The Claim

Carried out over 650,000 arrests, detentions, and deportations of illegal aliens — including the worst of the worst criminal illegal alien killers, rapists, gang members, and repeat offenders.

The Claim, Unpacked

What is literally being asserted?

Two claims: (1) That the administration carried out “over 650,000” enforcement actions comprising arrests, detentions, and deportations; and (2) that these actions targeted “the worst of the worst” — specifically “killers, rapists, gang members, and repeat offenders.”

What is being implied but not asserted?

That 650,000 represents 650,000 individual people who were subject to enforcement. That the majority or all of these people were dangerous criminals. That the language “arrests, detentions, and deportations” describes a scale of operations rather than three overlapping counts of the same people. That this enforcement was historically unprecedented.

What is conspicuously absent?

Any explanation of how “arrests, detentions, and deportations” are counted. These are sequential stages in the enforcement pipeline — the same person is typically arrested, then detained, then deported. Are these three separate counts of the same individuals, summed together? If so, 650,000 “enforcement actions” could represent as few as 217,000 individual people (650,000 / 3). Any breakdown of what percentage of the 650,000 actually had violent criminal records. Any comparison to prior administrations. Any acknowledgment that independent data shows less than 14% of ICE arrests involved people with violent criminal records, and 73.6% of ICE detainees had no criminal conviction at all.

Evidence Assessment

Established Facts

The “650,000” figure appears to combine arrest, detention, and deportation counts — which overlap substantially because they describe sequential stages of the same enforcement process. DHS’s own December 2025 press release cited “more than 595,000” arrests and “more than 622,000” deportations as separate figures. These cannot simply be added together, because a person who is arrested is typically also detained and then deported — meaning the same individual appears in all three counts. ICE produces data with linked identifiers across arrest, detention, and deportation datasets, confirming these are overlapping populations. The Deportation Data Project (UCLA/UC Berkeley) documented this pipeline: “After ICE arrests and processes individuals… officials may detain them while their immigration cases are pending.” The “650,000” figure thus likely represents between 290,000-400,000 unique individuals (the independent deportation estimates), not 650,000 separate people. [^003-a1]

Less than 14% of ICE arrests in the first year involved people with violent criminal records. CBS News obtained an internal DHS document showing that of approximately 393,000 ICE arrests between January 21, 2025 and January 31, 2026, only 13.9% involved individuals charged with or convicted of violent crimes. Less than 2% had homicide or sexual assault charges/convictions. Nearly 40% had no criminal record whatsoever. This was independently corroborated by multiple analyses: FactCheck.org found 43% of recent ICE arrests had no criminal record; the Cato Institute found only 5% of ICE detainees had violent criminal convictions; the New York Times found 7% had violent convictions; and Poynter/PolitiFact found only 5-7% had violent convictions. [^003-a2]

73.6% of ICE detainees had no criminal conviction as of February 2026. TRAC (Syracuse University) data current as of February 7, 2026 shows that 50,259 out of 68,289 people held in ICE detention — 73.6% — had no criminal conviction. The number of immigrants detained with no criminal record grew from 3,165 in February 2025 to 25,193 in January 2026 — a 2,450% increase. Austin Kocher (Syracuse University/TRAC) documented that 92% of ICE detention growth in FY 2026 was driven by people with no criminal convictions. [^003-a3]

DHS itself reported approximately 7,000 gang member arrests — 1.8% of total ICE arrests. DHS’s own January 20, 2026 press release stated that 7,000 gang members were arrested from “American streets” during the first year. CBS News separately reported approximately 7,500 (1.9%) accused of gang affiliation. Against 393,000 total arrests, gang members represent less than 2% of the total — not the dominant category the claim’s language implies. [^003-a4]

Strong Inferences

The “650,000” figure is structurally designed to be as large as possible through aggregation of overlapping categories. DHS reported approximately 595,000 arrests and 622,000-675,000 deportations as separate metrics. The claim’s construction — “arrests, detentions, and deportations” — treats each stage of the enforcement pipeline as a separate achievement. A person arrested by ICE in February, detained through June, and deported in July would appear in all three counts. The claim does not say “650,000 individual people” — it says “650,000 arrests, detentions, and deportations,” a formulation that allows multiple counting of the same individuals. Even the arrest figure alone (595,000) is likely inflated: ICE’s counting methodology permits multiple arrests of the same individual to be counted separately, and the Deportation Data Project notes that ICE data includes “people who have more than one initial book-in.” [^003-a5]

The “worst of the worst” framing inverts the actual composition of enforcement. The claim foregrounds “killers, rapists, gang members” as representative of the enforcement population. The data shows the opposite: the “worst of the worst” categories constitute a small fraction of total enforcement actions. Homicide charges/convictions: approximately 2,100 (0.5% of arrests). Sexual assault: approximately 5,400 (1.4%). Gang members: approximately 7,000 (1.8%). Combined “worst of the worst”: approximately 14,500, or roughly 3.7% of the 393,000 arrests. The remaining 96.3% were arrested for other reasons — including the 40% with no criminal record at all. The rhetorical structure places the exceptional cases first and uses them to characterize the entire enforcement operation. [^003-a6]

ICE arrest patterns shifted dramatically toward people without criminal histories over the course of the year. FactCheck.org documented that the share of ICE arrests involving people with no criminal record rose from 21.9% in the first three months to 43% by January 2026. Conversely, the share with criminal convictions dropped from 44.7% to 31.8%. The Deportation Data Project found that arrests of people without criminal convictions rose 700%, while arrests of people with violent crime convictions increased only 30%. This pattern — enforcement increasingly targeting non-criminals — contradicts the “worst of the worst” framing. [^003-a7]

This claim significantly overlaps with Items #2 and #4. Item #2 claims “2.6 million” removals. Item #3 claims “650,000 arrests, detentions, and deportations.” Item #4 claims “400,000 criminal aliens” deported. These draw from the same pool of enforcement actions. The deportation component of Item #3 is a subset of Item #2’s deportation component. The criminal aliens in Item #3 overlap entirely with Item #4. A reader encountering these sequentially could reasonably add them together, creating an inflated impression of the total scale of enforcement. [^003-a8]

Informed Speculation

The construction of the claim — “arrests, detentions, AND deportations” — may be deliberately ambiguous. The word “and” can be read as describing a unified category (arrests-plus-detentions-plus-deportations as one combined figure) or as listing three separate types of action. The former reading makes the number sound like 650,000 individual enforcement encounters. The latter reading acknowledges the overlap but loses the impressive-sounding scale. The ambiguity likely serves the purpose: audiences who read quickly will register “650,000 people,” while the technically defensible reading — “650,000 total enforcement actions across overlapping categories” — remains available as a fallback.

The “worst of the worst” framing, combined with the DHS practice of publishing individual case press releases about specific dangerous criminals (DHS issued at least 15 separate “worst of the worst” press releases in January 2026 alone), creates a narrative environment where anecdotes crowd out data. A reader who has seen stories about individual murderers and rapists will naturally assume these cases are representative. The data shows they are exceptional — less than 4% of the total.

Structural Analysis

Cui bono from the framing: The claim serves dual purposes: it makes the enforcement operation sound larger than it is (by aggregating overlapping counts) and more targeted at dangerous criminals than it is (by foregrounding the “worst of the worst” while the data shows the majority had no violent criminal record). This framing justifies the $170 billion congressional appropriation for immigration enforcement and the expansion of ICE from 6,000 to 18,000 agents.

The padding lens: The “650,000” figure is a textbook example of metric padding. By listing “arrests, detentions, and deportations” as if they were additive rather than sequential stages of the same process, the claim inflates the apparent scale of operations by roughly 2-3x. The actual number of unique individuals affected is likely in the range of 290,000-400,000 (the independent deportation estimates from TRAC and MPI).

Stated vs. revealed preferences: The administration states it is targeting the “worst of the worst.” The revealed preference — documented by Kocher, TRAC, CBS, FactCheck.org, and the Deportation Data Project — is to arrest and detain increasingly large numbers of people with no criminal record. The 700% increase in arrests of people without criminal convictions vs. the 30% increase in arrests of people with violent convictions tells the real story: the enforcement apparatus has expanded primarily by sweeping up non-criminals.

The denominator problem: The claim presents absolute numbers without denominators. “Over 650,000” sounds massive. Against the estimated 11 million unauthorized immigrants in the United States, it represents approximately 4-6% of the target population (and likely less, because the 650,000 includes substantial double-counting). Against annual federal arrest totals across all categories, it provides essential context about relative enforcement priorities.

Context the Framing Omits

Prior administrations conducted comparable enforcement without claiming it as historically unprecedented. Obama deported approximately 400,000 people per year during his first term and was labeled “deporter-in-chief.” Biden’s FY 2024 saw approximately 778,000 total repatriations. Trump’s independently verified deportation figures (290,000-400,000 through the first year) are below these benchmarks, not above them.

The “70% criminal aliens” claim relies on conflating charges with convictions. DHS Secretary Noem repeatedly claimed 70% of ICE detainees are “violent criminals.” PolitiFact rated this claim as misleading: the actual figure is approximately 26% with criminal convictions and 26% with pending charges (which may never result in conviction). Only 5-7% had violent crime convictions. Traffic violations were the most common criminal offense among those with convictions.

Detention costs are substantial and rising. The Brennan Center documented that ICE detention costs approximately $152 per person per day. With 68,000+ people in detention — 73.6% having no criminal conviction — the annual cost of detaining non-criminals alone exceeds $2.7 billion. Case management alternatives cost $4 per day and achieve 99% compliance rates.

The claim overlaps with Items #2 and #4, creating an inflated total impression. Item #2 (2.6 million removals), Item #3 (650,000 arrests/detentions/deportations), and Item #4 (400,000 criminal aliens) are not independent counts — they draw from the same underlying enforcement actions. A reader encountering all three is likely to sum them, producing an implied total of 3.65 million enforcement actions, when the actual number of unique individuals affected is likely below 500,000.

Verdict

Factual core: Misleading. The administration did conduct significant immigration enforcement — that is not in dispute. ICE arrests increased substantially, reaching approximately 393,000 in the first year. But the “650,000” figure inflates the scale by aggregating overlapping categories (arrest, detention, deportation of the same people). The actual number of unique individuals subject to enforcement is likely in the range of 290,000-400,000 based on independent estimates. The “worst of the worst” framing is contradicted by the administration’s own data: less than 14% of arrests involved violent criminal records, less than 2% involved gang membership, and 73.6% of detainees had no criminal conviction.

Framing as “win”: Misleading. The claim uses two techniques to inflate perceived impact: (1) aggregating overlapping counts to produce a large number, and (2) using emotionally loaded language (“killers, rapists, gang members”) that characterizes less than 4% of enforcement actions as representative of the whole. The enforcement expansion has primarily targeted people without criminal records — the opposite of the “worst of the worst” narrative.

What a reader should understand: ICE did arrest approximately 393,000 people in the first year — a real increase in enforcement. But the “650,000” figure counts the same people multiple times across arrest, detention, and deportation. The “worst of the worst” language describes less than 4% of those arrested. The majority of the enforcement expansion has been directed at people with no criminal record. The claim, like Items #2 and #4, draws from the same pool of enforcement actions, and the three items together create an inflated impression of an enforcement operation that, by independent analysis, is comparable in scale to prior administrations.

Cross-References

• Item #1: “Negative net migration” — overlapping narrative of enforcement scale
• Item #2: “2.6 million removals” — the deportation component of this claim is a subset of Item #2; same double-counting and methodology concerns apply
• Item #4: “400,000 criminal aliens deported” — directly overlaps with the “criminal” subset of this claim; the criminal breakdown data analyzed here applies there
• Item #5: “Two million self-deportations” — the deterrence narrative relies on enforcement scale claims like this one
• Item #6: “Lowest border crossings since 1970s” — related enforcement narrative; border decline began under Biden

Sources

Brennan Center for Justice. “The Deportation Industrial Complex: Arrests and Detentions by the Numbers.” 2025. https://www.brennancenter.org/our-work/analysis-opinion/deportation-industrial-complex-arrests-and-detentions-numbers

Cato Institute. “5% of People Detained By ICE Have Violent Convictions, 73% No Convictions.” 2025. https://www.cato.org/blog/5-ice-detainees-have-violent-convictions-73-no-convictions

CBS News. “Less Than 14% of Those Arrested by ICE in Trump’s First Year Had Violent Criminal Records.” January 18, 2026. https://www.cbsnews.com/news/ice-arrests-violent-criminal-records-trump-first-year/

DHS. “DHS Sets the Stage for Another Historic, Record-Breaking Year Under President Trump.” January 20, 2026. https://www.dhs.gov/news/2026/01/20/dhs-sets-stage-another-historic-record-breaking-year-under-president-trump

DHS. “MAKING AMERICA SAFE: DHS Arrested 7,000 Gang Members from American Streets in First Year of Trump Administration.” January 20, 2026. https://www.dhs.gov/news/2026/01/20/making-america-safe-dhs-arrested-7000-gang-members-american-streets-first-year

DHS. “Thanks to President Trump and Secretary Noem, More than 2.5 Million Illegal Aliens Left the U.S.” December 10, 2025. https://www.dhs.gov/news/2025/12/10/thanks-president-trump-and-secretary-noem-more-25-million-illegal-aliens-left-us

Deportation Data Project (UCLA/UC Berkeley). “Immigration Enforcement in the First Nine Months of the Second Trump Administration.” January 27, 2026. https://deportationdata.org/analysis/immigration-enforcement-first-nine-months-trump.html

FactCheck.org. “As ICE Arrests Increased, a Higher Portion Had No U.S. Criminal Record.” January 2026. https://www.factcheck.org/2026/01/as-ice-arrests-increased-a-higher-portion-had-no-u-s-criminal-record/

Kocher, Austin. “92% of ICE Detention Growth in FY 2026 Driven by Immigrants with No Criminal Convictions.” January 2026. https://austinkocher.substack.com/p/92-of-ice-detention-growth-in-fy

Kocher, Austin. “Show Your Work: The Math Behind My Claim that 92% of ICE’s Detention Growth Comes from People with No Criminal Convictions.” January 2026. https://austinkocher.substack.com/p/show-your-work-the-math-behind-my

Migration Policy Institute. “Unleashing Power in New Ways: Immigration in the First Year of Trump 2.0.” January 2026. https://www.migrationpolicy.org/article/trump-2-immigration-1st-year

NBC News. “Immigration Enforcement Tracker.” Updated March 2, 2026. https://www.nbcnews.com/data-graphics/us-immigration-tracker-follow-arrests-detentions-border-crossings-rcna189148

PolitiFact. “Do 70% of Immigrant Detainees Have Criminal Convictions or Charges? Fact-Checking Kristi Noem.” January 23, 2026. https://www.politifact.com/article/2026/jan/23/Kristi-NoemICE-detention-criminal-conviction-70/

Poynter. “Kristi Noem Said Most Immigrants in ICE Detention Are Violent Criminals. The Data Says Otherwise.” 2026. https://www.poynter.org/fact-checking/2026/how-many-arrested-immigrants-violent-crime-convictions/

TRAC (Syracuse University). “Immigration Detention Quick Facts.” Data through February 7, 2026. https://tracreports.org/immigration/quickfacts/

TRAC (Syracuse University). “Taking Stock: Trump Administration Record on Detention and Removals.” November 2025. https://tracreports.org/reports/767/