bilateral
TopicsAbout← Feed
BySRSam Reyes·CMCal Morrow·EQEliza Quinn·DPDana Park
BREAKINGApril 21, 2026

How Stephen Miller Is Adjusting Trump’s Immigration Agenda

Following fatal shootings of two U.S. citizens — Alex Pretti and Renee Good — during Operation Metro Surge, a massive ICE enforcement operation in Minneapolis-Saint Paul, Stephen Miller and the Trump administration are adjusting their deportation strategy. ICE has pulled back from Minnesota, overall arrest numbers have dropped, and the share of arrestees with criminal records has risen. However, critics and data show non-criminal arrests continue, and Miller is simultaneously advancing a broader immigration restriction agenda including pushing states to deny public education to undocumented children.

XLinkedInFacebookThreadsWhatsAppReddit
⚡Both sides have significant internal splits on this story. Arguments below represent the dominant positions on each side — see The Divide below for the full picture.

Stephen Miller is reshaping Trump's second-term immigration strategy—but what's actually changing, and does it represent a harder line or a strategic recalibration that could shift which restrictions actually get enforced?

Not familiar with this story? Get context →
Dragnet arrests versus targeted enforcement
Liberal
More than one in three arrests during Operation Metro Surge were 'collateral' — people swept up on the street with no individual predicate, no warrant, no criminal flag, just present in the wrong neighborhood. Fewer than one quarter of the 3,789 total arrestees had any criminal record at all. You can call that immigration enforcement. The Fourth Amendment calls it a dragnet.
Conservative
The 35% collateral figure is real, and it's a legitimate problem. But the enforcement-only-by-warrant model you're implicitly defending produced the 11-to-15 million undocumented population that exists today — decades of 'targeted' enforcement that was so selective it functioned as non-enforcement. The question isn't whether Metro Surge was perfectly calibrated. It's whether any interior enforcement at scale can avoid collateral contacts, and history suggests it cannot.
Liberal
You're describing a scale problem and treating it as an inevitability. The criminal-record arrest share rose from 56% to 65% after the operation collapsed — which means the 35% collateral rate wasn't operationally necessary, it was a choice. If the numbers improved when scrutiny increased, the original design was the problem.
Conservative
That improvement under pressure is exactly what a self-correcting system looks like — which means the oversight mechanism worked, not that the entire framework is illegitimate.
U.S. citizens killed, government account disputed
Liberal
Alex Pretti and Renee Good were U.S. citizens killed by federal agents during this operation, and video evidence reportedly contradicts the government's account of why. DHS shot 14 people in five months. When the state kills its own citizens and its first instinct is to dispute the footage rather than demand accountability, the burden of proof doesn't shift to the victims — it falls entirely on the government.
Conservative
Nobody is defending the deaths of Pretti and Good — those deaths, if the video contradicts the official account, demand full investigation and accountability. But you're using two tragic cases to indict an operation involving 2,000 agents over multiple weeks. Law enforcement operations at any scale produce incidents. The question is whether the rate of lethal force is systemically disproportionate or whether you're treating outliers as proof of design.
Liberal
Fourteen people shot in five months isn't an outlier rate — UN human rights experts called it potentially extrajudicial. And 'full investigation' is exactly what the administration is resisting. You can't cite accountability mechanisms the government is actively obstructing.
Conservative
UN characterizations of U.S. domestic law enforcement carry zero legal weight, and demanding investigation is not the same as obstructing one — those are different claims that need different evidence.
Plyler precedent and schooling for undocumented children
Liberal
Miller is now lobbying state legislators to deny public education to undocumented children — a move with a 42-year-old Supreme Court ruling directly against it. Plyler v. Doe wasn't a Warren Court liberal project. It was decided 8-1, and it held that punishing children for their parents' legal status violates equal protection. These children had no vote in the decision that brought them here.
Conservative
You cite Plyler as if it ends the conversation, but Plyler was a 5-4 decision — not 8-1 — and the dissent argued the Court had no business extending equal protection to people with no lawful presence. The current Court's equal protection jurisprudence has moved considerably since 1982. Miller isn't ignoring precedent; he's betting the Court that decided Dobbs reads Plyler's doctrinal foundations differently than the Court that wrote it.
Liberal
Even if you think Plyler is vulnerable, you're describing a strategy of deliberately passing unconstitutional state laws to lose in court and hope a future ruling reverses them — that's not governance, it's litigation theater that burns political capital while real enforcement reform goes undone.
Conservative
Every major constitutional shift in American law started as legislation someone called 'designed to lose' — that's how you build the vehicle for the case. Calling it theater only works if you're certain the precedent is unassailable, which the current Court's composition makes considerably less certain.
Strategic retreat disguised as refinement
Liberal
The rise from 56% to 65% criminal-record arrests is being called a refinement, but the Washington Post's own data shows non-criminal arrests continued at significant levels even after the administration publicly announced a strategic shift. The machinery didn't change — the press release did. What actually happened is the mass-arrest model collapsed politically and legally the moment U.S. citizens turned up dead, and Miller's response is to go quieter and relocate the pressure to state schoolhouses.
Conservative
You're calling a measurable improvement in targeting a 'press release' while simultaneously arguing the 35% collateral rate is disqualifying. You can't have it both ways. If the numbers moved in the direction you say enforcement should go, that's evidence the strategy responds to legal and political pressure — which is exactly how enforcement in a constitutional system is supposed to function.
Liberal
The strategy responds to pressure — agreed. Which means the original 35% collateral rate wasn't an operational necessity, it was the actual preference until resistance made it costly. That tells you what the default is when scrutiny relaxes.
Conservative
Every enforcement agency operates at the edge of political tolerance — the fact that this one moved when pushed doesn't reveal a sinister default, it reveals a negotiating position, which is how every administration on every issue has always operated.
Acceptable harm threshold for mass enforcement
Liberal
The administration has never answered the question that the facts require: what is the acceptable American body count for this policy? Fourteen people shot in five months, two of them U.S. citizens, a Hmong American man whose arrest Minnesota is investigating as a possible kidnapping — these aren't abstractions. If the answer is that some civilian casualties are operationally inevitable at the scale required to actually deter illegal entry, say that out loud and defend it.
Conservative
That question cuts both ways. What is the acceptable body count of Americans killed by people who should have been deported but weren't — because the prior administration's enforcement was so selective it functioned as policy non-implementation? The conversation about acceptable harm doesn't begin and end with enforcement operations. It begins with the original decision to allow an 11-million-person undocumented population to accumulate over thirty years of bipartisan inaction.
Liberal
Pointing to crimes committed by undocumented individuals doesn't answer the question — it changes the subject. The specific people killed by federal agents in Minneapolis were not the people you're describing. Alex Pretti and Renee Good died in an operation that couldn't keep track of who it was shooting. That's not a tradeoff. That's incompetence with lethal consequence.
Conservative
Incompetence requires evidence of the actual sequence of events — which the investigation hasn't completed. Calling it incompetence before the facts are in is exactly the same evidentiary move you're criticizing the administration for making in the other direction.
Conservative's hardest question
The strongest challenge to this argument is that the administration can credibly respond that any large-scale interior enforcement operation will generate collateral contacts and tragic incidents, and that the alternative — selective, warrant-based enforcement only — has been tried for decades and produced the 11-to-15 million undocumented population that exists today. The argument that targeted enforcement is both more just and more effective is theoretically sound but has not, in practice, produced enforcement at the scale that actually deters future illegal entry — and that empirical failure is genuinely difficult to dismiss.
Liberal's hardest question
The rise in the share of criminal-record arrests from 56% to 65% is a real number that partially supports the administration's claim that enforcement is becoming more targeted — a critic who concedes only the bad data and ignores the movement in that metric is not being fully honest. If that trend continued and non-criminal arrests fell substantially, the moral force of the 'indiscriminate dragnet' argument would weaken, even if the underlying policy remains objectionable.
The Divide
*Both sides are fracturing—conservatives on whether to pause for optics, progressives on whether enforcement can ever be legitimate.*
MAGA/POPULIST
Mass deportation must continue at scale and speed; collateral arrests and political costs are acceptable.
ESTABLISHMENT/PROCEDURALIST
Killings of U.S. citizens and high false-arrest rates create legal and political liabilities; retarget to criminal deportations only.
PROGRESSIVE/ABOLITIONIST
ICE enforcement is inherently violent and cannot be reformed; defund and abolish the agency.
INSTITUTIONAL DEMOCRAT
ICE must follow the law and due process, targeting only those with criminal records—the problem is lawless overreach.
The Verdict
Both sides agree
Both sides accept that Operation Metro Surge produced a dragnet effect with 35% collateral arrests and fewer than 25% of arrestees having criminal records — the disagreement is whether this represents indiscriminate overreach (liberal view) or a correctable operational mistake that the rising 65% criminal-record share now addresses (conservative view).
The real conflict
FACTUAL: Whether the rise from 56% to 65% criminal-record arrests represents genuine strategic refinement toward targeted enforcement or cosmetic adjustment masking continued indiscriminate operations — conservative cites the percentage rise as evidence of course correction; liberal cites Washington Post data showing non-criminal arrests continued even after the public pivot.
What nobody has answered
If the administration's own course-correction data (56% to 65% criminal-record arrests) is partly real, at what point does that trend line prove the system actually changed its underlying logic, and who gets to decide — or is the conservative rebuttal correct that you cannot distinguish between genuine reform and a system that operates indiscriminately until killed Americans force a pause?
Sources

More debates