Data Investigation • February 2026

The Detention Expansion

2025 saw ICE's budget nearly triple and its workforce double. But deportations barely moved. Where is all that capacity going—and what does the geographic distribution reveal about what's actually happening?

22,000+
ICE Personnel
↑ 120% in 2025
65,735
Detainees
↑ 65% in 2025
$28.7B
FY2025 Budget
↑ 193% from FY2024
290,603
Removals (YTD)
↑ 7% from FY2024

The numbers don't add up the way you'd expect. Immigration and Customs Enforcement received a 193% budget increase, more than doubled its workforce, and expanded detention capacity by 65%—yet actual deportations increased just 7% year-over-year. The gap between infrastructure growth and operational output is the central puzzle of the 2025 ICE expansion, and understanding it requires looking past the headlines to the geographic and structural data underneath.

This report presents that data: where ICE operates, how enforcement varies dramatically by state, who is being detained, and how 2025's policy changes reshaped the entire system. The visualizations below are interactive—explore them to understand your state's position in this national picture.

The Enforcement-Detention Gap

The defining feature of the 2025 expansion isn't increased deportations—it's increased detention.

+65%
Detention Population Growth
39,703 → 65,735
vs
+7%
Deportation Growth
271,484 → 290,603

The Divergence

Detention population vs. monthly removals, 2025

Release-to-Removal Ratio

For every person released, how many are deported?

Key Finding
The release-to-deportation ratio shifted from 1:1.6 in December 2024 to 1:14.3 by late 2025. For every person released pending a hearing, 14 are now deported directly from custody—or remain detained.

This gap suggests three possible interpretations. First, the system may be preparing for future capacity—building infrastructure now that will be utilized later. Second, detention itself may be functioning as policy, with prolonged confinement serving as deterrence independent of eventual removal. Third, bottlenecks elsewhere (immigration courts, receiving countries, transportation logistics) may be limiting throughput despite expanded capacity.

The data supports elements of all three, but the geographic distribution provides crucial context. Where detention is concentrated, who operates the facilities, and which states see the highest enforcement activity tells a story about political economy as much as immigration policy.

Geography Is Policy

Your exposure to ICE enforcement depends primarily on where you live—and your state government's political alignment.

Low → High
ERO Field Office
Major Detention Facility
What the map reveals
Texas accounts for 27% of all detainees and 23% of all arrests despite having 8.5% of the U.S. population. The concentration isn't random—it maps onto states with mandatory local-federal cooperation, established private prison infrastructure, and political alignment with enforcement expansion.

Per-Capita Rates Tell the Real Story

Total numbers obscure a crucial pattern. California has the third-highest arrest count, but its per-capita arrest rate is far lower than interior states like Tennessee, Utah, and Oklahoma. The difference? Sanctuary policies and limited jail cooperation.

This creates radically different lived realities. An undocumented resident in Nashville faces a per-capita arrest rate of 45.9 per 100,000—more than double California's rate despite Tennessee having a far smaller undocumented population. Political geography, not demographic geography, determines enforcement intensity.

Detention by State

Top 10 states by current detainee population

Arrest Rate Comparison

Per 100,000 residents, Jan-Jun 2025

The Vanishing Criminal Rationale

For years, ICE justified its operations by emphasizing focus on public safety threats—those with criminal convictions. The 2025 data shows this framework has collapsed.

73.6%
of current ICE detainees have no criminal conviction—up from 42% in January 2025

Detainee Criminal Status

Share with vs. without criminal convictions

Dallas Field Office Breakdown

2025 arrests by criminal history

The Dallas field office data is particularly revealing: 62% of those arrested in 2025 had no criminal conviction whatsoever. This isn't targeted removal of dangerous individuals—it's broad population enforcement. Whether one views this as appropriate policy or mission creep depends on priors, but the data itself is unambiguous about what's happening.

Policy Implication
The shift from criminal-priority to broad enforcement changes the fundamental character of ICE operations. The "dangerous criminals" framing that has historically justified the agency's existence no longer matches its operational reality.

Follow the Money: Private Prison Geography

The detention expansion flows through private contractors. Understanding who profits—and where—reveals the economic logic driving facility placement.

Facility Operators

86% of detainees in for-profit facilities

Budget Allocation

FY2025 $28.7B breakdown

📍

GEO Group

Operates 22,000+ beds across 20 facilities. Key contracts include South Texas IPC (Pearsall), Northwest IPC (Tacoma), and the recently reopened Delaney Hall in Newark under a $1.2 billion contract.

📍

CoreCivic

Manages major facilities including Stewart Detention Center (GA), Adams County (MS), California City (CA), and the South Texas Family Residential Center in Dilley—the largest family detention facility at 2,400 beds.

📍

LaSalle Corrections

Dominates Louisiana detention with Winn Correctional Center (1,576 beds) and a newly opened facility at Angola—the infamous Louisiana State Penitentiary now also housing ICE detainees.

The geographic concentration isn't coincidental. Texas, Louisiana, Georgia, and Mississippi—the states with the highest detention populations—share two characteristics: established private prison infrastructure and regulatory environments favorable to these contractors. The $45 billion detention allocation over four years flows disproportionately to these states and these companies.

"The top 20 detention facilities hold 59% of all ICE detainees. The system is remarkably concentrated—a handful of private contractors in a handful of states control the majority of America's immigration detention capacity."

This creates infrastructure lock-in. Private prison contracts, facility construction, and personnel hiring generate institutional momentum that persists across administrations. The machinery being built in 2025 will exist for whoever holds power next—and the companies holding these contracts have strong incentives to ensure continued utilization.

Historical Context

The 2025 expansion in the context of detention trends since 2020.

Detention Population Over Time

Average daily population, 2020-2025. The January 2025 inflection point is visible.

Reading This Chart

The COVID-era collapse (2020-2021), the gradual rebuild (2022-2024), and the 2025 surge are all visible. Note that the current population exceeds even the pre-COVID high of early 2020. The vertical dashed line marks January 2025—the inflection point when the current expansion began.

What the Data Does (and Doesn't) Support

Immigration enforcement generates strong claims across the political spectrum. The 2025 data complicates many of them—on all sides. Below, we examine common arguments against the evidence.

"We're just enforcing existing laws"

+

This framing presents enforcement levels as legally determined rather than policy choices. The data shows otherwise.

The same laws existed in 2023 when detention populations were 28,000. The 2025 expansion to 65,000+ reflects resource allocation decisions, not legal requirements.

The Laken Riley Act created new mandatory detention categories. The expedited removal expansion changed which laws apply where. Law enforcement always involves discretionary choices about priorities and resources. The 193% budget increase wasn't mandated by statute—it was a policy choice about how much to spend on enforcement versus other priorities.

What's accurate: The legal authority for these actions exists. What's misleading: The implication that current enforcement levels are required rather than chosen.

"ICE focuses on criminals and public safety threats"

+

This was the explicit Biden-era priority framework, and it's demonstrably no longer operative.

73.6% of current detainees have no criminal conviction—up from 42% in January 2025. In the Dallas field office, 62% of 2025 arrests involved people with no criminal history.

The shift is categorical. When nearly three-quarters of detainees have no criminal conviction, the "dangerous criminals" framing doesn't match operational reality. This doesn't necessarily mean current enforcement is illegitimate—one can argue for broad enforcement on other grounds—but it does mean the public justification has decoupled from actual practice.

What's accurate: ICE does arrest and deport people with criminal convictions. What's misleading: The suggestion that criminals are the primary or even majority target of current operations.

"Sanctuary cities obstruct legitimate enforcement"

+

The data shows sanctuary policies do reduce enforcement intensity—but "obstruction" implies something illegitimate.

California's per-capita arrest rate is roughly half that of cooperative interior states like Tennessee (45.9/100k) and Utah (47.0/100k), despite having larger undocumented populations.

Sanctuary jurisdictions have made democratic choices about local resource allocation. They've decided not to use local police and jails to extend federal immigration enforcement. Whether this constitutes "obstruction" or "local autonomy" depends on one's view of federalism and enforcement priorities.

The flip side: approximately 50% of ICE arrests still originate from local jails through cooperation programs. The states that have chosen cooperation see dramatically higher per-capita enforcement—their residents bear disproportionate exposure to federal immigration enforcement.

What's accurate: Sanctuary policies reduce ICE's operational effectiveness in those jurisdictions. What's contested: Whether local jurisdictions have the right to make those choices, and whether the variation constitutes "obstruction" or federalism working as designed.

"Mass deportation is logistically impossible"

+

The infrastructure scaling suggests serious preparation, even if current throughput lags.

Personnel doubled in 5 months. Bed capacity is targeting 107,000+. $75 billion has been allocated over 4 years. 33,000 federal employees from other agencies have been detailed to assist ICE.

The question isn't raw capability but political sustainability and economic cost. The 7% increase in actual removals despite massive resource growth suggests bottlenecks exist—immigration courts, receiving countries willing to accept deportees, transportation logistics. But the detention infrastructure is being built regardless.

Whether this infrastructure enables future mass removal, primarily serves as deterrence through detention, or represents wasteful over-building remains to be seen. But dismissing the possibility based on logistics underestimates the scale of resources now being deployed.

What's accurate: Current removal numbers haven't matched the infrastructure growth. What's uncertain: Whether this represents temporary bottlenecks or fundamental limits.

"This is unprecedented authoritarianism"

+

Historical context matters for this claim. Some aspects are genuinely novel; others have precedent.

Obama-era deportations peaked higher than current removal numbers. Detention populations hit similar levels in early 2020 before COVID.

What is genuinely novel in 2025:

  • The speed of infrastructure scaling (workforce doubled in 5 months)
  • The elimination of sensitive location protections (schools, churches, hospitals)
  • The shift away from criminal prioritization to broad enforcement
  • The use of 33,000+ detailees from other agencies
  • The budget scale ($28.7B is nearly triple the prior baseline)

"Unprecedented" requires specificity about which aspects. The machinery is more expansive than before, but the underlying legal authorities aren't new. Whether this represents a difference in degree or kind is a genuine question.

What's accurate: Several specific elements have no direct precedent. What's overstated: Claims that immigration enforcement itself is new or that current removal numbers exceed historical peaks.

"Interior enforcement is a new development"

+

Interior enforcement has always existed. What's changed is the balance.

Interior removals exceeded border deportations for the first time since 2014—a decade-long shift in enforcement geography.

The policy question is about the balance between border and interior enforcement and what that balance signals about enforcement philosophy. Border-focused enforcement primarily affects new arrivals; interior-focused enforcement affects established residents, often with deep community ties, U.S. citizen children, and years of presence.

The 2025 shift toward interior enforcement—enabled by the expanded expedited removal authority that now applies nationwide rather than just within 100 miles of the border—represents a change in who bears the enforcement burden. It's not that interior enforcement is new; it's that it's now dominant.

What's accurate: The interior-to-border ratio has shifted significantly. What's misleading: Suggestions that interior enforcement didn't previously exist.

How We Got Here: The 2025 Policy Timeline

The expansion didn't happen overnight. A series of policy changes, executive orders, and legislative actions created the current system.

January 20, 2025
Executive Orders Signed
Multiple immigration-focused executive orders signed on inauguration day, declaring a "national emergency" at the border and directing agency expansion.
January 21, 2025
Sensitive Locations Policy Rescinded
ICE ends decade-old protections limiting enforcement at schools, churches, hospitals, and courthouses. New policy relies on "agent discretion" rather than categorical limits.
January 29, 2025
Laken Riley Act Signed
Creates new mandatory detention categories for undocumented immigrants charged with theft, shoplifting, burglary, assault on law enforcement, or violent offenses.
March 2025
Expedited Removal Expanded Nationwide
Expedited removal now applies anywhere in the U.S. to anyone who cannot prove 2+ years of continuous presence. Eliminates the previous 100-mile border zone and 14-day arrival window.
July 4, 2025
"One Big Beautiful Bill" Signed
Allocates $75 billion over four years: $45B for detention expansion, $14B for transportation/removal, $8B for hiring. This is the funding mechanism for everything that followed.
July–November 2025
ICE Hiring Surge
12,000+ officers hired in under 5 months through $50,000 signing bonuses, shortened 6-week training (down from 13), and eliminated age requirements (minimum lowered to 18, upper limit removed).
October 2025
Interior Removals Exceed Border Deportations
For the first time since 2014, removals from the U.S. interior outnumber border deportations—a fundamental shift in enforcement geography.

The Workforce Question

Doubling an agency's workforce in five months raises inevitable questions about quality, training, and institutional culture.

Workforce Composition

Personnel by division (2025-2026 estimates)

Hiring Surge Timeline

Cumulative new hires, July-November 2025

Institutional Concern
The House Homeland Security Committee has requested a GAO review of ICE hiring practices, citing concerns about reduced vetting standards. Training was cut from 13 weeks to 6 weeks; age requirements were eliminated; 85%+ of new hires came from law enforcement backgrounds, but institutional acculturation necessarily suffers at this pace.

Beyond direct ICE hires, approximately 33,000 federal employees from other agencies—DOJ, IRS, CBP, USCIS, DOD—have been detailed to support ICE operations. This effectively triples the available workforce when including detailees, but raises questions about expertise and mission alignment.

The hiring surge creates institutional facts that persist. These 12,000+ new officers will remain federal employees across administrations. The infrastructure being built—both physical and human—represents long-term commitments regardless of future policy direction.

The Infrastructure: Offices and Facilities

The physical footprint of ICE operations across the United States.

ERO Field Office Network (25 offices)

Office Location States/Territories Covered Region
AtlantaAtlanta, GAGeorgia, North Carolina, South CarolinaSoutheast
BaltimoreBaltimore, MDMarylandNortheast
BostonBoston, MACT, ME, MA, NH, RI, VTNortheast
BuffaloBuffalo, NYUpstate New YorkNortheast
ChicagoChicago, ILIL, IN, WI, MO, KS, KYMidwest
DallasDallas, TXNorth Texas, OklahomaTexas
DenverDenver, COColorado, WyomingMountain
DetroitDetroit, MIMichigan, OhioMidwest
El PasoEl Paso, TXWest Texas, New MexicoTexas
HarlingenHarlingen, TXRio Grande ValleyTexas
HoustonHouston, TXSoutheast TexasTexas
Los AngelesLos Angeles, CASouthern CaliforniaWest
MiamiMiami, FLFlorida, Puerto Rico, USVISoutheast
NewarkNewark, NJNew JerseyNortheast
New OrleansNew Orleans, LALA, AR, MS, AL, TNSoutheast
New York CityNew York, NYNYC metro areaNortheast
PhiladelphiaPhiladelphia, PAPA, DE, WVNortheast
PhoenixPhoenix, AZArizonaSouthwest
Salt Lake CitySalt Lake City, UTUT, ID, MT, NVMountain
San AntonioSan Antonio, TXCentral TexasTexas
San DiegoSan Diego, CASan Diego & Imperial CountiesWest
San FranciscoSan Francisco, CANorthern CA, HI, Guam, SaipanWest
SeattleSeattle, WAWA, OR, AKNorthwest
St. PaulSt. Paul, MNMN, ND, SD, IA, NEMidwest
Washington DCWashington, DCDC, VirginiaNortheast

Major Detention Facilities (Top 15 by population)

Facility Name Location Operator Avg. Daily Pop.
Camp East MontanaEl Paso, TXICE/DOD2,774
California City DCCalifornia City, CACoreCivic2,560
South Texas FRCDilley, TXCoreCivic2,400
Adams County DCNatchez, MSCoreCivic2,224
Stewart DCLumpkin, GACoreCivic2,044
Adelanto IPCAdelanto, CAGEO Group1,940
South Texas IPCPearsall, TXGEO Group1,904
Winn CCWinnfield, LALaSalle1,576
Northwest IPCTacoma, WAGEO Group1,575
Otay Mesa DCSan Diego, CACoreCivic1,482
Port Isabel SPCLos Fresnos, TXICE-owned1,200
Krome North SPCMiami, FLICE-owned1,000
Karnes County RCKarnes City, TXGEO Group830
Broward TCPompano Beach, FLGEO Group700
Delaney HallNewark, NJGEO Group400

What This Means

The 2025 ICE expansion is real, substantial, and concentrated in specific geographies. The data supports several conclusions regardless of one's policy preferences:

Resource allocation reveals priorities. A 193% budget increase and 120% workforce expansion represent an enormous commitment of public resources. The $28.7 billion FY2025 budget exceeds the entire Department of Justice federal prison system. Where governments spend money tells us what they actually prioritize, independent of rhetoric.

Infrastructure creates lock-in. Private prison contracts, facility construction, and 12,000+ new federal employees generate institutional momentum that persists across administrations. The machinery being built in 2025 will exist—and require utilization or expensive dismantling—regardless of future policy direction.

Geography determines exposure. Immigration enforcement is not experienced uniformly. The five Texas field offices, the Louisiana-Mississippi-Georgia detention corridor, and the per-capita enforcement rates that vary by a factor of four between states create radically different lived realities for immigrant communities depending on state-level political decisions.

The metrics mismatch matters. The gap between detention growth (65%) and removal growth (7%) is the most analytically interesting tension in this data. Either the system is inefficient, detention itself is the policy goal, or external bottlenecks are binding. Each interpretation has different implications for what comes next.

The Bottom Line
The 2025 data shows a system optimized for detention rather than removal, concentrated in politically cooperative states, operated primarily by private contractors, and increasingly focused on non-criminal populations. Whether one views this as appropriate enforcement or problematic overreach, the data itself is clear about what's actually happening on the ground.