A federal judge has denied a request by former Milwaukee judge Hannah Dugan to either overturn her December conviction for obstructing an immigration proceeding. In a 32-page decision filed Tuesday, U.S. District Judge Lynn Adelman disagreed with Dugan’s attorneys that a recent appeals court decision that immigration enforcement actions are not “proceedings” means  her conviction should be overturned. 

In early 2025, Dugan was in court when she learned of that plain-clothes immigration agents were in the hall outside her courtroom, waiting to arrest Eduardo Flores-Ruiz, who was appearing for a routine hearing, for being in the country illegally. How to handle immigration enforcement had been a top concern among local judges, following a string of arrests by federal agents in and around the Milwaukee County Courthouse. Neither the county sheriff nor the Milwaukee Police Department participate in immigration enforcement and detention. 

Dugan confronted the agents in her judge’s robes, and directed them to check in with the chief judge. When most of the arrest team went to the office, Dugan went back into her courtroom and called Flores-Ruiz, set a new court date, and then led  him and his attorney out of the courtroom through a non-public hallway. Flores-Ruiz rode down in an elevator with one of the federal agents and was arrested outside after a brief foot pursuit. Dugan was later charged with obstructing agents and concealing Flores-Ruiz from them.

The appeals court decision centered around United States v. Hernandez, involving a man who’d been charged with obstruction after escaping immigration custody. In that case, the defendant argued that executing his removal order was not part of the proceeding, which ended once the final removal order was issued. When Dugan went to trial, the Hernandez case was still on appeal. Adelman instructed Dugan’s jury that “‘pending proceeding’ simply means any process taking place in the manner and form prescribed for conducting business by or before a department or government agency, including all steps and stages in such an action from its inception to its conclusion.” Federal agents testified that there are a number of steps in an immigration proceeding, beginning before an arrest and continuing through removal.

When Adelman approved the jury instructions, which were recommended by prosecutors, he relied on United States v. Hernandez. Adelman said that the instructions proposed by the defense team defined “proceeding” too narrowly. “In the present case, the government alleged that defendant obstructed an arrest (step eight),” Adelman wrote in his decision denying Dugan acquittal or a re-trial. Adelman also wrote that motions for reconsideration serve the limited function of correcting errors of law or fact, or to present newly discovered evidence. “Motions for reconsideration are not to be granted lightly,” Adelman wrote. 

Dugan’s attorneys argued that Adelman originally relied on the Hernandez case in deciding that “proceeding” was broad enough to include obstructing immigration agents from executing an administrative warrant. Now that an appeals court reversed the decision in Hernandez and ruled that  immigration enforcement is not a “pending proceeding” as presented to Dugan’s jury, they said Adelman should reconsider his own decision and grant an acquittal. 

Prosecutors countered that the Hernandez ruling “is neither binding nor persuasive, and thus fails to satisfy the ‘heavy burden required for consideration,’” according to Adelman’s summary of the arguments. 

“The problem for the defense is that this case did not involve some random encounter on the street,” Adelman wrote. “It was a targeted operation, conducted pursuant to agency procedures, including the issuance of an arrest warrant for a specific person.” Adelman also wrote that classifying immigration enforcement as “mere police activity” is inappropriate, as “the FBI, ICE can issue its own warrants and adjudicate and effectuate a removal…without the involvement of a court.”

For those reasons, Adelman ruled that the burden for reconsideration in Dugan’s case had not been met. A new court date for Dugan’s sentencing or other proceedings was not yet available Tuesday in online court records.

A quarter of immigration arrests since August were labeled by U.S. Immigration and Customs Enforcement as “collateral,” a type of arrest and detention that’s been challenged in court as an end run around civil rights.

Public outrage and lawsuits over the arrests may be tamping down the large-scale sweeps that foster them, but tens of thousands were arrested this way between August and early March.

Immigration arrests are usually based on warrants obtained ahead of time, showing either a removal order from immigration court or evidence of a crime or charge that makes the person subject to deportation.

But collateral arrests can result from street sweeps and raids in which a person is singled out for questioning based on appearance or proximity to someone wanted on a warrant. That person could be taken into custody if agents think they may be subject to deportation and also likely to flee if released.

Labeled for the first time ever, the collateral arrests are reported from August to early March in ICE arrest data obtained by the Deportation Data Project and analyzed by Stateline. In that time there were about 64,000 collateral arrests, a quarter of the 253,000 total arrests by ICE.

About 70% of the collateral arrests were for people with immigration-related crimes or violations alone, compared with 41% for arrests with warrants. Less than 2% of those with collateral arrests were convicted of a violent crime, one-third the rate of other arrests, and only 18% were convicted of any crime, compared with 33% for other arrests.

The collateral arrests contributed to an overall pattern of lower and lower shares of arrests for serious crimes, and more for immigration offenses alone.

Arrests climbed from about 12,000 in January 2025 to more than 40,000 in December, but fell back to 30,000 this February. The share of people with only immigration-related crimes and violations rose to more than half in December and January, the peak months for collateral arrests, and the share of violent criminals fell from 10% to 4% of arrests in that time.

New policy

ICE announced a new policy in January to issue warrants in real time if agents think an immigrant is deportable and “likely to escape,” though that policy faces a court challenge.

Total arrests and collateral arrests have been falling since December, whether because of the new policy or because of cutbacks in the large-scale street sweeps that tend to produce them.

One factor is public outrage over raids sweeping up noncriminals in places like Minneapolis and Chicago, said Colleen Putzel-Kavanaugh, an associate policy analyst for the nonpartisan Migration Policy Institute.

“The sort of large operations within big cities, as they were occurring, seems to have subsided somewhat,” Putzel-Kavanaugh said. “After the kind of public outcry following Minneapolis, it seems as though, at least for now, that tactic has kind of been paused.”

The Trump administration’s focus on mass deportation opened the way for more collateral street arrests with less investigation, she added.

“If it’s a more targeted arrest, they would take the time to sort of essentially have an investigation. It’s a pretty resource-intensive way that just would not yield the kind of numbers ICE was being told to produce,” she said.

The new policy was filed in court papers in February as a response to a lawsuit over ICE sweeps in the District of Columbia last year, alleging ICE agents “have flooded the streets of the nation’s capital, indiscriminately arresting without warrants and without probable cause District residents whom the agents perceive to be Latino.”

The case resulted in a preliminary injunction in December requiring a halt to warrantless arrests without establishing probable cause that the person is living here illegally and is a flight risk.

One plaintiff in the class-action case, José Escobar Molina, said in the lawsuit that agents in two cars pulled up to him as he approached his work truck on Aug. 21, grabbing him by the arms and legs and handcuffing him without asking any questions. Escobar, 47, said in the court papers that he’s lived in the district for 25 years and has had temporary protected status as a Salvadoran native the whole time. He was held overnight in Virginia before being released.

Other lawsuits are also challenging collateral arrests, such as an incident in Idaho in which agents with warrants for five people ended up arresting 105 immigrants at a Latino community event in October.

In North Carolina, four U.S. citizens and a visa holder sued in February, saying they were arrested in the Charlotte’s Web immigration crackdown in November without warrants, as is typical of collateral arrests.

“I have a lot of fear that this will happen to me again. I was essentially kidnapped based only on the color of my skin. That really weighs on me,” said Yoshi Cuenca Villamar, one of the citizens and a North Carolina native, in a statement announcing the lawsuit. He said he was doing landscaping work Nov. 15 when agents pushed him to the ground and handcuffed him, then held him in a car before releasing him.

One Illinois case that started in the first Trump administration challenged warrantless arrests and traffic stops used as a pretext for immigration arrests. A 2022 settlement required ICE to document “reasonable suspicion” of illegal status before arresting somebody. The case continues since a judge found in February that the new ICE policy of issuing warrants in real time after a detention violates the consent decree.

Shares of collateral arrests

In the months since August where collateral arrests are now labeled, the District of Columbia and Illinois stand out with high shares of collateral arrests. More than half the arrests in the district were collateral, as were 41% of those in Illinois. There were eight states in which at least 30% of arrests were collateral: Alabama, Maryland, West Virginia, Arizona, Pennsylvania, New Hampshire, Maine and Minnesota.

West Virginia, where there was a “statewide surge” of immigration enforcement in January with state and local cooperation, stands out for its high rate of total arrests as well as a large share of collateral arrests.

ICE labeled 1,300 arrests during Operation Metro Surge as ‘collateral’

For the eight months between August and early March, West Virginia had 1,831 arrests, or 1 in 10 of the state’s noncitizen population as of 2024, the latest data available. That’s by far the largest share in the country, followed by 7% in Wyoming (where truck drivers were targeted for immigration arrests in February) and 4% in Mississippi.

West Virginia Republican Gov. Patrick Morrisey, in a statement, cited the cooperation of state and local agencies with ICE through the 287(g) program that assists with immigration enforcement. He praised ICE, saying “they have removed dangerous illegal immigrants from our communities and made our state safer for families and law-abiding citizens.”

Few of those arrested in the surge were violent criminals, however. More than half of those arrested during the surge were collateral arrests, and only 1% — nine immigrants — had a violent crime conviction, according to the Stateline analysis. More than three-quarters, about 500 people, had only an immigration-related violation or crime.

Judges didn’t always agree that collateral arrests and detentions in the West Virginia surge were legal under the U.S. Constitution. U.S. District Judge Joseph Goodwin, a Clinton appointee, ordered two detainees released in January. He noted that “similar seizures and detentions are occurring frequently across the country” without any evidence they’re necessary as required by the Constitution.

Stateline reporter Tim Henderson can be reached at thenderson@stateline.org.