Freedom within the Framework - Part II
I. The Assumption That Breaks
Part I of this series ended with an appeal to cooperation. The argument, built on game theory and the evidence from Camden, was essentially this: the relational contract between law enforcement and citizens is a repeated game, and in repeated games, cooperation is the rational strategy. Both sides have skin in the ongoing relationship. The shadow of the future disciplines behavior. Given the right incentives and enough time, trust is achievable.
That argument rests on a premise so fundamental it went unstated. It assumes that both parties to the contract are playing the same game.
What happens when they are not? What happens when a party enters a community with no relationship to it, no intention of staying, no accountability to the people it is policing, and no interest in what happens after it leaves? What happens when the force that shows up is not a local precinct operating under the scrutiny of the city council and the eyes of neighbors, but a deployment of thousands of agents sent from out of state, operating under rules written somewhere else, answering to a federal authority that the community has no mechanism to reach?
The answer, as Minneapolis learned in the winter of 2026, is that the relational contract does not just break down. It becomes structurally impossible. And the consequences of that impossibility extend far beyond the duration of the operation itself.
II. Operation Metro Surge
Operation Metro Surge was a federal immigration enforcement operation conducted by ICE and Customs and Border Protection with the stated purpose of apprehending undocumented immigrants. Beginning in December 2025, it initially targeted the Twin Cities and later expanded to all of Minnesota. The Department of Homeland Security called it “the largest immigration enforcement operation ever carried out,” involving the detention of roughly 3,000 people.
The operation’s publicly stated rationale was immigration enforcement. Its methods became a different story. Since December, immigration enforcement agents shot three people and killed two; racially profiled people, asking them to produce proof of legal residency; detained legal immigrants and shipped them across state lines; caused numerous car crashes; deployed chemical irritants on public school property; smashed the car windows of observers and arrested them before releasing them without charges.
On January 28, 2026, Minnesota’s chief federal district judge found that ICE had violated at least 96 court orders in Minnesota since January 1. On February 3, another federal judge stated that the “overwhelming majority” of cases brought before him by ICE involved people lawfully present in the United States.
Into this environment came two deaths that forced the country to confront a question Part I could only gesture at: what do you do when the law itself becomes the source of the threat?
III. Renee Good and Alex Pretti
Renée Nicole Good was a 37-year-old US citizen, a writer and poet who lived in Minneapolis with her partner and a six-year-old child. Good’s killing was the ninth time in five states and Washington, D.C., that ICE agents had opened fire on people since September 2025. At least four other people had died during federal deportation operations since the enforcement surge began.
On January 24, 2026, Alex Jeffrey Pretti, a 37-year-old American intensive care nurse for the United States Department of Veterans Affairs, was shot multiple times and killed by two United States Customs and Border Protection officers in Minneapolis. Pretti was filming law enforcement agents with his phone and directing traffic. At one point, he stood between an agent and a woman the agent had pushed to the ground, putting his arm around her. He was then pepper sprayed and wrestled to the ground by several federal agents, with around six surrounding him when he was shot and killed.
Bystander video was verified and reviewed by Reuters, the BBC, The Wall Street Journal, and numerous other organizations. Numerous news organizations noted that multiple specific claims made by Trump administration officials about the killing were contradicted by the video footage.
While ICE officers are not required to wear body cameras, and only approximately 4,400 of 22,000 ICE officers had even been issued one by June 2025, DHS indicated body camera footage from at least four different angles of the incident existed and were under review.
Then came the official response. At a news conference, Border Patrol Commander-at-Large Gregory Bovino referred to Pretti and Good as “suspects” when asked whether federal leadership planned to take any accountability for the two deaths. “You’re correct, two suspects have been shot,” Bovino said.
Kristi Noem’s insistence that the victims in both cases were “domestic terrorists” made a credible federal investigation politically impossible from the start.
Minnesota prosecutors requested a partnership with the federal government to investigate the Good case, as they had done in past shootings involving federal agents. When the Trump administration refused to cooperate, Minnesota prosecutors escalated, sending a series of strongly worded legal letters demanding evidence. More than a month after the deadline set by prosecutors, the Trump administration still had not turned over the materials.
The state of Minnesota, Hennepin County, and the Minnesota Bureau of Criminal Apprehension ultimately filed suit against the Department of Justice and the Department of Homeland Security, alleging that the defendants were withholding investigative evidence about the killings in order to shield the federal officers involved.
Two American citizens, dead on camera, in a city that had no say in the operation that killed them, being called suspects by the agency responsible, while the evidence needed to investigate was withheld from the state with jurisdiction. This is not a breakdown of the relational contract. It is the deliberate dismantling of it.
IV. The Game Theory Failure
Return to Part I’s core framework. Robert Axelrod’s central insight was that cooperation emerges in repeated games. When players know they will encounter each other again, defection becomes costlier because tomorrow’s interaction carries the consequence of today’s choice. The “shadow of the future” is what makes the contract possible.
Now consider the structural position of a federal agent deployed to Minneapolis from another state for a defined operation. That agent will not patrol these neighborhoods next year. Will not attend the community meeting next month. Will not see the family of the person they arrested, or do not arrest, at the grocery store on Saturday. The relationship is, by design, a one-shot game.
In the classic Prisoner’s Dilemma, a one-shot game has a clear dominant strategy: defect. The logic is airtight. If you will never interact with this person again, there is no future cooperation to protect, no reputation to maintain with this specific community, no repeated-game incentive to build trust. The individually rational choice is always to maximize the immediate objective, which in this case meant arrests, not relationships.
This is not a critique of any individual agent’s character. It is a critique of the structure. The incentives were configured to produce defection as the rational outcome, and they did.
But Operation Metro Surge was not simply a one-shot game played by strangers. It was something structurally worse: an external party introducing a one-shot game into a community that had been, painstakingly, trying to build the repeated one. The local Minneapolis Police Department, whatever its own fraught history, exists in an ongoing relationship with its residents. Officers live in or near the communities they serve. They attend community meetings. They are subject to local political accountability. They have, at minimum, the structural conditions that make a relational contract possible.
When thousands of federal agents arrive, operating under different rules, answerable to a different authority, with no accountability to local governance and no stake in what the community looks like after they leave, that local relational contract does not simply pause. It gets contaminated. Residents who might have distinguished between the Minneapolis Police Department and ICE find the distinction increasingly difficult to maintain when the two appear to operate in the same space. Fear, once ignited, does not sort itself neatly by agency affiliation.
Car horns and whistles became a constant in parts of the Twin Cities, alerting immigrants to lock their doors and citizens to come out and film when ICE was active. The infrastructure of community cooperation, the thing that takes years to build, collapsed into an improvised warning system in weeks.
V. The Collateral Damage
The scale of disruption in Minneapolis was not incidental. It was measurable.
In one month alone, the city experienced at least $203.1 million in economic impact, representing losses to the economy, community livelihoods, neighbors’ mental health, and to food and shelter security. That figure included $47 million in lost wages from residents unable to safely go to work, $81 million in restaurant and small business revenue losses, and $15.7 million in additional rent assistance needed due to lost income. Approximately 76,000 people, mostly immigrants, refugees, and people of color, required urgent relief assistance.
Community leaders described receiving hundreds of calls daily from people reporting ICE agents at their businesses and in their neighborhoods. Business owners reported federal agents entering restaurants pretending to be customers, then returning later to arrest employees or the owners themselves.
Schools closed. Clinics reported patients avoiding care. The basic civic infrastructure of a functioning city, the trust that allows people to show up, to participate, to be visible, collapsed.
None of this chaos, it should be noted, was necessarily the intended goal. The stated goal was immigration enforcement. But the structural conditions of the operation, agents with no relationship to the community, no requirement to wear body cameras, no accountability to local governance, and no repeated-game stake in what comes next, made this level of community damage not a side effect but a near-certain outcome.
An NPR/PBS/Marist poll conducted in February found that two thirds of Americans said ICE had gone too far. That number is significant not because polls determine policy, but because it suggests that even among Americans who support immigration enforcement as a principle, the specific execution of Operation Metro Surge registered as a violation of something fundamental.
What was violated was not just a legal standard. It was the intuitive expectation, held by most Americans regardless of politics, that law enforcement operates within a recognizable relationship with the public it serves. When that expectation is violated at scale, the damage spreads well beyond the immediate community.
VI. The Accountability Gap
Part I argued that the single most important driver of the trust deficit between law enforcement and citizens is a systems design problem: we measure the wrong things, reward the wrong behaviors, and then act surprised when we get the wrong outcomes.
Operation Metro Surge compressed that failure into a matter of weeks, and added a new dimension that the standard reform literature has not fully grappled with: what happens when accountability mechanisms themselves are removed from the equation?
Local police departments, whatever their flaws, operate within a web of accountability structures. City councils, police commissions, local prosecutors, community boards, civil litigation in local courts. These mechanisms are imperfect and often inadequate. But they exist, and their existence creates at least a structural possibility of course correction.
Federal agents involved in the shootings of Good and Pretti did not speak publicly. Minnesota prosecutors found themselves denied access to crime scenes, kicked off joint investigations, and unable to obtain basic evidentiary cooperation from the federal government. The Minnesota Bureau of Criminal Apprehension, the agency that would normally investigate police shootings, was effectively stymied.
The Justice Department indicated it was opening a federal civil rights investigation into Pretti’s killing, but said a similar federal probe was not warranted in the killing of Good.
When the agents responsible for a civilian death answer only to the federal authority that deployed them, and that authority declines to cooperate with independent investigation, the entire accountability loop closes. There is no external check. There is no mechanism by which the community’s judgment about what happened can translate into institutional consequence. The game becomes, by design, unaccountable.
This is the powder keg. Not the violence itself, as horrific as it was. The powder keg is the combination of traumatic events and the absence of any credible path toward accountability. Communities can absorb difficult truths when they believe the systems around them are trying to find those truths. What they cannot absorb, what history shows produces the most severe and durable breakdowns in civic trust, is the experience of watching something happen in plain sight and being told, in effect, that it did not happen, or that the people it happened to deserved it.
The victims were called suspects. The evidence was withheld. The investigation was blocked. The operation was defended.
VII. The Structural Problem Has a Name
What Minneapolis experienced was not a policing failure in the traditional sense. It was the consequence of deploying law enforcement in a context that structurally prevents the conditions for legitimate authority from forming.
Legitimate authority, as political scientist Tom Tyler’s research has repeatedly demonstrated, depends not just on outcomes but on procedural justice: the perception that the process was fair, that people were treated with dignity, that there was an opportunity to be heard. When those conditions are absent, even outcomes that might otherwise be accepted as legitimate are rejected.
Federal agents operating under masks, without body cameras, without local accountability, answering factual disputes about on-camera shootings with press releases that contradicted the video record, could not produce procedural legitimacy by definition. The process was not perceived as fair because in several documented respects it was not fair.
The deeper structural problem is what happens to local law enforcement when federal operations behave this way in the same geography. When local police work with ICE, it makes it harder for the community to be aware of immigration enforcement happening near them. In Texas, where local police cooperation with federal authorities has been long-standing, a whistle network like Minneapolis’s was never possible, because enforcement was never visible enough to track. The visibility of Operation Metro Surge was unusual. What was not unusual was the underlying dynamic: federal priorities being executed through, alongside, or in spite of local law enforcement relationships that took years to build.
The question for any city, and for any local law enforcement agency, is not whether federal operations can be controlled. They generally cannot. The question is whether the relational contract with residents can survive them, and what it takes to rebuild it when it does not.
VIII. What the Game Requires
Part I ended with an appeal to cooperation as the only rational long-run strategy. That conclusion stands. But it requires a clarification that Minneapolis made impossible to avoid.
Cooperation is the rational strategy in a repeated game between parties who share a stake in the ongoing relationship. It is not the rational strategy when one party is playing a finite game with no future stake, no accountability to the other party, and no cost for defection beyond the duration of the operation.
The path forward, then, is not simply a matter of reforming the incentives within a single institution. It requires something more structurally demanding: ensuring that any law enforcement authority operating within a community is accountable to that community in some meaningful and enforceable way. That means body cameras as a non-negotiable baseline. It means independent investigative authority that cannot be blocked by the agency under investigation. It means metrics that measure something other than arrest counts. And it means recognizing that sending thousands of agents from out of state into a city that has not consented to their presence, operating under rules that immunize them from local accountability, is not simply a different approach to public safety. It is the structural opposite of one.
The family of Renee Good released a statement after the killing of Alex Pretti: “ICE agents can leave Minneapolis. The residents of Minnesota cannot.”
That sentence contains the whole argument. Accountability, cooperation, and trust require permanence. They require the knowledge that the person holding power over you will still be there tomorrow, will see you at the school pickup, will answer to the same city council you vote for. When that condition is absent, the relational contract is not strained. It is impossible.
What Minneapolis showed is that the question of freedom within the framework is not just a question for local precincts and community boards. It is a question that runs all the way up to the structure of federal power and the conditions under which that power can legitimately operate among citizens who did not choose it, cannot recall it, and have no means of holding it to account.
That is the question this moment is forcing us to answer. And how we answer it will shape the relationship between law enforcement and citizens for a generation.