ICE Isn’t Cool.

Born from Fear: The Post-9/11 Origins of ICE On March 1, 2003, the United States government created a new kind of law enforcement agency, one that was neither quite a police force nor quite a court, but something in between and accountable to neither. U.S. Immigration and Customs Enforcement, or ICE, was the product of panic. The September 11 attacks had torn a hole in the national psyche, and Congress responded as it often does in crisis: by reorganizing things.

The Homeland Security Act of 2002 dissolved the century-old Immigration and Naturalization Service and scattered its functions among three new agencies under a freshly minted Department of Homeland Security. U.S. Citizenship and Immigration Services would handle lawful applications. Customs and Border Protection would guard the ports and perimeter. And ICE would handle everything else: interior enforcement, deportation, and criminal investigations related to immigration, customs, and trade.¹ The critical distinction between ICE and CBP is routinely blurred in public discourse and deserves to be clear. CBP is the agency you encounter at the airport or land crossing; it guards the perimeter. ICE operates inside the country: in neighborhoods, workplaces, courthouses, homes. Where CBP holds the line at the border, ICE draws an interior border wherever it decides to stand. Congress granted ICE, in the agency’s own words, “a unique combination of civil and criminal authorities.”² That combination, and the lack of clear guardrails around it, is where much of what follows begins.

It bears noting that ICE does perform legitimate law enforcement work. Its Homeland Security Investigations arm runs human trafficking operations, dismantles transnational criminal networks, and pursues financial crimes. This piece does not contest those functions. What it contests is the unchecked growth of the Enforcement and Removal Operations arm, and the agency’s systematic avoidance of the oversight mechanisms we demand of every other law enforcement body in the country.

The Warrant Problem: When an Agency Authorizes Itself Here is a foundational principle of American constitutional law: before the government can enter your home and take you away, it must persuade a neutral, independent magistrate judge that it has probable cause. That judge’s signature is not a formality; it is the Fourth Amendment made paper, the mechanism by which the executive branch is prevented from authorizing itself to act.

ICE, however, has long operated on a parallel legal track. Rather than judicial warrants signed by judges, it frequently relies on administrative warrants (Form I-200, Warrant for Arrest of Alien, and Form I-205, Warrant of Removal), issued by immigration officers within the same agency seeking to make the arrest. In May 2025, DHS quietly changed ICE policy to allow agents to enter private residences using only these administrative warrants.³ A whistleblower complaint subsequently revealed the policy change; DHS then defended it publicly.

The legal challenge is structural, not procedural. As the Brennan Center for Justice has analyzed in detail, administrative warrants are issued by immigration officers, not by the “neutral and detached magistrates” the Supreme Court has identified as the constitutional safeguard. The same agency requesting authorization is the agency granting it.⁴ In January 2026, U.S. District Judge Jeffrey Bryan ruled that an ICE agent’s forcible entry into a home using only an administrative warrant violated Fourth Amendment rights.⁵ The government’s defense has been revealing. DHS has argued that Congress never created a mechanism for ICE to obtain judicial warrants in purely civil immigration cases, and that this gap justifies bypassing Fourth Amendment requirements. The Brennan Center’s response is correct: if Congress declined to create such a mechanism, that may reflect a deliberate judgment that civil immigration violations should not be pursued by entering people’s homes. An institutional inconvenience is not a constitutional exception.

The government invokes the Supreme Court’s 1960 decision in Abel v. United States,6 which acknowledged historical legislative recognition of administrative arrest authority for deportable aliens. What the administration elides is that Abel did not involve forcible home entry, and that the accumulation of administrative warrant authority over seventy years has never produced a principled limiting rule. Even the Cato Institute has noted that returning to core warrant requirements, specifically issuance by a judicial officer and probable cause, would benefit courts and executive officials alike.

A Civil Infraction, Treated as a Crime Let us be precise about what most undocumented people in this country have actually done. Estimates consistently place visa overstays at between 40 and 45 percent of the total undocumented population, roughly four to five million people. In fiscal year 2023 alone, over 510,000 new undocumented residents became so not by crossing a border illegally but by remaining past the expiration of a valid visa.⁷ They came through airports and legal crossings. They showed their documents. They were admitted. And then they stayed.

Overstaying a visa is a civil immigration violation. Not a misdemeanor. Not a felony. A civil infraction, carrying consequences for immigration status but not constituting a crime in the penal sense. As immigration attorney Rosanna Berardi has noted, “the U.S. Supreme Court has been clear that deportation is a civil process, not a criminal punishment.”⁸ The apparatus deployed against these individuals, including armed agents, pre-dawn raids, detention facilities and family separation, is indistinguishable from criminal enforcement being applied to a civil wrong.

Then there is the constitutional question, which is routinely misrepresented in political rhetoric. The Constitution does not say “citizens.” It says “persons.” The Fifth Amendment’s due process clause, the Fourth Amendment’s protections against unreasonable search and seizure, the Fourteenth Amendment’s equal protection clause: all speak of persons, not citizens. The Supreme Court affirmed this in Plyler v. Doe (1982) and Zadvydas v. Davis (2001).⁹ In a 2014 joint interview, both Justice Ruth Bader Ginsburg and Justice Antonin Scalia agreed: “Every person who is here, documented or undocumented,” is protected by the Constitution.¹⁰ The Big Beautiful Bill and the Deportation- Industrial Complex On July 4, 2025, President Trump signed the One Big Beautiful Bill Act into law. Inside its pages was the single largest investment in immigration enforcement in American history.

ICE received $75 billion in supplemental funding available through September 2029.11 To put that in context: the entire Federal Bureau of Prisons budget in fiscal year 2025 was $8.6 billion. The U.S. Marine Corps proposed $57 billion.¹² ICE’s new war chest, when combined with its regular appropriation, made it the best-funded federal law enforcement agency in the country by a substantial margin.

The American Immigration Council has estimated that the additional detention spending could fund an increase in ICE detention capacity to at least 116,000 beds.¹³ The bill separately funds 80,000 new ICE beds, a $10,000 bonus for ICE and Border Patrol agents for four years, and full funding for the 287(g) program. The Brennan Center has described the resulting architecture as a “deportation-industrial complex,” a machine built not for public safety outcomes, but for removal volume.¹⁴ As of early 2026, the Cato Institute found no comprehensive public accounting of how the more than $190 billion in OBBBA funds allocated to DHS have been or will be spent, an extraordinary gap given the raw size of the cash infusion.¹⁵ One is left to ask why the removal of people who harvest crops and care for children warrants a budget that rivals the United States Marine Corps.

Who Builds Your House? Who Pays Your Social Security? Undocumented immigrants are woven into the fabric of industries that white-collar America depends on but would prefer not to examine closely. Construction, agriculture, food service, domestic work, and landscaping collectively employ the largest shares of undocumented workers. In Washington state, apple orchards depend on them. In the South, cotton fields do.

The argument that these workers displace native-born Americans does not survive contact with the labor market evidence. Research by economists including Giovanni Peri at UC Davis, and published by the National Bureau of Economic Research, consistently finds that immigration is a net complement to, not a substitute for, native labor.¹⁶ The Social Security argument is more stark and more urgent. In 2022, unauthorized immigrants contributed approximately $25.7 billion in Social Security payroll taxes, typically through wages reported under borrowed or incomplete Social Security numbers. They are categorically ineligible to collect a dollar of those benefits.¹⁷ The Social Security Trustees’ own projections show that higher immigration levels reduce the program’s projected shortfall, while lower immigration levels worsen it.

The demographic arithmetic is not disputed. The U.S. fertility rate currently stands at approximately 1.8 births per woman, below the replacement rate of 2.1. Social Security was designed for a country where more than five workers supported each beneficiary; that ratio is now below three and falling, projected to drop below 2.5 by mid- century.¹⁸ The people being deported are, in a quite literal actuarial sense, part of the workforce projected to fund the retirements of the Americans calling for their removal.

The Irony of the Nativist Argument There is a particular irony to the argument that America must be defended against the arrival of people who want to make better lives here, an irony that becomes visible the moment you consult the historical record.

The land belonged to people now called Native Americans. It was taken through warfare, treaty violation, forced removal, and deliberate destruction of culture and population, by European immigrants: the British, the Irish, the Italians, the Germans, the Norwegians, and the rest. Every family in America of European descent is descended from immigrants who arrived, competed for resources, and displaced or destroyed what was here before them.

If we are going to apply the logic of nativism consistently, if the argument is that people should not come here and change the character of what was here before, then the line for return flights should start with those whose ancestors made that argument impossible to make with a straight face. It should not start with the people picking strawberries in California.

History also gives us a useful empirical test. The Italian, Irish, Eastern European, and Chinese immigrants of the late nineteenth and early twentieth centuries were described in terms remarkably similar to those used today about Central American and Mexican migrants: criminality, cultural incompatibility, economic displacement, threats to national identity. In each case, those fears were proved wrong.

Blood on the Snow: Minneapolis and the Accountability Vacuum Between January 7 and January 24, 2026, three people were shot by federal immigration agents during Operation Metro Surge, ICE’s large- scale deployment to Minneapolis. What follows is based on publicly available video evidence, court filings, statements by the Hennepin County Attorney, and reporting by NPR, the Associated Press, and CNN; federal investigations remain ongoing and some facts are contested.

Renée Good, 33, a mother of three, was fatally shot by an ICE agent on January 7. Alex Pretti, 37, an intensive care nurse, was killed by two DHS officers on January 24. Julio Sosa-Celis, a Venezuelan immigrant, was shot through a closed door on January 14 and survived with a wound to his right leg.¹⁹ Federal authorities initially claimed the Sosa-Celis shooting was justified because the victim and his cousin had attacked an agent with a broom handle and a snow shovel. A federal judge later dismissed all charges against Sosa-Celis after finding new evidence “materially inconsistent” with the federal account. Authorities subsequently opened a criminal investigation into whether two ICE agents had lied under oath.

Hennepin County Attorney Mary Moriarty charged ICE agent Christian Castro with four counts of second-degree assault and one count of falsely reporting a crime.²⁰ He was arrested in Texas on May 29, 2026. DHS called the arrest “unlawful” and a “political stunt.”

What makes these cases operationally distinctive is not the shootings themselves but what happened afterward. When state investigators arrived at the scene of the Pretti shooting, federal authorities physically blocked them from accessing it. Renée Good’s car, potentially critical evidence, was transported to an FBI warehouse and, as of April 2026, had never been forensically examined.²¹ Minnesota ultimately sued the federal government to obtain access to evidence in all three shootings.²² In police-involved shootings across the country, there are protocols. Force investigation teams, separate from the shooting agency, examine the evidence and reach conclusions subject to review. When officers lie about what happened, there are consequences. The architecture of accountability exists, imperfect and unevenly applied, but present. Not so with ICE. This is not a posture consistent with an agency confident in its own conduct.

The Political Architecture of Exemption Why does ICE operate under different legal standards than other law enforcement agencies? The answer involves several converging structural factors.

First, immigration law occupies an unusual position in American jurisprudence. The Supreme Court’s “plenary power” doctrine, developed in the late nineteenth century in the explicit context of Chinese exclusion laws that would today be recognized as unconstitutional, holds that Congress has nearly unreviewable authority over immigration. Courts have been reluctant to apply ordinary constitutional scrutiny to immigration enforcement that they would apply everywhere else.

Second, the people most affected by ICE’s practices cannot vote. They cannot lobby. They have no electoral leverage. The political incentive structure therefore runs almost entirely in one direction: toward more enforcement, not toward accountability. Agencies whose primary constituency cannot vote tend, over time, to drift toward the preferences of those who can.

Third, there is the deliberate construction of dehumanizing language. People who came here legally and overstayed a visa are routinely described as “illegal aliens,” a phrase chosen not for legal precision but for political effect. It is easier to authorize aggressive enforcement against an “illegal alien” than against a mother of three, a nurse, a man answering his door.

The defense that critics offer for ICE’s legal exceptionalism, the argument that it is constitutionally authorized, that Congress intended it, that courts have upheld it, is worth taking seriously as law. Statutory authority is real. But authorized is not the same as wise, and legal is not the same as consistent with the values a democratic society should expect of its law enforcement institutions.

What Reform Would Actually Look Like Op-eds that enumerate problems without proposing solutions are criticism without responsibility. Here, then, is what meaningful reform of ICE would require: not abolition, but accountability.

First, a judicial warrant requirement for home entry. Congress should amend the Immigration and Nationality Act to require that ICE obtain a judicial warrant, signed by a federal magistrate judge with a finding of probable cause, before entering a private residence. This is the standard applied to every other law enforcement agency in the country. The May 2025 administrative policy change enabling warrantless home entry should be reversed and codified by statute.

Second, an independent oversight body with subpoena power. ICE currently reports to DHS, which reports to the President. Congressional oversight is episodic and partisan. What is needed is a standing, independent Inspector General with jurisdiction over use-of-force incidents, detention conditions, and warrant practices, with the authority to compel document production and testimony.

Third, mandatory data transparency. ICE should be required to publish quarterly: use-of-force incident data broken down by type, outcome, and demographics; detainee medical incident data; warrant type used in each arrest; and complaint disposition rates. This data exists for municipal police departments. It should exist for a federal agency spending $75 billion of public money.

Fourth, a narrowing of ERO’s mission. ICE’s own FOIA-released data has shown that in prior enforcement surges, the majority of people arrested had no criminal record. Refocusing Enforcement and Removal Operations on individuals with serious criminal convictions would improve public legitimacy and concentrate resources where the genuine public safety case is strongest.

Fifth, a federal-state evidence-sharing protocol. The Minneapolis episode demonstrated what happens when there is no mandatory protocol for federal-state cooperation after a shooting: evidence disappears, investigations stall, accountability becomes impossible. A statutory requirement that federal agents cooperate with state investigators in use-of-force incidents, including preserving and sharing evidence, would address this gap.

None of these proposals abolish ICE. None of them open the borders. All of them ask only that a federal law enforcement agency with extraordinary power and an extraordinary budget be subject to the same oversight norms we apply to everyone else.

The Country We Choose to Be America has always been most itself when it has been most open, most willing to absorb the ambitious, the desperate, and the dreaming and make something new of the encounter. The history of this country is, in large part, the history of people who were not wanted and became indispensable.

ICE, as currently configured and funded, is a $75 billion bet that the country’s future lies in exclusion: in removing the people who pick the fruit, care for the children, and pay into retirement funds they will never collect from. That bet is demographically wrong, economically costly, legally questionable, and operationally dangerous, as the available evidence from Minneapolis shows.

We built this agency in fear, after a catastrophe, without quite thinking through what we were creating. We have spent twenty-three years learning what it does when it operates largely outside normal accountability structures. The answer, it turns out, is what most unchecked institutions do: it expands, it cuts corners, it resists oversight, and it produces outcomes that its founders would not have authorized.

The question now is whether we have the political will to demand that it behave like law enforcement in a democracy, subject to courts, to independent investigation, to the basic dignity owed every person present on American soil, and to the transparency that is the minimum obligation of any agency spending public money and carrying government weapons. That is not a radical demand. It is, or should be, a minimum one.