Unsupervised ICE: Built for Violence

The supervisor standing between an ICE agent and tragedy now has 73% more substandard agents to watch.

Part 1 established that two-thirds of ICE's Enforcement and Removal Operations (ERO) surge hires have no discernible law enforcement background. Part 2 established that the academy is not screening them — attrition fell to one-third of historical levels as it surrendered its role of weeding out problem candidates.

So who watches these undertrained, unfiltered officers once they hit the field?

In nine months, ICE first-line supervision ratio degraded 73% — from 6.4 agents per supervisor to 11.2. The professional benchmark is 6:1. ICE blew past this supervisor ratio and kept going.

These are not abstract management ratios. The first-line supervisor — called a Supervisory Detention and Deportation Officer, or SDDO — is the person who rides along on operations, reviews use-of-force decisions, and catches problem behaviors to prevent misconduct like shooting unarmed civilians. That SDDO now has 73% more agents to watch — most of them inexperienced hires from Part 1 who survived the turnstile academy in Part 2 and were then dropped into the field for Part 3.

First-Line Supervision Matters

In law enforcement, this supervision is not a level of bureaucracy — it is a safety and training mechanism. A SDDO is responsible for managing field agents' conduct and ensuring they all understand and conform to best practices in the field.

The Police Executive Research Forum (PERF), which partners with the Department of Justice's COPS Office on law enforcement research and training, surveyed agencies nationally in 2018 and found that agencies reported an ideal first-line supervisor-to-officer ratio of 6:1. This reflects a consensus on the practical limit of how many armed officers one supervisor can meaningfully oversee during operations. A State Department synthesis of law enforcement studies found that officers receiving active supervisory review were up to 50% less likely to use force in subsequent weeks.

While not a direct mapping onto federal law enforcement, it does show the importance of front-line supervision — and the limits of any one supervisor.

ICE Enforcement and Removal Operations supervisory hierarchy works like this:

The SDDO is who matters here. Other supervisors are important, but 66% of all ERO supervisory personnel are these SDDO – on the ground when field agents do everything from stopping a car to killing civilians. By focusing our analysis on SDDOs, we are tracking the workforce in position to actually mitigate harm in real time.

Killing Supervision

I tracked ERO's workforce month by month from February through November 2025 using OPM's Federal Workforce Data portal. The pattern is consistent across months, and the projected trajectory is severe.

Supervisors are leaving and ICE is not replacing them.

In February, ERO was close to the 6:1 benchmark ratio at 6.4. By November, each SDDO was responsible for 11.2 agents. A 73% increase in nine months.

The field agent count grew 69% while the number of SDDOs shrank. The surge poured thousands of new agents into the field while the first-line supervisory layer quietly contracted beneath them. Look at September: the month the hiring surge visibly hit, agent headcount jumped by over 1,100 in a single month. The SDDO count dropped.

To restore the professional 6:1 standard today, ERO would need 1,565 SDDOs — 725 more than it has. That is an 86% increase in the number of first-line field supervisors. But there is no evidence in the data for any a pipeline producing them.

SDDOs Retire

The SDDOs are not quitting in protest. They are not being fired. They are retiring. Between February and November 2025, 38 SDDOs departed. Of those, 35 — 92% — retired. One quit. One left for other reasons. Zero were terminated.

The average SDDO who retired had 26.4 years of federal law enforcement experience. That is 26 years of knowing when a situation is escalating. Twenty-six years of practical experience in training and managing field agents just walks out the door with every retirement. Federal law enforcement officers can retire at 50 with 20 years of service, or at any age with 25. These SDDOs are leaving because they have earned the right to. ICE is just not replacing them at a rate to address their ratio problem.

Replacements Scarce

To be clear: ICE did promote agents into SDDO positions in 2025. Internal promotions happened every month — in May, roughly 16 agents were promoted to SDDO to replace 8 who retired. But then September hits. 1,131 new agents in one month. Then 1,102 more in October. Then 1,383 more in November. Internal promotion does not match that pace. In November, ICE promoted 37 new SDDOs — the biggest single-month influx of first-line supervisors in the entire period. Four outside supervisors transferred into ERO from other agencies during this time. These hires moved the ratio from what would have been 12.5:1 down to 11.2:1.

It barely dented the problem. In fact, the net from promotions against departure left ICE with 22 fewer SDDO in November 2025 than just in February that year.

And here is where it gets worse: the data shows who got promoted. The November SDDO promotions had a higher median Length of Service (LOS) than the supervisors they replaced. Retirement-eligible SDDOs — those with 20 or more years of service — held steady at around 45% of the corps even as headcount fluctuated (387 in February, 380 in November). Roughly 58% of the new SDDOs were already retirement-eligible the day they became supervisors.

The average SDDO's length of service keeps rising — from 19.7 years in February to 20.2 years by November. So ICE is filling the SDDO ranks with people who will likely have shorter tenure in office than their predecessors. The first-line supervisory pool is aging in place. Half of ERO's supervisory corps can walk out the door tomorrow. Many will, and when they do, the SDDO ratio — already at 11.2:1 — will get worse.

ICE's response to the SDDO shortage is promoting an insufficient number of people who could themselves walk out the door tomorrow retired just as their predecessors did in 2025.

Hell's New Babysitters

The compound problem comes into focus when you look at how the agent workforce changed beneath the SDDOs' feet. In February, before the surge, field agents were overwhelmingly experienced. By November, they were not.

In February, 97% field agents had at least a year of experience. By November, one in four had less than a year on the job. By December, at observed rates, it will be one in three. The experienced core barely grew — it was the rookie layer that exploded, from 157 to 2,427 in nine months.

So, in February, a SDDO supervised 6.4 agents, only 3% of whom had less than a year of experience. By November, each SDDO watched roughly 11 field agents, with 25% being rookies. By December, at projected rates, each SDDO would watch roughly 12.5 — with nearly 33% rookies. The SDDO cohort trying to ride herd during field operations will increasingly consist of newly promoted supervisors — individuals with years in the field but only weeks acquiring supervisory skills — while simultaneously managing an ever more inexperienced agent corps.

A Purposeful Failure

Just for comparison and some perspective, the 2006 Border Patrol surge, which saw roughly 20% of recruits fail to graduate from the academy and had stable supervision ratios, still produced 144 corruption arrests and 125 convictions. ICE's 2025 surge has one-fifth the academy filtering and nearly double the first-line supervisory strain.

Nothing outlined in this series relies on exotic or arcane truths about public safety administration to evaluate ICE. Rather, these articles show how ICE departs from well-known, established, and widespread practices. The three parts of this series detail, with numbers, how disturbing the ascension of ICE is for the public.

Part 1 — The Lie: DHS claims 85% of surge hires are experienced law enforcement officers. OPM payroll data shows 66.5% are true rookies with no discernible law enforcement background.

Part 2 — The Rubber Stamp: The academy that is supposed to filter unqualified recruits has stopped doing so. Attrition collapsed from 12.2% to 4.2%, and the ratio of departures inverted — 74% quit, only 26% are removed by administration. Two out of three trainees who would have washed out in past years are now put in the field.

Part 3 — Unsupervised: First-line SDDO supervision degraded 73% in nine months. Each first-line supervisor now watches nearly twice the professionally recommended number of field agents. The agents being supervised are increasingly inexperienced. And there is no indication ICE promotions will ever be enough to return the supervisory ratio to the standard 6 to 1.

At this point, failure is no longer isolated or accidental: it has persisted across hiring, training, and supervision despite readily available benchmarks and internal data that call out for change. These are not three separate problems, but one unifying theme – ICE designed to foster violence and intimidation.

Lord of the ICE Flies

The Trump administration is creating a "Lord of the Flies" ICE designed to hire and keep the worst of us and allow them to follow all their worst instincts in the field without restraint. There is no other way to understand such a willful violation of basic standards in law enforcement management.

This is no law enforcement. It is organized thuggery — Trump’s own Sturmabteilung in brown camouflage — turned loose on political targets in cities across the United States. The record reveals a minimally supervised cadre of agents, armed with unchecked views of their powers and venomous racism, organizing themselves around a single principle: “What would Trump want me to do?”

I know my numbers and charts are of no immediate help to people across this country suffering at the hands of ICE. This work is being taken on by neighbors protecting neighbors everyday across the United States.

My hope is this work gives opponents more insight for future political fights against ICE and provides a framework for history to judge what is happening today.

This data is limited in that it does not allow individuals to be tracked over time; the analysis therefore relies on cohort than person-level outcomes. Overall, while not perfect data, it does reveal the directions and magnitude of ICE's deliberate malpractice.

Data Sources

All data is publicly available from the Office of Personnel Management:

FedScope/data.opm.gov: https://data.opm.gov (replacement for FedScope, which sunset January 28, 2026)

Employment Stock Data: Federal Workforce Data monthly files, February–November 2025

Separations Data: Monthly files, February–November 2025

Filters Applied:

• Agency: HSBB (Immigration and Customs Enforcement)
• Occupation: 1801 (Enforcement and Removal Operations)
• Supervisory Status: 2, 5 (supervisors); 8 (field agents)
• First-line supervisor defined as GS-13 SDDO (Supervisory Detention and Deportation Officer)
• Data window: February–November 2025 (same-platform FWD comparison)

Key External Sources:

• PERF, "Promoting Excellence in First-Line Supervision" (2018), p. 19 — 6:1 ideal supervisor-to-officer ratio
• GAO-13-59 (December 2012) — 2006 Border Patrol surge: 144 corruption arrests, 125 convictions
• GAO-07-997T (June 2007) — Border Patrol academy costs and attrition during 2006 surge
• U.S. Department of State, "Best Practices for Curbing Excessive Use of Force" (January 2024) — Active supervisory review reduces subsequent use of force by up to 50%
• DHS OIG-19-07 (November 2018) — ERO organizational structure and training
• USAJOBS vacancy announcements — SDDO (GS-1801-13) position descriptions

Methodological Notes:

**December data from OPM is not yet available for publication. This projection reflects the narrow band (12.3:1 to 12.8:1) produced by four different estimation scenarios, all of which converge because the underlying dynamic is consistent: agents pouring in, SDDOs flat or declining.

Position title mapping (GS-13 = SDDO) is inferred from USAJOBS vacancy announcements for Supervisory Detention and Deportation Officer positions, ICE ERO leadership biographies, and career ladder documentation. Some GS-13 supervisors may hold non-SDDO positions such as instructors or program managers. The mapping is approximate but consistent with published ERO organizational structure.

Internal promotions are not directly observable in OPM data. They are inferred from month-over-month stock changes minus external hires visible in accessions data. The November SDDO increase of +33 with only 4 external hires for the entire period is consistent with a batch promotion event.

This analysis uses the same-platform Federal Workforce Data (FWD) window from February through November 2025. No cross-platform comparisons are required, avoiding the methodological complications of comparing legacy FedScope to the replacement FWD system.

This concludes the three-part series on ICE's hiring surge. The full dataset, methodology, and replication materials are available on request.