The Trump administration in 2025 set a limit of 7,500 for refugees for FY 2026, but it then set a monthly target of 4,500 white South Africkaner refugees — that is, a an annual target of about 50,000. Go here. Also, here is the executive order. Settlement in the U.S. appears to be done independently of establishee refugee settlement infrastructure, which the administration, as it attempted in the first Trump administration, seeks to destroy.

It has been well known that arrests by ICE of unauthorized persons have been trending towards persons without a criminal record. This report provides clear evidence of how that trend evolved in 2025. The report looks at arrests through October 2025.

In October, 2024, an average of 200 persons a day were taken by ICE from local jails or other lock ups, versus about 75 from worksites and other locations.  By October 2025, the average daily arrest rate for each hovered around 550.

Throughout, DHS was saying that it was detaining the “worst of the worst.” It could loosely back that up with the arrests from jails and from the probably extremely low numbers of persons with a lurid criminal background that ICE arrested on the streets.

However, the trends documented in this report indicate that to boost arrests from the 1,000 person a day level to much higher requires either more states to agree to allow coordination between local jails and ICE, or to arrest a lot more outside of jails. This is not news.

A question: does this history in any way imply in any reasonable way that unauthorized persons are relatively more criminal than authorized persons?  One can firmly say no – but there is one characteristic that should be recognized.  I do not have the figures before me now, but I believe that, at least perhaps 15 years ago, the unauthorized populations contained relatively more poor, young men – whose run-ins with the law are higher in American society.   The studies showing that the crime rate among unauthorized is lower than among others I trust – but also note that the demographics, at least in the past, might suggest otherwise.

In a court case before an Alabama district court, attorneys say DHS has adopted a three-pronged policy of aggressive arrest and detention. The case involves Leonardo Garcia Venegas, a U.S. citizen construction worker of Mexican descent, twice detained during immigration raids at private, non-public construction sites despite presenting an Alabama REAL ID:

The brief alleges DHS has adopted a Warrantless Entry Policy permitting officers to enter fenced, posted construction sites and partially built homes without judicial warrants, treating them as “open areas.” It also alleges a Preemptive Detention Policy under which officers detain Latino construction workers based on appearance and occupation alone, without individualized suspicion, effectively replacing targeted investigation with mass, location-based sweeps. Finally, it alleges a Continued Detention Policy authorizing officers to maintain custody even after individuals present documentation establishing lawful presence, including DHS-certified REAL IDs.

They assert there are  agency-wide directives reflecting a reinterpretation of DHS authority, constitute final agency action, and violate the Fourth Amendment, federal statutes, and binding regulations. The plaintiffs do not have smoking gun internal documents. They rely on public DHS press releases and memoranda, including references to a Lyons Memo authorizing use of administrative warrants, public statements, ICE training materials cited in other litigation, news reports, and repeated on-the-ground enforcement conduct, including the plaintiff’s own detentions.

Operation Midway Blitz was a highly publicized immigration enforcement campaign launched by ICE and other federal agencies in the Chicago area. It began in September 6-9, 2025, deploying hundreds of ICE agents plus National Guard, ostensibly targeting undocumented immigrants with criminal records.  In retrospect, this campaign was intended as the first of a large number to be waged against blue cities, with an ill-trained and led force, with the primary intent to create a war zone atmosphere which would work to the administration’s benefit. This kind of campaign was effectively ended by Minneapolis.

Henceforth, for DHS to meet its goal of well over 1,000 arrests a day it will have to arrest persons at work or in their homes — which will provoke perhaps a broader negative response from the public and courts.

DHS’ press release said, “This ICE operation will target the criminal illegal aliens who flocked to Chicago and Illinois because they knew Governor Pritzker and his sanctuary policies would protect them and allow them to roam free on American streets.”

DHS sought to invoke shock and awe, in an approach that was later discredited in its Minneapolis campaign. The first full day of the campaign,  a TV station reported “On an anxiety-filled day for the city, there were sporadic reports of arrests in Chicago’s Latino community as hundreds of federal agents move in — and many residents hunker down.”

By mid-November, DHS claimed nearly many arrests across six Midwestern states. This figure was not significantly higher than what might have happened had DHS not undertaken such a publicly provocative campaign. DHS’ figures have been suspect from early in the Trump administration. Out of 1,852 people arrested in the Chicago area before October 7, as many as 1,100 had already been deported or had taken voluntary departure. (Go here.)

Protests grew at the Broadview ICE facility. Federal agents rdeployed tear gas, pepper balls, and flashbang grenades. Silverio Villegas-Gonzalez was fatally shot by ICE. Marimar Martinez was shot five times. Mayor Brandon Johnson and Governor J.B. Pritzker actively opposed the operation, with Johnson signing an executive order barring federal use of city property for immigration enforcement.

The operation wound down by mid-November 2025. Border Patrol Commander Bovino and approximately 150 agents departed for Charlotte, North Carolina. The National Guard withdrew from Chicago on December 31, 2025, after being stationed for only one day in October before courts blocked further deployment.

Antecedents to Minneapolis

Arising out of the Chicago campaign were ICE interpretations of the law which gave its personnel wide discretion on whom to arrest.  A “flight risk” was not just the risk of a person upon encounter to flee, but any risk that the person might not show up at a future hearing. In November a federal judge in Chicago ordered the release of 600 persons whom he said had been illegally detained by warrantless arrests. This issue of arrest discretion has yet to be definitively addressed by the courts.   (Go here, here, here and here).

However, DHS credibility was challenged. Judge Sara Ellis called Border Patrol Commander Gregory Bovino’s testimony “not credible,” accusing him of “outright lying” about incidents.

The single most controversial case was that of Marimor Martinez.  In October 2025, citizen Martinez was shot five times by a CBP agent. DHS claimed she had used her car as a weapon and labeled her a “domestic terrorist,” leading to federal assault and attempted‑murder charges. Later‑released video, diagrams, and texts undermined agents’ accounts, showing that they rammed her car and then opened fire. Prosecutors dismissed all charges with prejudice.  (Go here)

Latino voters are poised to play a pivotal role in determining control of Congress in 2026.

Redistricting strategies built on static Latino support are now highly vulnerable. Small shifts in Latino voter behavior can upend partisan control. I track here the Hispanic vote through many elections, and note that the 2024 Hispanic vote for Trump was about 10 points higher.  This is in the process of reversing.

The American Business Immigration Coalition gets into seat-specific analysis for the  Hispanic vote in November 2026. It analyzed eight Texas and Florida seats

In Texas, the report says looks at five newly drawn congressional districts where eligible Hispanic voters is high enough to affect outcomes. Under a scenario of a 10% more toward Democratic candidate than the 2024 vote, at least one formerly Republican-leaning seat would become competitive. With a somewhat larger 15% shift, two Republican-held districts, particularly the open seats in South and South-Central Texas, would move Democratic. In sum, a reasonable estimate is that 1 to 2 Republican seats could shift Democratic.

The news from South Texas is that disruption to workforces and households from immigrant law enforcement is turning the area away from Republicans.

In Florida, two of three Miami area districts could move into Democratic-leaning territory even with a 10% Hispanic swing, and the third becomes competitive. With a stronger 15% swing, all three districts trend Democratic. In sum, a plausible estimate is that 2 Republican seats would flip to Democratic control.

Caution: the Hispanic voting deficit

The number of eligible Hispanic voters has been increasing by about 3% a year. The white vote has been declining. But Hispanics are poor in registering to vote: only 40% of eligible Hispanic voters get around to register and vote, compared with 55% for whites. Much of this can be explained by whites being older and with more formal education, factors associated with higher voting rates.

The Trump administration is most likely planning to announce sometime this year that it has evidence that large numbers of non-citizens are on state voting rolls, aided by non-profits it alleges recruit non-citizens.  The DOJ and their State attorney general allies will (1) cite large number of non-citizens on voter rolls, (2) sue to close down non-profits which work to register naturalized immigrants and others, and (3) conduct highly publicized raids to collect data.

What the administration has done so far

According to the Brennan Center (as of  February 13) since mid 2025 the DOJ has asked states to sign a confidential memorandum of understanding through which the states provide voter rolls and other data. Here is a copy of the memorandum.

11 states have either provided or said they will provide their full statewide voter registration lists, including driver’s license and Social Security numbers: Alaska, Arkansas, Indiana, Kansas, Louisiana, Mississippi, Nebraska, South Dakota, Tennessee, Texas, and Wyoming.

The Brennan Center writes that the DOJ has sued Washington, DC, and 24 states for refusing to provide their statewide voter registration lists with driver’s license and Social Security numbers: Arizona, California, Colorado, Connecticut, Delaware, Georgia, Hawai’i, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, Nevada, New Hampshire, New Mexico, New York, Oregon, Pennsylvania, Rhode Island, Vermont, Virginia, Wisconsin, and Washington. The cases in California, Michigan, and Oregon have been dismissed.

On January 28, 2026, the FBI executed a search warrant for records related to the 2020 election at the Fulton County, GA Election Hub and Operation Center.  This is a one-off gambit tied expressly to the 2020 elections and the administration has no way to copy this in other states.

What the DOJ will attempt to do

A half dozen or more states have in the past ten years tried to identify non-citizens who are on the voter rolls. The standard approach is to match the citizen/non-citizen designation on Registry of Motor Vehicle records with voter rolls.  There are many cases where a person on the voter rolls has self-identified as a non-citizen.  When data system inconsistences are revolved, this number shrinks by 95-99%, and among the remainders there may be no evidence of actual voting.  The overwhelmingly dominant reason for the apparently positive match is that state data bases are not aligned to reconcile continuously and because drivers mistakenly list themselves a non-citizens on RMV records.

These investigations have uniformly ended in embarrassment (here).

Texas serves as an example

In 2019, Texas Secretary of State David Whitley, using the standard method of matching state databases, assigned 95,000 voters for citizenship checks and set them up for possible criminal investigation.(Go here.) That campaign was terminated and resulted in a settlement of three federal lawsuits. Whitley resigned his job due to the fallout. Texas had at the time about 17 million registered voters. Following the settlement, some counties undertook a more careful protocol to identify non-citizens on voting lists. They found several hundred. One person was a staff member of the El Paso County election staff and had had a naturalization party at the office years before.

On August 21 2024, Attorney General Ken Paxton, according to the office’s press release, “has opened an investigation into reports [sic] that organizations operating in Texas may be unlawfully registering noncitizens to vote in violation of state and federal law.” Paxton said that his investigation involved undercover agents and raids on homes of individuals.

Paxton then tried to close down a Latino organization, Jolt Initiative. Per the Texas Tribune, Paxton filed a lawsuit in state court accusing Jolt of submitting “unlawful voter registration applications,” specifying in a press release that the group was “attempting to register illegals, who are all criminals.” The suit sought to revoke Jolt’s nonprofit charter through a legal mechanism known as a quo warranto petition. A Federal judge ruled against Paxton in November 2025.