The Parallel Army: Understanding ICE's Transformation
Budget tripled. Personnel doubled in four months. Training cut from 16 weeks to 6. Recruitment ads geofenced at gun shows with white nationalist imagery. This article analyzes ICE's rapid transformation through the lens of authoritarian "counterbalancing"—the creation of parallel security forces loyal to regime rather than constitution—and what Sudan's 2023 catastrophe warns about where this leads.
Analytical Frame: Counterbalancing, Praetorianism, and Leader-Centered Coercion
In comparative politics, leaders who fear either (a) resistance from professional coercive institutions, or (b) “legitimacy deflation” in the Nordlinger sense, often respond by restructuring coercive power rather than persuading opponents. One core strategy is counterbalancing: fragmenting coercive capacity across multiple agencies so that no single institution can block the leader—or mount a coup.
Erica De Bruin’s work provides the central empirical framework: counterbalancing can protect regimes by reducing coup success, but it tends to increase coup attempts and can raise the risk of violent conflict between rival coercive organizations. This is the “counterbalancing paradox.”
Huntington’s praetorianism adds a political warning: when coercive institutions take on leader-centered or partisan roles—especially under conditions of institutional polarization—coercion becomes a tool of political settlement rather than law enforcement. The crucial question is not whether a U.S. agency is “the same as” a presidential guard abroad, but whether it becomes functionally analogous: unusually well-funded, operationally insulated, politically recruited, and deployable for domestic intimidation.
The Numbers
The transformation of Immigration and Customs Enforcement in 2025 defies historical precedent:
Budget: ICE's budget tripled from approximately $10 billion in FY2024 to roughly $28-30 billion annually. The One Big Beautiful Bill Act, signed July 4, 2025, provided $170 billion in supplemental funding over four years—creating what critics call "fiscal autonomy" that bypasses the annual appropriations process.
Personnel: ICE's workforce more than doubled from approximately 10,000 to 22,000+ agents in roughly four months, driven by a $100 million "wartime recruitment" campaign.
Training: The standard training program was compressed from 16 weeks to 6-8 weeks. Some recruits were fast-tracked to just 4 weeks. The mandatory language course was eliminated.
Funding autonomy: The multi-year supplemental funding means ICE can remain operational even during government shutdowns, independent of annual continuing resolutions that constrain other agencies.
By January 2026, NPR reported that ICE had $85 billion at its disposal—more than Poland's entire military budget. It had become the highest-funded federal law enforcement agency.
The Recruitment Strategy
The $100 million "wartime recruitment" campaign revealed explicit ideological targeting.
Internal planning documents showed ICE targeting digital advertising toward audiences with interests in "military culture," "tactical gear," and "guns." Ads were geofenced around military bases, gun shows, and UFC fights.
The imagery was telling:
- Some recruitment materials utilized white nationalist slogans and imagery, including the caption "Which way, American man?"
- One ad featured a song by the white nationalist group Pine Tree Riots
- Ads showed tactical gear, bomber jets, and retro-style vans with captions like "Want to deport illegals with your absolute boys? Think about how many illegals you could fit in this bad boy!"
The "wartime" framing was consistent throughout. Not law enforcement recruitment. Wartime recruitment. The enemy wasn't named, but the messaging was clear: this was combat, not civil service.
Former ICE Director Sarah Saldaña warned that this approach "tends to inculcate in people a certain aggressiveness that may not be necessary in 85 percent of what you do." She feared attracting "combat-hungry or inexperienced people" rather than professional law enforcement officers.
The Training Compression
Professional law enforcement requires extensive training. The consequences of inadequate preparation—improper use of force, rights violations, procedural errors—can be catastrophic.
The standard ICE training program existed because decades of experience demonstrated it was necessary. Sixteen weeks covered legal authority, proper procedures, de-escalation techniques, language skills, and constitutional constraints.
Cutting training to 6-8 weeks—or 4 weeks for "experienced" recruits—meant eliminating most of this content. The compression happened because speed mattered more than professionalism. The administration wanted bodies deployed, not officers trained.
An AI tool used to parse applications reportedly flagged the "majority" of applicants incorrectly as experienced law enforcement officers, allowing them to bypass even the reduced training. Quality control had been sacrificed for quantity.
The result: thousands of agents deployed with minimal preparation for the constitutional and legal complexities of their work. As one analysis noted, this created an influx of "untrained recruits eager for combat."
The Oversight Dismantlement
Simultaneously with the expansion, the infrastructure for oversight was destroyed.
The Office for Civil Rights and Civil Liberties (CRCL) was effectively dismantled in March 2025. More than 500 civil rights complaints were open when work was suspended. More than 100 employees lost their jobs. CRCL had previously processed 3,000 complaints annually, including investigations of medical neglect and sexual assault at detention centers.
DHS spokesperson Tricia McLaughlin characterized these offices as "internal adversaries that slow down operations." The message was explicit: oversight was an obstacle to be eliminated, not a constraint to be respected.
DHS removed 160+ CRCL investigative memos from its website without announcement. The institutional memory of past misconduct—the documented record of what had gone wrong and why—was erased.
In O’Donnell’s terms, this is a direct strike on horizontal accountability: removing internal oversight produces discretion without monitoring—exactly the condition under which abuse becomes both more likely and harder to prove.
The Anonymity
A signature feature of the transformed ICE was operational anonymity.
Human Rights Watch documented that "federal immigration enforcement agents now commonly operate masked and without visible identification." ProPublica recorded over 50 cases since 2025 where ICE agents smashed car windows for arrests, compared to 8 in the previous decade.
Residents described masked men forcing people into unmarked vehicles. Agents wore plain clothes and refused to identify themselves during raids. People couldn't tell whether the armed men taking their neighbors were federal officers or criminal imposters.
Reagan-appointed Judge William G. Young issued a blistering condemnation: "ICE goes masked for a single reason—to terrorize Americans into quiescence... We have never tolerated an armed masked secret police."
Young called ICE's stated rationale for masking "disingenuous, squalid and dishonorable." He compared the agency's tactics to the "cowardly desperados" of the Ku Klux Klan.
The academic point here is not rhetorical. In coercion studies, anonymity and unaccountability are classic indicators of repression capacity: when agents cannot be identified, personal liability decreases, civilian fear increases, and the deterrent effects of law weaken.
The Sensitive Locations Policy
On January 20-21, 2025—the first days of the new administration—ICE rescinded its "sensitive locations" policy. This policy had protected schools, churches, hospitals, and courthouses from enforcement actions.
The policy existed because enforcement in these locations causes collateral damage beyond the targeted individuals. When ICE raids churches, immigrants stop attending religious services. When ICE waits outside hospitals, people stop seeking medical care. When ICE operates at schools, parents stop sending children to class.
The rescission meant nowhere was off-limits. A parent could be arrested dropping their child at school. A patient could be detained seeking emergency medical care. A witness could be seized leaving a courthouse.
The policy change wasn't about enforcement efficiency—most targets could be apprehended elsewhere. It was about creating pervasive fear. When nowhere is safe, the immigrant population (and those who might be mistaken for immigrants) lives in constant terror.
What Is Counterbalancing?
Political scientists use "counterbalancing" to describe a specific coup-proofing strategy: rulers fragment coercive power by creating parallel security organizations—presidential guards, militarized police, party militias—that check and balance the regular military.
Research by Erica De Bruin (Hamilton College) found that 63% of autocracies between 1946-2010 used counterbalancing. Her State Security Forces Dataset covers 375 security organizations across 110 countries.
The logic is straightforward: if the regular military might refuse illegal orders or mount a coup, create an alternative force that won't. The parallel force owes its existence and funding to the regime, not the state. Its personnel are selected for loyalty rather than professionalism. Its institutional culture prioritizes regime protection over constitutional constraint.
ICE is not a presidential guard in the classical sense—it's a statutory federal agency. But the relevant question is functional: has it become unusually resourced, operationally insulated, politically aligned in recruitment, and capable of domestic operations that can intimidate potential opposition?
By every measure, the answer after 2025 is yes.
The Counterbalancing Paradox
De Bruin's research reveals a dangerous tradeoff: counterbalancing reduces the likelihood that coup attempts will succeed by creating incentives for some soldiers to resist, but it is not associated with fewer coup attempts—and creating a new security force actually increases the odds of a coup attempt in the following year.
Additional research identifies that heavily coup-proofed regimes see considerable increases in civil war likelihood when coup risk is high. The parallel forces designed to protect the regime can themselves become threats—rival armed organizations that might fight each other rather than defend the state.
This is the counterbalancing paradox: the strategy that protects against coups creates new and potentially worse risks.
Sudan: The Warning
The 2023 Sudanese civil war provides the definitive case study of counterbalancing producing state collapse.
President Omar al-Bashir created the Rapid Support Forces (RSF) in 2013 as a counterweight to the regular military in one of the world's most coup-prone countries. The RSF emerged from the Janjaweed militias used during the Darfur genocide. It reported directly to Bashir rather than through military chains of command.
The parallel force worked as designed—for a while. It protected Bashir from military plots. But the RSF developed independent revenue through gold mining and mercenary contracts. It grew wealthy enough that its commander could pledge over $1 billion to Sudan's Central Bank.
Then the parallel force turned on its creator. In 2019, the RSF joined the regular military in overthrowing Bashir. In 2023, when integration plans threatened its independence, the RSF went to war against its former coup partner.
The humanitarian consequences represent counterbalancing's ultimate danger: 12+ million people displaced, an estimated 150,000-250,000 killed, 25 million acutely food insecure, and de facto partition of the country. The U.S. designated the RSF as having committed genocide in January 2025.
The force designed to prevent coups destroyed the country.
The Comparative Pattern
ICE's transformation maps onto patterns from authoritarian security architectures:
| Feature | ICE (US 2025) | IRGC (Iran) | RSF (Sudan) |
|---|---|---|---|
| Personnel surge | 120% in 4 months | Constant Basij expansion | Rapid from Janjaweed |
| Funding | Off-budget $170B | Direct oil/business control | Gold mining |
| Legal constraint | "Unreviewed" warrants | Answers to Supreme Leader | Independent of army |
| Domestic use | Targeting protesters | Internal dissent suppression | Civil war vs army |
The comparison isn't to claim equivalence—the institutional contexts differ significantly. But the structural features align in ways the academic literature on authoritarian consolidation finds significant.
The Judge's Assessment
ProPublica's October 2025 investigation—"Unfettered and Unaccountable: How Trump Is Building a Violent, Shadowy Federal Police Force"—quoted current and former national security officials describing ICE as "an unfettered and unaccountable national police force" that "could eventually be turned against any groups the administration labels a threat."
The investigation documented more than 170 U.S. citizens detained by immigration agents. Approximately 50 were detained based solely on questions about their citizenship—almost all Latino.
Judge William Young's ruling called the tactics "indistinguishable from the world's most repressive regimes." He found that the administration was using immigration law to suppress First Amendment-protected speech, targeting student activists for deportation to "tamp down" political activity on college campuses.
The academic point here tracks Huntington and Nordlinger: once coercive institutions become leader-centered and legitimacy fractures, coercion shifts from enforcing law to enforcing political settlement.
This is the thirteenth article in a series examining democratic decline. The next article documents extrajudicial killings in the Caribbean—117-126 dead in approximately 36 boat strikes—and how the legal architecture of the war on terror has been repurposed for peacetime execution.