement concentrated in Black and Latino neighborhoods (NAACP LDF, 2026).
The pipeline within the pipeline. The NAACP LDF describes the mechanism as a “prison-to-deportation pipeline”: a routine traffic stop in a Black or Latino neighborhood becomes “a death knell for a person’s immigration status” (NAACP LDF, 2026). The 287(g) program, which deputizes local police to perform immigration functions, “has turned traffic stops into a racialized weapon of mass exclusion” (NAACP LDF, 2026). The data supports the claim: in Florida, Hispanic drivers are arrested at nearly twice their representation in traffic stops and face 2.1 times higher odds of arrest (ACLU of Florida, 2026). A Department of Justice investigation found that Latino drivers in one jurisdiction were up to nine times more likely to be stopped than other drivers (ACLU Maryland, 2026). At the federal level, 65 percent of the agencies participating in 287(g) have “established histories of racial profiling and other civil rights violations” (CourtListener, 2025).
The scholarship. The legal literature names the mechanism. A 2026 article in Race and Justice argues that immigration enforcement has “inverted Fourth Amendment protections for Latino Americans,” shifting the burden from the government’s obligation to justify seizures to the citizen’s need to carry documentation to avoid detention (Brown, 2026). The article draws the historical parallel directly: “selectively imposed documentary demands function analogously to Jim Crow literacy tests—ostensibly neutral proof requirements administered discriminatorily to produce two-tiered citizenship” (Brown, 2026). The demand for papers is not incidental to racial profiling; it is “the operational means through which ethnic targeting functions” (Brown, 2026).
The detention apparatus does not need a racial category to reach racial minorities. It needs a status-neutral legal form—the administrative warrant, the “reasonable suspicion” standard, the 287(g) agreement, the documentation demand—and a racialized discretion that fills the form with the bodies the discretion selects. The form is the bridge. The discretion is what crosses it. The citizens in the 506 cases are the proof that the bridge is open and in use.
The bridge also answers the question of what kind of apparatus this is. A group that was excluded from the beginning and is now being reached through status-neutral instruments is not being integrated and then reversed. It is being excluded by the ordinary operation of an apparatus that was built when the group was already outside. The danger is not that it requires a purge. It is that it requires nothing new. The bridge is already built. The discretion is already racialized. The machinery is already running.
V. The Medical and Labor Regimes: The Carceral Substrate
The pipeline’s terminal form is the prison, and the prison’s logic extends to the body.
Forced labor. The Thirteenth Amendment’s exception clause permits involuntary servitude “as a punishment for crime whereof the party shall have been duly convicted.” In ICE detention, detainees are routinely required to perform essential labor—cooking, cleaning, maintenance—for as little as **$1 per day** (Human Trafficking Search, 2026). In May 2026, hundreds of detainees at **Delaney Hall** in New Jersey launched a hunger and labor strike against **GEO Group**, protesting spoiled food and pay as low as $1 a day for cleaning kitchens and bathrooms, sometimes without gloves or soap (Democracy Now!, 2026). Detainees said they felt compelled to work; those who refused faced retaliation, including threats of transfer or isolation (Human Trafficking Search, 2026). ICE has rewritten its detention standards to eliminate the $1-a-day pay requirement at the request of private prison operators, effectively shielding companies like GEO Group from minimum-wage lawsuits (The Independent, 2026; Prison Legal News, 2026).
Involuntary medical procedures. A review of federal court records found nineteen cases since January 2025 in which ICE was approved to perform involuntary medical procedures on hunger strikers in detention (The Guardian, 2026). In September 2026, a federal judge extended an order allowing ICE to perform a percutaneous endoscopic gastrostomy (PEG)—a surgical procedure involving an incision through the abdominal wall to insert a feeding tube directly into the stomach—on a Cuban man detained near Houston (The Guardian, 2026). This was the first time a PEG had been authorized against an ICE hunger striker. The World Medical Association states that force-feeding is ethically unacceptable when a mentally competent individual can make an informed decision to refuse food. The legal form is being used not to prohibit the intervention but to legitimize it.
Forced sterilization. A whistleblower complaint alleged that unnecessary hysterectomies were performed on immigrant women in ICE custody at the Irwin County Detention Center in Georgia (Inter-American Law Review, 2026). The history of federally funded sterilization programs targeting immigrant populations and the eugenics boards in thirty-three states is the precedent the current apparatus inherits (CNN, 2022).
VI. The Historical Substrate: Experimentation on Vulnerable Populations
The claim that the carceral substrate can extend to medical experimentation is supported by the historical record, and the record should be stated plainly because it establishes the pattern.
The Nuremberg Code’s mandate of voluntary consent was not adopted as part of American medical research ethics after the Second World War. Until the early 1970s, approximately 90% of all pharmaceutical research was conducted on prisoners (Hornblum, 1998). From 1962 to 1966 alone, 33 pharmaceutical companies tested 153 experimental drugs at Holmesburg Prison in Philadelphia. The experiments at Holmesburg, conducted under the direction of a University of Pennsylvania dermatologist, exposed the predominantly Black prison population to pharmaceuticals, viruses, dioxin, and chemical warfare agents (CNN, 2022). The Tuskegee Syphilis Study, running from 1932 to 1972, denied treatment to hundreds of impoverished Black men even after penicillin became available.
The formal prohibition that emerged in the 1970s did not eliminate the practice; it changed its legal form. The contemporary apparatus shows the adaptation: the prohibition on non-consensual experimentation has been met with court-authorized procedures that provide a legal veneer for interventions that would otherwise be unlawful. The mechanism is the same as the one in the deportation pipeline: the legal form is used not to constrain the apparatus but to facilitate it, by supplying the authorization that the apparatus requires.
PART TWO: THE FOREIGN APPARATUS
VII. The Foreign Facilities
The United States has sent more than 25,000 people to third countries under arrangements it does not publish in full. Human Rights Watch has documented more than 500 deportees sent to African countries alone, where they had no contacts and were often arbitrarily detained or returned to the countries they fled (HRW, 2026). The Berkeley Human Rights Center’s Architecture of Cruelty (2026) has mapped the network country by country, and what it shows is that the American government did not build a system abroad. It bought access to systems that already existed. The systems differ by what the receiving state already had, and the differences are not cosmetic.
The hotel and the return ticket
In Sierra Leone, the arrangement is a hotel near the airport and a scheduled departure. The ECOWAS citizens on the flights already held a regional right of stay, so the country received them as regional travellers who had arrived in an inconvenient way. A contractor housed them for up to two weeks and described the model in its own literature as a temporary transit location with no long-term settlement (Berkeley, 2026). There is no prison, no armed guard, no curfew. The harm, where it exists, is in the exit: the transit is designed to complete a removal, and the removal returns people to places they had asked not to be sent.
Palau is the same regime without the transit. Deportees are screened for health and criminal record, permitted to live and work, and resettled rather than held. Rwanda is a qualified member of the group: the Gashora transit centre provides accommodation, healthcare, and training, with UNHCR involvement in status determination (Berkeley, 2026). The regime is humane and it expires. Sierra Leone’s cap is fourteen days, extendable to thirty. When the clock runs out, the person either goes home or becomes someone else’s problem.
The guarded compound
In South Sudan, the arrangement is a house with a guard on the door. Eight men were sent in July 2025 and held in a compound under armed guard in Juba, in a country so unstable that the State Department warns Americans not to travel there (Amnesty International, 2026). The guard is not a policy. It is a substitute for the infrastructure that South Sudan does not have. A regular traveller in Juba faces the same shortages of electricity, water, and medical care. The deportees were held in the best accommodation available because the alternative was leaving them unguarded in a city where nobody could be left unguarded.
The DRC is the same case with a different building: a hotel from which a group was reportedly prevented from leaving, raising what the Berkeley researchers call credible fears of de facto detention (Berkeley, 2026). The harm is the state’s poverty, not the state’s design. And the regime is unstable in the specific sense that it improves as the state’s capacity improves, or it collapses when the state collapses.
The room with no legal category
In Equatorial Guinea, the arrangement is a hotel that functions as a prison. About thirty deportees remain at the Hotel Bamy in Malabo, under twenty-four-hour armed guard (NYT, 2026). They have no legal status, because Equatorial Guinea has no framework that could grant them one. They have not been charged with any crime. Lawyers who reached them reported that men had been bound and beaten and that authorities had tried to confiscate their phones (The Guardian, 2026).
The detention is not a sentence and not a stated policy. It is what happens when a person is present in a jurisdiction that has no legal category in which to place them. Cameroon is the conditional version. Deportees were detained on arrival, released only after a hunger strike, and then placed under a curfew, prohibited from working while their asylum claims are pending, and required to return to the compound by sundown (HRW, 2026). A release that requires you to come back at nightfall and forbids you to earn is not a release. It is a confinement administered with a clock instead of a door.
This arrangement is essentially the same that Germany, Denmark, Finland and Sweden used in the 2015 migration wave, with the same types of violations. It resolved in regularisation and settlement with limited deportations. This is the likely path for countries lacking legal capacity.
This is the least stable arrangement of all, because it cannot be defended, only tolerated, and tolerating it is expensive: guards, facilities, lawsuits, hunger strikes. The Cameroonian hunger strike is the mechanism in miniature. Two months of confinement, a strike, and the state’s answer was not to build a legal framework but to relax the confinement. Relaxation is cheaper than construction, and it accelerates with every lawsuit.
The prison
In Eswatini, the arrangement is a prison. Deportees are held at the Matsapha Correctional Centre, a high-security facility, without charge and without access to counsel (Berkeley, 2026). For the first two months they were allowed outdoors for fifteen minutes a day and given one weekly phone call.
El Salvador’s CECOT is the same arrangement at its outer limit. The facility’s director has said that once an inmate enters CECOT, he never leaves (BBC, 2026). Former inmates have described severe mistreatment, sexual abuse, and torture (BBC, 2026). Kosovo belongs here: the Gjilan prison, the facility leased to Denmark for two hundred million euros, with abuse documented as part of the prison culture (Berkeley, 2026). Uzbekistan belongs here: immigration detainees are held in criminal prisons, in a system with documented torture risk and no Refugee Convention accession (Berkeley, 2026).
The prison is not an aberration in these countries. It is the institution they already operate, and the deportee is a new category of inmate added to it. Eswatini did not build a facility for the deportees. It added a category to a facility it already operated. El Salvador did the same. Kosovo did the same. The United States did not purchase a carceral system abroad. It purchased access to carceral systems that already existed, and the systems that already existed are the ones that will still be there when the arrangements are renegotiated.
This is the only arrangement that does not erode, because it does not require the state to build anything. The state already has the building, the staff, and the budget. Adding a foreigner to a national prison costs almost nothing, requires no new law, and produces the most control per unit of effort.
The informal economy
Mexico is the largest destination and the least visible arrangement. Nearly twenty thousand third-country deportees have been sent there, given a ten-day permission, and left without a right to work (Berkeley, 2026). There is no prison and no guard. The coercion is subsistence. A person who cannot work legally and cannot leave within ten days is a person who will work informally, and the informality is the mechanism.
What the separation produces
The arrangements do not drift toward one another. They drift toward two ends.
The hotel and the return ticket terminate, because they are scheduled. The guarded compound improves or collapses with the state that hosts it. The room with no legal category is a residue that cannot be maintained at scale and cannot be defended in court, so it moves toward the informal economy as soon as the confinement is challenged. What remains at the end of the drift is the prison, which is designed and durable, and the informal economy, which is not designed at all and therefore cannot be dismantled.
The transit hotel and the guarded compound are waypoints. The legal vacuum is a residue. The prison and the limbo are the equilibria. The program selects for them because they are the cheapest stable forms, and the cheapest stable arrangement is the one that accepts the removal as a premise and acts on it in a logical way. That is why the United States is buying access to prisons abroad for the same reason it is building warehouses at home.
PART THREE: EUROPE
VIII. Why the European Legal Order Is Structurally Different
The European legal order does not contain the channels the U.S. pipeline requires. This is not a claim about virtue. It is a claim about the form of the statute, and the form determines what can be done without a new law.
The closed list. Article 5(1) of the European Convention on Human Rights permits deprivation of liberty only in the cases exhaustively enumerated in the text, and the European Court of Human Rights has confirmed that the list is closed (ECtHR, Factsheet V). There is no category for the preventive detention of a citizen suspected of posing a future threat. The Parliamentary Assembly of the Council of Europe has stated expressly that “purely preventive detention of persons suspected of intending to commit a criminal offence is not authorised by Article 5″ (PACE, 2011).
The non-deportable citizen. Article 20 TFEU establishes citizenship of the Union, and the CJEU has held that a Member State cannot expel its own nationals (Ruiz Zambrano, C-34/09; McCarthy, C-256/11). The category “deportable EU citizen in the home state” does not exist in the legal order. Directive 2004/38/EC permits removal from a host Member State only on “serious grounds” of public policy or public security, and for a citizen with ten years’ residence or a minor, only on “imperative grounds” of public security—a threat to the very existence of the state.
The absence of a domestic security statute. The NDAA and the Insurrection Act are the two U.S. statutory channels that could be read to authorize domestic detention of citizens outside the criminal process. The EU has no equivalent. The European Parliament has explicitly criticized the US NDAA for authorizing indefinite detention. Member states retain emergency powers, but those powers operate through constitutional derogation under Article 15 ECHR, which requires a public emergency threatening the life of the nation, measures “strictly required by the exigencies,” and safeguards such as parliamentary control (Lawless v. Ireland, 1961). Derogation is a legal event, not an administrative decision.
The consequence is that the U.S. pipeline is legally available and operationally built. The EU equivalent is legally unavailable and operationally absent. The U.S. form is a delegation; the European form is a prohibition. A delegation can be used without a new law. A prohibition must be repealed, derogated, or captured. Those are legal events, and legal events are the friction that keeps the European core from moving at the speed of the American system.
The brake is not absolute. The EU’s Return Regulation, adopted in 2026, extends detention periods to a maximum of 24 months and creates a legal basis for return hubs outside the EU (European Parliament, 2026). The rule-of-law backsliding in Hungary and Poland demonstrates that the legal form can be eroded from within, and the Article 7 procedure has been slow and politically constrained. A member state that captured its constitutional court and its ordinary judiciary could, in principle, use the derogation channel to build a domestic detention regime. But the direction of travel is slower, and the friction is real.
IX. The European Core and Periphery
The European case is not one thing. It splits along the border between the core and the periphery, and the split is not a measure of how much racism exists in each place. It is a measure of what the receiving state’s legal order will permit without a new law, and how far along the sequence of exclusion each zone has traveled.
The scale used here is written rather than numerical, because the point is the category, not the decimal. A zone is Low when there is no apparatus and no project. Moderate when there is rhetorical consolidation and cultural pressure but no machinery. Elevated when the machinery is under construction but the commitment has not been executed. High when the machinery is committed outward and the groundwork is laid domestically. Severe when the machinery is committed both outward and inward.
| Zone | Structure | Stage | Danger |
|---|---|---|---|
| European Core (UK, FR, DE) | Foundational, then codified | Rhetorical consolidation, with cultural pressure | Moderate |
| Visegrád (PL, HU, CZ, SK) | Citizenship stripping | Committed at the border; rhetorical domestically | Severe |
| Balkans (RS, HR, RO, BG) | Citizenship stripping | Committed at the border; rhetorical domestically | High |
| Italy | Citizenship stripping | Committed at the border; rhetorical domestically | Elevated |
| Baltics (EE, LV, LT) | Chronic marginalization | Rhetorical consolidation | Elevated |
| Finland | Foundational, then codified | Rhetorical consolidation, with cultural pressure | Moderate |
| Iberia (ES, PT) | Chronic marginalization | Rhetorical consolidation, with cultural pressure | Moderate |
The European core. The core is at the stage of rhetorical consolidation, with cultural pressure beginning to register. The backlash is consolidated as rhetoric: the delegitimization of the institutions that produced partial inclusion, the framing of diversity as conferring “unearned racial advantage,” the finding that diversity efforts in the UK newsroom have been “performed rather than embedded” (Runnymede, 2025). The cultural pressure is present but not yet measurable at the scale of the United States. The apparatus is not being built. The domestic Black populations are not yet being framed as a demographic threat in the American register. This is the safest of the zones, but it is not safe. It is the stage before the machinery.
Visegrád is the most dangerous zone in the European periphery. Poland completed a 186-kilometre steel wall on its border with Belarus, suspended the right to claim asylum, and legalised pushbacks (Border Forensics, 2025). The Polish-Belarusian border is a “lethal pushback regime” in which “the framing of migration within racial logics of enmity” is explicit (Border Forensics, 2025). Hungary has been accused of promoting an “ethnically homogeneous nation” while depicting racialized minorities as threats (ECRI, 2025). The apparatus is committed outward with full state backing, and the ideological groundwork for a domestic variant is being laid.
The Balkans combine a committed border regime with domestic political actors actively deploying racialized threat narratives. In Romania, a far-right MP was sent to trial for inciting hatred against workers from Asia and Africa (Balkan Insight, 2026). The apparatus is present and the rhetoric is being tested.
Finland has built a border wall covering a part of the Finnish-Russian border and passed legislation allowing pushbacks of unauthorized border crossers, but in practice their authorities have only returned Russian citizens pursued by Russia on criminal charges once determined a public security risk in Finland. Theory and practice are two different things in Finland. Then there’s the Finns Party, whose MPs have made racist comments and promoted the “great replacement” conspiracy theory (YLE, 2025). The state’s institutional apparatus has produced reports documenting widespread racism. The structure is foundational exclusion, and the tension between the government’s rhetorical direction and the institutional friction that remains is the whole of the Finnish case.
Italy has a government explicitly committed to the same logic at the border: swift deportation of foreign offenders, easier citizenship stripping, a cap on foreign children in classrooms (Ministero dell’Interno, 2026). Italy’s political culture is described as “colour-blind,” which coexists with “overt anti-Blackness” and institutional racism (ECRI, 2025). The fiscal cleavage is the transmission mechanism: pension spending is 16.3% of GDP against a Eurozone average of 12.5%, while education spending is 4% against 4.6% (OECD, 2025). The median voter is 58; the pensioner bloc is over 16 million. The nativist coalition’s fiscal priority is the pension cheque, and the left’s is the university and the training college. The pensioner–youth cleavage is the reason the left’s “right to remain” is fiscally expensive in a way the sovereignty frame is not.
The Baltics are a different case. The primary axis is ethnicity and citizenship, directed largely at Russian-speaking minorities, not race. A Lithuanian sports official openly declared “I am racist” after sabotaging a Black athlete (Delfi, 2026). The structure is chronic marginalization, because the exclusion is not performed through a formalized project of citizenship stripping for a racialized group, and the danger reflects the second stage.
Iberia is the zone where the machinery has the least traction. The structure is chronic marginalization, characterized by systemic barriers and “banal racism” rather than a codified project. Vox and Chega poll around 20% each, and their role is as coalition partners rather than senior governing forces. The institutional and legal framework provides more friction than anywhere else in the periphery.
X. The Italian Pivot
Italy is the only periphery zone that combines the capacity and the intent to tilt the EU, and it is also the only one where the constraints that currently hold the nativist coalition back are the reason the danger is latent rather than active. Meloni is uniquely constrained. She is also, if those constraints lift, uniquely dangerous, because she is the one leader in Europe positioned to rewrite the rules of the union from the inside rather than from the outside.
The constraints are real. The premierato reform cleared parliament but faces a confirmatory referendum, and a referendum is a legal event that can fail. The RAI reform, which shifts the balance of power over the public broadcaster toward the executive, is contested by the opposition and by the president. The justice reform, separating the careers of magistrates, is slow and litigated. The electoral law — the instrument that matters most, because a majoritarian premium converts a plurality into a supermajority and a supermajority into a constitution the successor coalition cannot easily reverse — has not been touched. The EU fiscal framework and the NRRP conditionality bind her budget. Her coalition includes Forza Italia, which is more pro-European than her own party. The Italian public is not uniformly nativist, and her government has been more Atlanticist and more pro-Ukraine than her coalition’s rhetoric suggested.
The constraints explain why Italy has not, so far, done what Hungary did. They do not explain why Italy is the pivot. That explanation is structural.
Italy is a founding member of the EU and the third-largest economy in the eurozone. It has a leader who has already assembled a coalition of 20 EU governments on migration policy, growing to fifteen countries plus Italy by March 2026 — a majority of the 27 (Chatham House, 2026). It has a leader who has explicitly positioned herself as a “central player” in Europe’s conservative mainstream, cultivating ties with the EPP’s Tusk and Merz while maintaining her nationalist base (TVP World, 2026). And it has a leader whose distinction from other nativists is the reason she is more dangerous, not less. She “represents forces seeking to move EU policy to the right while preserving its institutions”, whereas figures like Bardella favor a “Europe of nations” that would devolve power away from Brussels (TVP World, 2026). A system-breaker can be isolated or vetoed. A system-preserver with a large coalition can rewrite the rules from the inside.
The nativist bloc in the European Parliament has already demonstrated what that arithmetic looks like. The Return Regulation passed in June 2026 with 418 votes — the EPP voting with the ECR, PfE, and ESN (European Parliament, 2026). A right-leaning majority exists in the Parliament. What does not exist is a nativist majority in the Council. If Italy formally leads one — if the 20-government coalition on migration becomes a standing bloc on a broader agenda — the EPP has a choice between the nativist bloc and the progressive bloc. The arithmetic of the Parliament favors the nativist bloc. The arithmetic of the Council would favor it too, if Italy delivered the votes.
That is the danger. It is not that Meloni will do what Orbán did. It is that she will do what Orbán could not. She has the weight, the coalition, and the system-preserving posture to rewrite the rules of the union from inside the institutions rather than against them. The constraints are the only thing keeping that from happening. If they lift — if the referendum passes, if the electoral law is changed, if the coalition holds through a second term — Italy becomes the pivot on which the European project turns.
The energy-policy divide is a register, not the tipping factor. Meloni’s government installed approximately 28 extra gigawatts of renewable capacity and claims installed capacity grew by nearly 40% by October 2026 (Palazzo Chigi, 2026). The European Commission approved a €23 billion state aid scheme to support renewable deployment (European Commission, 2026). The nativist frame is not anti-renewable; it is sovereignty-first. The September 2026 hydrocarbon decree justified accelerated extraction in explicitly patriotic terms—”defending our nation’s energy security,” “generating higher royalties” (Palazzo Chigi, 2026)—and created government commissioners empowered to override regional authorities. The left is more consistently pro-renewable and pro-Green Deal, but it is not the sole supporter of renewables. The real divide is who gets to set the rules, and that divide is consistent with the second stage.
XI. The Runway
The nativist coalition’s runway is not the pensioner bloc’s remaining years. It is the window in which the young male cohort can be consolidated as a permanent base, and the window is open now.
The gender gap in political preference is now one of the widest social cleavages in European electorates. Young women are moving toward the pro-investment coalition; young men are drifting toward the nativist right at rates that the pensioner model cannot explain (AfD youth polling, 2024–2026; RN post-election analysis, 2024; Vox youth polling, 2025). The mechanism is not the pension. It is meaning. The pro-investment offer is a credential and a service role—procedural, compatible with the care-oriented forms of identity the investment state recognizes. The nativist offer is a self with a plot—defender, protector, the one who holds the line. Provision and sacrifice are not symmetrical offers. Provision is a floor. Sacrifice is a ceiling, and a cohort that has been told it will be provided for, and has not been told what it is for, will reach for the cause.
The runway is five to ten years without the lock, and twenty-five to forty with it. The lock instruments are identifiable and partly built in Italy: the premierato reform, the RAI reform, the justice reform. The electoral law is the instrument that has not been built, and it is the one that matters most. The electoral law is the one to watch.
PART FOUR: WHERE THINGS STAND
XII. The United States
The U.S. domestic apparatus is at the stage of apparatus building, with commitment not yet executed for the mass domestic population.
The apparatus is built: the beds, the warehouse conversions, the military-base contracting mechanism, the dual cages (criminal and civil), the administrative warrants, the transfer routes, the legal channels, and the medical and labor regimes. The rhetoric is consolidated: “the homegrowns are next,” the threat to denaturalize and deport citizens, the explicit discussion of sending citizens to CECOT (La Prensa Gráfica, 2026; Despacho 505, 2026). The operational practice is documented: 506 citizens detained, more than 100 citizenship-questioned, 70 children, 36 held incommunicado, a dozen deported (ProPublica, 2026). The bridge to racial minorities is documented: a federal court finding of a policy of racial profiling, HRW’s finding of widespread profiling, the Supreme Court’s authorization of race as a factor, and the data showing Latinos at 96% of deported ICE arrestees (Independent, 2026; HRW, 2026; UCLA, 2026).
What has not occurred is the systematic redirection of the apparatus onto the domestic population as a stated objective. The citizens in the pipeline are being caught incidentally, as the dragnet expands and the discretion racializes. They are not the stated target. The stated target is the immigrant. The bridge is the mechanism through which the incidental capture happens, and the bridge is the reason the distinction between “immigrant” and “citizen” is collapsing in practice even as it remains intact in law.
The threshold is the transition from incidental capture to designated intake. The indicators are a legal designation that produces the domestic population as a category subject to detention; a rhetorical transfer that names the domestic population as the threat; and an operational practice that processes citizens through the pipeline at a scale that cannot be explained as error. None of the three has been fully crossed. The legal channels are being tested; the rhetoric is present; the scale is not yet there.
The pipeline is provisional in the precise sense: the capacity is built and the thresholds are being crossed incrementally, but the commitment has not been made. The American risk is a decision that can be made within the existing machinery.
XIII. Europe
The European core is at the stage of rhetorical consolidation, with cultural pressure beginning to register. The European periphery is committed at the border and at the stage of rhetorical consolidation domestically. The two clocks are running separately.
The transition from apparatus building to commitment for a domestic population in Europe would require a legislative act that defines the category, and the act would be visible, contested, and reviewed. That is the brake, and it is real. The brake is under coordinated attack—the Return Regulation, the ECHR challenges by nine member states, the Hungarian non-compliance—but the attack requires legal events, and legal events are slower than decisions.
XIV. The Assessment for the Whole
The United States is Elevated. The first four stages are present and documented; the apparatus is built; the commitment has not been executed. The danger is a decision that can be made within the existing machinery. It is not High because the legal order retains friction—the courts have struck parts of the apparatus, the states retain independent authority, and the 2026 midterm cycle is an open variable. It is not Moderate because the friction is being removed systematically, and the removal is proceeding through procedural instruments that are individually defensible and collectively directional.
The European periphery is High to Severe, with Visegrád Severe. The periphery is dangerous not because it is uniformly worse, but because in its most extreme zones the apparatus for exclusion is already built and committed, and the rhetorical groundwork for its internal extension is being laid.
The European core is Moderate. It is safer than the periphery, and safer than the United States in the current phase, but the direction of travel is toward the second stage of a reversal, and the institutional friction that slows the reversal is the same friction the backlash is designed to remove.
THE BOTTOM LINE
The archipelago is a pipeline, not a place. It is assembled from converted warehouses, leased prisons, county-jail contracts, military-base task orders, and third-country agreements. It is held together by a procurement vehicle designed for warfighters and a legal form that acts first and reviews afterward.
The sequence has stages, and the stage that matters is the last one. Diffuse exclusion, rhetorical consolidation, cultural preparation, apparatus building, commitment. Everything before commitment is reversible. Commitment is the point of no easy return, because the apparatus is now doing what it was built to do and the law has been arranged to permit it.
The capacity is being built to hold a domestic population. The 92,600-bed target, the 11 warehouse purchases, the $38.3 billion plan, and the WEXMAC channel are calibrated to volume, and volume is status-neutral.
The legal channels exist. The NDAA’s silence, the Insurrection Act, and denaturalization are the statutory doors. They were not opened by this administration. They were left ajar by the last one, and the current one has tested the hinges.
The citizens are already in the pipeline. 506 detentions, a dozen deportations, 70 children, 36 held incommunicado. The apparatus does not need a new law to receive them. It needs only the discretion that is already documented.
The bridge to racial minorities is the bridge to citizens. A federal court found a policy of racial profiling. Human Rights Watch documented widespread profiling of Somali and Latino residents. The Supreme Court authorized race as a factor in “reasonable suspicion.” Latinos account for 96% of deported ICE arrestees. The 287(g) program turns traffic stops into a racialized weapon. The legal form is neutral; the discretion is not. That is the bridge, and it is the reason the 506 citizens are overwhelmingly people of color.
The historical substrate is documented. Ninety percent of pharmaceutical research was conducted on prisoners until the 1970s. The formal prohibition changed the legal form, not the practice. The court-authorized PEG tube is the current form.
The foreign side is the same apparatus. A hotel in Sierra Leone, a guarded house in South Sudan, a hotel in Equatorial Guinea, a prison in Eswatini, a prison in El Salvador, a prison in Kosovo, a prison in Uzbekistan, an informal economy in Mexico. The prison is the only stable form, because it accepts the unlawful deportation as a premise and acts upon it in a logical way. The United States is buying carceral capacity abroad because carceral capacity is the cheapest stable regime.
Europe is not immune, but it is structurally different. The EU citizen is non-deportable by law. The European form is a prohibition, and a prohibition must be repealed, derogated, or captured. That is a legal event, and legal events are the friction that keeps the European core from moving at the speed of the American system. The European core is Moderate. The periphery is High to Severe, with Visegrád at Severe. The difference is the form of the statute.
Italy is the pivot, and it is uniquely constrained. Meloni faces a confirmatory referendum, a contested RAI reform, a slow justice reform, an untouched electoral law, the EU fiscal framework, and a coalition that includes a more pro-European partner. Those constraints explain why Italy has not done what Hungary did. They do not explain why Italy is the pivot. Italy is a founding member and the third-largest eurozone economy, and it has a leader who has already assembled a 20-government coalition on migration and positioned herself as a system-preserver rather than a system-breaker. A system-breaker can be isolated or vetoed. A system-preserver with Italy’s weight and a large coalition can rewrite the rules from inside the institutions. If the constraints lift, Italy becomes the pivot on which the European project turns.
The runway is the young male cohort. The mechanism is not the pension. It is meaning. The pro-investment offer is a credential and a service role. The nativist offer is a self with a plot. Provision is a floor, sacrifice is a ceiling, and a cohort that has been told it will be provided for, and has not been told what it is for, will reach for the cause. The runway is five to ten years without the lock, twenty-five to forty with it. The electoral law is the one to watch.
The archipelago does not announce itself. It is assembled in procurement documents and sheriff’s agreements. It is held together by the ordinary operation of a state that has decided that the cheapest stable regime is the one that accepts the premise and acts upon it. The premise is unlawful deportation. The architecture is the prison. The bridge is the racialized discretion. The pipeline is running.
References
Above the Law. (2026, April 16). Sotomayor Apologizes For Possibly Hurting Kavanaugh’s Feelings Over The Racial Profiling He Invented.
ACLU of Florida. (2026, April 23). New Report Exposes Stark Racial Disparities in Florida Traffic Stops as Street-Level Immigration Enforcement Expands.
ACLU of Maine. (2026, January 25). Trump’s Threat to Invoke the Insurrection Act, Explained.
ACLU Maryland. (2026). Testimony on 287(g) and racial profiling.
Amnesty International. (2012, November 29). 5 Reasons the Feinstein/Paul Amendment Doesn’t Fix the NDAA.
Amnesty International. (2026, May 15). South Sudan: Further information: Deportees’ ongoing detention is unlawful.
Balkan Insight. (2026). Romania: Far-right MP sent to trial for inciting hatred.
BBC News. (2026, June 18). The cemetery of living men: Trump deportees tell of abuse in secretive mega-jail.
Berkeley Human Rights Center. (2026, October 5). Architecture of Cruelty: Third Country Removal Agreements and the Second Trump Administration.
Border Forensics. (2025). Lethal pushback regime: The Polish-Belarusian border.
Boston Globe. (2026, February 14). Full scope of ICE’s ‘new detention model’ outlined in records released by N.H. Governor Ayotte.
Brown, D. L. (2026). Papers, Please: How Immigration Enforcement Creates a New Literacy Test for Latino Citizenship. Race and Justice. https://journals.sagepub.com/doi/10.1177/21533687261498919
CBS News. (2026, September 25). ICE wasted millions of taxpayer dollars in rapid push to expand detention capacity, government watchdog finds.
Chatham House. (2026). Italy and the future of EU migration policy.
CNN. (2022, October 7). Philadelphia apologizes for history of prison experiments on Black men.
Council of Europe, Parliamentary Assembly. (2011). Resolution on preventive detention.
Council of Europe, European Commission against Racism and Intolerance. (2025). Country reports: Italy, Hungary, Finland.
CourtListener. (2025, June 11). Amicus Curiae Brief in New Hampshire Indonesian Community Support v. Trump (1st Cir., 25-1348).
Deccan Chronicle. (2026, February 14). US Reveals $38 Billion Plan to Transform How It Detains Migrants.
Delfi. (2026). Lithuanian sports official’s racist remarks.
Democracy Forward. (2026, March 23). The Denaturalization of U.S. Citizens.
Democracy Now! (2026, August 14). 140 Immigrants Hold Hunger Strike at For-Profit ICE Jail in Tacoma, Washington.
Democracy Now! (2026, October 7). ProPublica: 500+ U.S. Citizens Have Been Detained by ICE.
Despacho 505. (2026, April 29). Estados Unidos estudia la legalidad de enviar a estadounidenses a la megacárcel de Bukele.
European Commission. (2026). State aid: Commission approves €23 billion Italian scheme to support renewable deployment.
European Court of Human Rights. (n.d.). Factsheet V: Right to Liberty and Security.
European Parliament. (2026, June 17). European Parliament approves new rules for the return of irregular third-country nationals.
EUR-Lex. (2011). Case C-256/11, McCarthy.
EUR-Lex. (2011). Case C-34/09, Ruiz Zambrano.
The Guardian. (2026, September 5). ICE given permission to force-feed Cuban hunger striker with gastric tube.
The Guardian. (2026, September 18). Men deported from US bound and beaten in Equatorial Guinea detention hotel, lawyers say.
Hornblum, A. M. (1998). Acres of Skin: Human Experiments at Holmesburg Prison. Routledge.
Human Rights Watch. (2012, November 29). Indefinite Detention Is Already Bad, Don’t Add Discrimination.
Human Rights Watch. (2026, February 20). Abuses in Cameroon After US Deports Third-Country Nationals.
Human Rights Watch. (2026, June 18). “A Manufactured Crisis”: Minnesota Communities Terrorized by the Federal Government.
Human Rights Watch. (2026, October 8). US: Deportations to African Countries Not Their Own.
Human Trafficking Search. (2026, October 21). Forced labor in detention: freed ICE detainee speaks out.
The Independent. (2026, March 10). Trump-appointed judge says ICE has a ‘policy’ of racial profiling: ‘Evidence is compelling and troubling’.
The Independent. (2026, June 16). At contractor’s request, ICE ends pay requirement for migrants who work in detention.
Inter-American Law Review. (2026, April 21). A Global Call for ICE Investigation: Disturbing Allegations in Whistleblower Complaint Raise Concern Over Hysterectomies Performed at Georgia Detention Center.
La Prensa Gráfica. (2026, September 26). Trump anuncia que envía a criminales de Estados Unidos a la megacárcel salvadoreña.
Lawfare. (2023, January 18). Congress, the Courts, and Detention of Americans under the AUMF/NDAA.
Los Angeles Times. (2026, July 28). Feds targeted Latinos more after Supreme Court’s racial profiling decision, report says.
The Marshall Project. (2025, December 17). ICE Threw Thousands of Kids in Detention, Many For Longer Than Court-Prescribed Limit.
Migrant Insider. (2026, February 2). How the Pentagon is Quietly Building Trump’s Concentration Camps.
Ministero dell’Interno. (2026). Legislative proposals on citizenship and public order.
NAACP Legal Defense Fund. (2026, May 26). ICE made 400,000 arrests in 14 months; Black, Latino and immigrant neighborhoods bore the brunt.
New Republic. (2026, June 15). Secret Memo Exposes Trump Team’s Debate on Suspending Constitution.
Newsfromthestates.com. (2026, May 29). Trump admin pays Florida first $58 million in ‘Alligator Alcatraz’ reimbursements, state says.
New York Times. (2025, December 17). Trump Administration Aims to Strip More Foreign-Born Americans of Citizenship.
New York Times. (2026, September 18). A Central African Hotel Has Become a Nightmare Prison for U.S. Deportees.
NPR. (2025, December 24). Trump administration wants to set quota for denaturalizing American citizens.
OECD. (2025). Pensions at a Glance 2025.
Palazzo Chigi. (2026). Statements on renewable capacity and the hydrocarbons decree.
Politico. (2011, December 1). Senate votes to allow indefinite detention of Americans.
Prison Legal News. (2026, September 1). New ICE Detention Standards Appear Designed to Help GEO Group Avoid Paying Wages.
ProPublica. (2026, October 6). More Than 500 U.S. Citizens Have Been Detained, Including Handcuffed or Held at Gunpoint, in Trump’s Immigration Crackdown.
The Real News Network. (2026, February 6). US military helping Trump to build massive network of ‘concentration camps,’ Navy contract reveals.
Runnymede Trust. (2025). Race and leadership in the UK news media.
Svimez. (2026). Rapporto Svimez 2026 sull’economia del Mezzogiorno.
TVP World. (2026). Meloni and the future of the European right.
UCLA. (2026). Unseen Latino Deportations. UCLA Latino Policy and Politics Institute.
U.S. House Committee on Oversight and Reform. (2026, June 17). Ranking Member Robert Garcia Demands Answers from White House Chief of Staff Amidst Plan to Suspend Constitutional Rights and Turn U.S. Military Against American People.
The Washington Post. (2026, February 13). ICE unveils $38.3B detention expansion plan.
WYPR. (2026, April 20). An obscure contracting method allowed ICE to secretly buy a warehouse in Maryland.
YLE. (2025, December 17). Analysis: The government is in another racism scandal, but nothing will change.