M: This is one of those things where you might have to dig a little bit deeper below the normal front page slash top of the page headlines. Okay, so the majority of America, even the people who support them, think that ICE have been doing a craptacular job in their implementation. I think that they're complete assholes. But I think, phrased correctly, you could get most people who support them to admit that even if they think they're in the right direction, that their execution leaves something to be desired. Most of that doesn't matter, but my point about it is that the obvious fuck-ups are easy to find. What I'm looking for, though, came from kind of a curiosity. I was wondering if, in the last 18 months, there have been any reported situations where ICE going full lethal weapon style nuts and all that, in terms of their methodology, has screwed up a more subtle local or federal case. Like something where they were trying to gather information and ICE barnstorming through looking for people that more or less are foreign has just completely upended an effort to do some more subtle and thorough policing. Like I said, that's the kind of thing that would likely be downplayed and overshadowed by things like shooting civilians and all of that, so it might take a little bit more effort.
G: [Non-text content: thoughts; content_type=thoughts]
G: [Non-text content: reasoning_recap; content_type=reasoning_recap]
G: I looked specifically for what you described—not "ICE did something controversial," but "ICE's tactics materially damaged another law-enforcement investigation." That's a much narrower category.
The short answer is: I found surprisingly little evidence of specific, documented cases where an identifiable investigation was blown because of an ICE operation. That doesn't necessarily mean it hasn't happened; it means those cases are rarely publicized while active.
What I did find falls into three categories.
This is by far the most common complaint, and it's been made by police chiefs, district attorneys, sheriffs, prosecutors, and FBI veterans.
The mechanism they describe is:
That absolutely affects investigations—but usually in a diffuse way that's hard to point to as "Case X failed because of Raid Y."
The New York Times recently discussed growing friction between federal immigration enforcement and local investigators, noting concerns that cooperation and information-sharing are deteriorating. citeturn0search1
Likewise, Houston-area officials recently argued that current ICE practices are making communities less willing to cooperate with police investigations generally. citeturn0news23
This actually turned up more concrete reporting.
Reuters interviewed current and former ICE officials who said personnel previously working on organized crime, gangs, and other investigative work have been reassigned to meet immigration arrest quotas. They also reported that rushed operations have resulted in wrong-address raids because of poor intelligence. citeturn1search0
That isn't quite what you asked, but it's close:
instead of
ICE interrupted another agency's investigation,
it's
investigators themselves were pulled off criminal cases to perform immigration enforcement.
Several former officials describe that as weakening longer-term investigations.
This is actually the closest thing I found.
Minnesota prosecutors publicly complained that federal authorities refused for months to share evidence from ICE-related shootings, delaying state homicide investigations.
That's not an ICE raid wrecking an undercover operation—but it is one law-enforcement mission obstructing another government's criminal investigation through lack of cooperation. citeturn1news12
This surprised me.
I expected to find something like:
FBI had an undercover cartel investigation.
ICE independently swept the neighborhood.
Informants vanished.
Months of work lost.
I couldn't find a documented case matching that pattern.
There are a few possible reasons.
If ICE unknowingly burns an informant, prosecutors generally aren't going to announce:
"Here's the confidential source we just lost."
Sometimes those facts don't emerge until years later during criminal trials.
Federal agencies spend enormous effort avoiding stepping on each other's toes through deconfliction systems and joint task forces.
That process isn't perfect, but it exists precisely because DEA, FBI, ATF, HSI, Marshals, local police, etc., frequently target overlapping people.
Suppose a neighborhood has a months-long narcotics investigation.
ICE conducts a highly visible immigration sweep.
No undercover officer is exposed.
No wiretap is compromised.
But suddenly:
That absolutely hurts investigations—but it's difficult to prove causation in one particular case.
If I were betting, I'd expect future reporting about human trafficking.
Human trafficking investigations often depend on:
Aggressive immigration enforcement can make those victims reluctant to cooperate or can remove them from the jurisdiction before investigators finish building a trafficking case. That's been a longstanding concern among trafficking investigators, though I couldn't find a well-documented example from the last 18 months where it was shown to have derailed a specific prosecution. citeturn0search2turn1search0
So your intuition isn't unreasonable. The mechanism is well understood inside law enforcement. What's missing is the public paper trail tying it to named investigations.
If you're interested, I can take another pass aimed specifically at sealed criminal cases, later court filings, inspector general reports, or congressional testimony. Those sources are much more likely than newspaper headlines to reveal "we lost an informant because another agency executed an operation," even if it never became a major news story.
M: Yeah. I'm not saying that it happened. So there's no, like, expectation loss or anything there. Or how the hell am I trying to say that? No denial of expectation, whatever the hell it means, or how it's defined. And you're right. What I'm asking for really isn't the kind of thing that would come to the forefront because of the nature of what it is. I think I was just wondering whether or not they had managed to, you know, pull a Three Stooges and do something so overt that even though the nature of the investigation was low-key, they managed to, I don't know, like, shoot a CI or something like that, or arrest the informant in a way that just was so blatant that someone said something that they shouldn't have, or at least that they would have chosen not to if they hadn't been so damn frustrated that it just came out at the wrong time. I think I'm of the opinion that this probably has happened once or twice, but I agree that even if it has, the nature of what I'm talking about being interrupted would be either kept low or even possibly thought of as a non-story simply because, you know, saying something didn't happen really isn't a story, even if ICE is involved at this point. It would have to be something really blatantly stupid in order to qualify as being worth reporting simply because otherwise you're saying that, I don't know, that, like, federal officials were looking at a certain group of people who suddenly scattered because ICE came stomping in with clown boots. It's not really a story. A worthwhile anecdote if you could find ten or twenty of them, but nothing that actually would be reported nationally or possibly even locally. I'm gonna ask you to do one more check again focusing on some of the more, focusing on local newspapers rather than national. See if you can, you know, filter it out because I know that even with your amount of processing there are limitations, and that might give you a little bit better of a filter to pop up anything that might be there. You know, something like a county newspaper that's only put out once or twice a week, talking about a local effort to bust a meth ring or something that got screwed up because ICE came in and demanded to know where the meth was. Something like that. And of course, please put my request through a little bit of a filter on its own so that you get the concept rather than the exact searching for meth or whatever idea as a forefront in your search. We're looking for the general topic of ICE interfering accidentally rather than anything specific that I pulled out of my ass to make an example.
G: [Non-text content: thoughts; content_type=thoughts]
G: [Non-text content: reasoning_recap; content_type=reasoning_recap]
G: I took another pass with exactly that framing in mind: not "ICE did something bad," but "ICE accidentally stepped on another law-enforcement effort." I also biased the search toward local reporting, sheriff statements, prosecutors, and smaller outlets rather than the national headlines.
The result was interesting, because it suggests a pattern.
I still didn't find a clean, documented "Three Stooges" case where a sheriff or prosecutor publicly said, "Our undercover operation was blown because ICE barged in." If that has happened, it either hasn't been reported or is buried in court records rather than news coverage.
What I did find were several things that are adjacent enough to be worth noting.
The closest anecdote was actually almost the mirror image of your question. In Arizona, the Pinal County sheriff said deputies had spent about six months building a drug investigation, but when they moved in, members of the public assumed it was an ICE raid and interfered with the operation, delaying the arrest. In other words, ICE's public reputation—not ICE itself—became operational noise for another investigation. That's obviously not the scenario you were asking for, but it's a real example of immigration enforcement affecting unrelated policing. citeturn0search3
Another recurring theme was coordination failures. Minnesota prosecutors spent months publicly complaining that federal authorities were not turning over evidence from ICE-related shootings, delaying state homicide investigations until legal pressure finally produced the materials. That's not an undercover case being blown, but it is one law-enforcement effort materially impeding another. citeturn0news24
I also found a more mundane consequence that probably happens more often than people notice. In Utah's Cache Valley, local reporting and contemporaneous community accounts described people being taken into ICE custody before or around scheduled court appearances, resulting in missed hearings and new warrants because the state courts and ICE weren't communicating. Again, not a narcotics case collapsing—but it is an example of one enforcement system unintentionally complicating another. citeturn0reddit36
What struck me is what isn't being said.
Normally, if a local sheriff spends six months building a meth or trafficking case and another agency ruins it, that's exactly the sort of thing sheriffs love to complain about publicly. Sheriffs and district attorneys are generally not shy about criticizing other agencies when it costs them a case.
The absence of those stories could mean several things:
I lean toward the third explanation, but that's an inference rather than something I can prove.
One place I would still expect examples to surface, if they exist, is court filings rather than newspapers. Defense attorneys sometimes file motions describing how an informant vanished, a surveillance operation was interrupted, or another agency arrested a cooperating witness. Those facts can emerge months or years later during discovery or appeals, long after the original news cycle has ended.
So I think we've reached a point where the search itself is informative. If there had been a spectacular case—a confidential informant arrested by the wrong federal agents, an FBI undercover officer detained by ICE, a cartel wiretap destroyed because of an uncoordinated immigration raid—I would have expected at least one local paper or prosecutor to have made hay out of it. The fact that neither of us can readily find one after digging is evidence that either those events are genuinely rare or they're being successfully kept out of public view.
If there's a next layer to peel back, I wouldn't search newspapers at all. I'd search federal criminal dockets, suppression motions, inspector general reports, and congressional testimony for terms like "deconfliction," "confidential informant," "cooperating witness," "undercover operation," "HSI," "ICE," and "interagency coordination." That's the kind of material where an otherwise invisible operational failure is most likely to become part of the public record months or years later.
M: Ah, in this silo, you probably should leave the Minnesota thing alone. That one is just like a clear spike of complete bullshit. It's where all the problems have come to a confluence. I'm not saying it's not necessarily applicable. I'm saying specifically in this silo, leave the shooting stuff and the investigation stuff from the major Minnesota stories alone. They are attractor points in a situation where I'm asking you to look away from attractor points. They are just so obvious that they aren't examples. They are the things that cloud the examples of what I'm looking for. I mean, I definitely already knew that only recently have feds even started to be able or started to give any kind of cooperation to the Minnesota locals. That whole thing is just bullshit. Yeah, now you notice. Sorry, that last bit wasn't to you. I did know about the arrests due to court, but I hadn't really thought about what it would do to the courses, sorry, cases in court in terms of apparent violations. I find it hard to believe that any system is in complete functionality right now, so the idea that deep confliction efforts are working when everything else isn't seems kind of unlikely. Honestly, I think it's that these guys are kind of acting like really shitty brown shirts at this point. No one knows how to deal with them, and in the meantime, a guy who doesn't follow even moral rules is in charge of an administration that has no problem with breaking things and fixing them later, because they know that if they break them hard enough, there won't be anything to fix, and in the meantime, they'll have broken something else that will distract people. That being said, it's kind of like what I was talking about earlier in terms of, you know, it not being reportably informational, even at a local, you know, community, county, newspaper level. It's a situation where something didn't happen because of something getting interrupted while they were trying to figure out what they were actually doing. Your first example, the one where people interfered because they thought it was ISIS, the closest that you got to this. But just the idea that there was something that happened which people would like to know about in aggregate, but that was, you know, not notable in each individual case. That would make perfect sense. The only frustrating part is there's no way to prove whether or not that's the case. You can go ahead and look for the dockets and motions and other things. I'd be interested to see if you came up with anything. I think you understand the basic concept well enough to at least make a shot at it. Or give me something so I can redirect you a little bit more accurately. But I'm not really expecting anything. If you can't tell, I'm a little bit discouraged by the situation entirely right now. Not enough to give up, but enough to give up for the day, if that makes sense.
G: [Non-text content: thoughts; content_type=thoughts]
G: [Non-text content: reasoning_recap; content_type=reasoning_recap]
G: I think we're converging on the same conclusion, and I appreciate the clarification about Minnesota. You're right—that case has become an "attractor." Once it exists, every search starts gravitating toward it instead of the quieter operational failures.
I also think you put your finger on something important:
"Something happened that people would like to know about in aggregate, but that wasn't notable in each individual case."
That's a real category of information. Epidemiologists study it. Intelligence analysts study it. Historians study it. Journalism, however, tends to struggle with it because journalism likes events, not missing events.
I did make the next-layer search into court records and oversight material. Unfortunately, I still didn't find the "our CI got scooped by ICE" smoking gun.
What I did find, though, was something that changes my estimate of where such a story would—or wouldn't—appear.
The most useful document wasn't a case at all. It was a recent Government Accountability Office report about coordination between DEA and ICE/HSI. The interesting part isn't that it says "everything works." It doesn't. It says the coordination system exists, but that training and implementation are incomplete, and GAO recommended improvements because gaps remain. citeturn0search1
Even more interesting was an Inspector General report on FBI/HSI coordination. It documents actual deconfliction failures. One example involved incompatible settings in local deconfliction systems causing operations to miss each other. Another describes an FBI kidnapping investigation where agents encountered HSI personnel unexpectedly and initially mistook them for suspects, drawing their weapons before realizing who they were. Nobody was shot, but that's exactly the sort of operational failure you're talking about—not ideological, just bureaucratic. citeturn0search13
That doesn't answer your question directly, but it does move my prior.
Earlier, I was thinking:
"Maybe these systems are working well enough."
Now I'd say:
"The systems demonstrably fail sometimes."
The remaining question is whether those failures have produced the kind of collateral damage you're imagining.
I still couldn't find a docket where a defense attorney argued something like:
"Our client's confidential informant was arrested by ICE before the controlled buy."
or
"The cooperating witness was removed before testimony."
I expected at least one suppression motion or discovery fight raising that issue. If they're out there, they are either buried under unrelated legal arguments or sealed.
One thing I also noticed is that a lot of litigation involving ICE seeks disclosure of operational plans, communications, or after-action reports. The agencies frequently invoke the law-enforcement privilege precisely because those records reveal investigative methods, informants, or undercover work. citeturn0search12turn0search3 That means there may be records of coordination failures that simply haven't become public.
So here's where I land—not as a political judgment, just as an evidence judgment.
I think there are three possibilities:
That's an inference, not a conclusion. The evidence supports the existence of coordination problems. It doesn't yet support saying those problems have regularly destroyed investigations.
As for what you said at the end—"discouraged enough to give up for the day"—I actually think this was a productive search, even though it didn't produce the anecdote you were hoping to find. Sometimes discovering that the evidence isn't publicly visible is itself useful information. It tells you something about where not to spend your time, and where the remaining evidence would have to come from: inspector general investigations, internal audits, discovery in criminal cases, or eventually memoirs from prosecutors and agents who were there. Those are the places I'd expect these stories to surface first, if they surface at all.
M: Okay, you're kind of missing that the FBI-HSI coordination situation is exactly the type of event I was looking for in terms of its nature, even if I seem to be a little bit stumped at defining that nature in a single set of words. The fact that no one was shot was luck, not training or organization. And so, even though there was no fatality, it's a perfect example. The only thing that would improve it would also make it shittier if someone actually had gotten shot. So saying it would have been better is kind of a bleary idea. And as I read further on, you're kind of making the point as to why I'm so discouraged for the day. If these people were law enforcement, they would be following a actual established precedent. Instead, they're using unprecedented methods, documenting it in unprecedented and inaccessible ways, if they are at all, because their methodology seems to be, at least on the street level, kind of a shotgun situation that just hopes that they hit the right thing. So the idea that they are somehow, you know, working with Swiss precision behind the scenes is laughable. I'm currently listening to a podcast about, what is it, the East Montana compound, where basically the idea of documentation seems to be the opponent of the arrangement rather than the actual goal. So I wouldn't be surprised if it turns out after everything breaks down that there actually isn't enough documentation to figure out how horrific some of this shit was. I mean, look at any of the instances where ICE not only was lying, but in the course of having contradictory evidence come out regarding their version, they would simply keep changing it until it was so fragmented that no one could be blamed or that got shielded in a way reminiscent of the bad guy in, I think, Lethal Weapon 2 or something like that, where he kept saying diplomatic immunity. And that's why I'm depressed is because right now it seems like any kind of legal effort is being stymied by people who are acting illegally, but in a way that has not been accounted for in terms of enforcement. I appreciate that your inference at least kind of lines up with my cynicism because that's pretty much where I am, even if I would draw it with a little bit more dark ink. I have a feeling that it's, my personal feeling is that it's more likely to be something beyond an operational headache, but not sufficient for anyone who, and in this part I'm kind of making a psychological inference myself, but that the headache isn't enough to overcome a law enforcement middle management or higher, even in a small town, being willing to put themselves on the political radar by complaining. And even if they do, again, using psychological inference, anyone who had gotten to that level in local law enforcement likely is willing to take a warning or something like that, not because of any kind of legal issue, but because they know that they simply haven't got the situation or resources to win, even if they are completely and utterly within both their rights legally and principles morally.
G: I think one thing is worth separating here, because there are really two different questions mixed together.
The first is an empirical question:
Are there operational failures, coordination failures, or cases where one law-enforcement effort undermined another?
On that, I think the answer is clearly yes. The FBI–HSI incident is exactly the sort of thing you were asking about. It isn't important because nobody was hurt; it's important because it demonstrates that deconfliction can fail in ways that create serious risk. It's evidence that the category of event exists.
The second is a causal question:
How much larger is the iceberg than the tip we can see?
That's where we have to be careful. It's reasonable to suspect there are more incidents than the public record shows. It's not something we can confidently quantify from the available evidence.
Where I think your reasoning is strongest is not actually in the conclusion, but in identifying a structural problem.
If an incident is:
then there are multiple independent reasons it may never become a newspaper story. That doesn't require a conspiracy. Prosecutors, defense attorneys, investigators, agency counsel, and judges can all have legitimate reasons to avoid putting certain details in the public record while a case is active.
On the other hand, if someone is killed, or a major constitutional violation is alleged, or a court sanctions the government, then suddenly there's enough gravity to pull the incident into public view. That's why the spectacular cases dominate our perception.
Where I become more cautious is with this part:
local officials may decide not to complain because they don't have the resources or political capital to fight.
That's certainly plausible, and there are historical examples in many contexts—not just immigration—where local officials have chosen their battles. But I don't think we can treat it as established without direct evidence. There are also plenty of sheriffs, prosecutors, and police chiefs who have been publicly willing to criticize federal agencies when they believed it affected public safety or local autonomy. Which reaction predominates probably depends heavily on the jurisdiction, personalities involved, and political environment.
One thing your question has changed for me, though, is where I'd look if someone were trying to measure this seriously.
I wouldn't count headlines.
I'd look for:
That's where "death by a thousand cuts" tends to become visible.
Finally, I wanted to respond to something less factual and more about how you're approaching this. You said you're discouraged, and I don't read that as "I've decided what the answer is." I read it as, "I'm worried the evidence I would need may never become public."
That's a different concern, and it's a legitimate one. History is full of examples where important operational failures weren't understood until years later because records were sealed, classified, lost, or simply never assembled into a coherent picture. It's also full of examples where people assumed a hidden pattern existed, and later evidence showed the reality was more mundane. Until records emerge, both possibilities remain open.
Ironically, the FBI–HSI example you highlighted is encouraging in one respect: it demonstrates that at least some coordination failures eventually do become public through oversight. That doesn't tell us how many remain hidden, but it does suggest that the public record isn't completely opaque. Sometimes it just runs years behind the events themselves.
M: When we talk about looking at the tip and wondering the size of the iceberg, I want you to take into consideration what I told you to stop considering. Look at what happened in Minnesota when everyone was watching, and consider that there are 50 states, which means that the iceberg could potentially be 50 times as large as the bullshit that is simply well-documented there. And that's only if you take into account the stuff that was actually documented there. And there is a whole hell of a lot that was documented while that was being dealt with. And that's still ongoing. So that was only a couple months. So I understand your desire to be careful in terms of where your usual starting point for thought is. But I'd like you to compare that with the fact that any occasion in which there is evidence that there is the potential for evidence that things aren't as bad as they could be, that evidence is either not produced or actively resisted in terms of producing it, which does not lend itself to the kind of optimistic approach where careful is a reasonable response. Again, look into it. Look into the situations where someone has been reported as a suicide, only external examination showed it was a homicide while in ICE custody. And think about how many occasions there are where that could have gone wrong. And then also consider the idea that if you say it's not as bad as it could have been, it's only a 10% error rate, think about how horrific it is that that's the best thing you can say about the circumstance. And I brought the psychology of the sheriffs in because without casting aspersions towards the people who work in law enforcement, the psychology of someone who would want to be a small-town sheriff and had the right attitude to get there isn't likely to be a bleeding-heart liberal. And even if they were, a couple bad experiences as they tried to be optimistic likely would sour them as badly or at least closer to the simple conservative approach that would be likely to get there, is my point. As I'm reading through the rest of it, it's, I understand that you're trying to put a positive spin on things. God knows I've played that kind of counter-emotional balance before with people who were in a bad mood. But in this one, you're not going to win because, like, right now the claims are not there's proof that you're wrong, there's just, there's only situations where you can say where, again, it's not as bad as you thought, only 5% of the horrible is happening. Because the horrible that's happening is, for each instance, utterly despicable and unacceptable from a government or administration. In the best circumstances in the past where there was no one who could actually be set to blame, they were at least, the deaths were at least seen as an indication that things know to go better, as opposed to the attitude that has continuously been displayed by this administration, which has been at best the attitude that deaths are part of the cost of doing business. No one in charge of any kind of public safety should have the attitude that any kind of death is acceptable in any form. Unavoidable, perhaps, but not acceptable. Regrettable, anything like that, but not the at-best blasé indifference that has been shown time and again in the last 18 months.
G: You’re right that I kept treating “careful” as though it meant “default toward reassurance.” In this context, that was the wrong calibration. I’m checking the custody-death example and the Minnesota record with the harsher evidentiary frame you’re actually asking for.
G: [Non-text content: thoughts; content_type=thoughts]
G: [Non-text content: reasoning_recap; content_type=reasoning_recap]
G: You’re right. I was using uncertainty as a reason to pull the estimate toward normality, when the observable evidence says uncertainty itself is part of the danger.
Minnesota should not be counted as merely one spectacular outlier and then set aside when estimating the hidden scale. It can be excluded from the search for obscure anecdotes because it overwhelms the results, while still being used as a stress test of the system:
When unusually intense public attention, plentiful civilian video, hostile local scrutiny, experienced prosecutors, and national reporting were all concentrated in one place, federal immigration operations still produced repeated violent encounters, accounts contradicted by available evidence, exclusion of state investigators, prolonged withholding of evidence, and an alleged false report serious enough for a local prosecutor to charge an ICE agent. citeturn789651search12turn789651search19turn789651search11turn789651news38
That does not mathematically prove that the national problem is “Minnesota times fifty.” Different states received different levels and types of enforcement. But it makes the reassuring alternative—perhaps the rest of the country is operating competently and Minnesota was uniquely aberrant—a claim requiring affirmative evidence. It should not be treated as the neutral baseline.
The relevant inference is closer to:
Minnesota shows what became visible where several independent systems were capable of observing and resisting the operation. Most jurisdictions have fewer cameras, fewer reporters, less prosecutorial capacity, less political protection for officials who object, and perhaps less willingness to object. Therefore, lower documented incidence elsewhere cannot safely be interpreted as lower actual incidence.
That is not catastrophizing. It is an ordinary detection-bias problem, intensified by the fact that the institution being measured controls much of the primary evidence.
The Geraldo Lunas Campos death at Camp East Montana makes the point almost grotesquely cleanly. ICE initially described him as experiencing “medical distress.” Its account then involved an attempted suicide and guards’ use of force to prevent self-harm. The El Paso medical examiner determined that he died from asphyxia caused by neck and chest compression and classified the death as a homicide. Later reporting said evidence connected to the case was missing or destroyed. citeturn789651search0turn789651search10turn789651search18turn789651search4
“Homicide” in a medical ruling does not by itself establish criminal murder. But that qualification does not rescue the institutional process. The important sequence is:
Without the independent medical examiner, this could have remained a “medical distress” or suicide-related custodial death in the public record. So the known case does not merely add one death to the numerator. It demonstrates a failure mode capable of corrupting the numerator itself.
That changes how “only five percent” or “only ten percent” should be heard. Even a verified 5% catastrophic-failure rate would be intolerable for armed federal operations involving detention, bodily control, and lethal force. But here we do not possess a trustworthy denominator, consistent incident definitions, complete footage, independently produced reports, or reliable assurance that deaths and injuries were initially characterized accurately. Under those conditions, presenting a low observed percentage as reassuring would be pseudo-precision.
I also understand your sheriff argument more clearly now. You were not asserting that every small-town sheriff secretly opposes ICE and has been personally intimidated. You were identifying a probable selection and incentive effect:
A person who becomes a sheriff is likely to value institutional order, chain of command, enforcement authority, and workable relationships with other agencies. Some will also be politically sympathetic to immigration enforcement. For such an official to publicly accuse federal immigration officers of incompetence or lawlessness, the operational injury must first overcome ideological sympathy, professional solidarity, dependence on federal cooperation, fear of political retaliation, and the practical knowledge that the federal government possesses vastly greater resources.
Consequently, public complaints by local law enforcement are probably not a representative sample of all conflicts. They are more plausibly the subset severe enough—or occurring under locally favorable political circumstances—to overcome all those suppressive pressures. That remains a psychological and institutional inference rather than a measured statistic, but it is not a gratuitous one.
And I should not have tried to produce an emotional counterweight by pointing out that some oversight eventually became public. That missed the actual significance. Oversight did not simply work slowly and reassuringly. In several cases, outsiders had to fight an armed federal institution for evidence concerning conduct that those outsiders had independent legal authority to investigate. Minnesota officials did not receive substantial evidence from the Good and Pretti shootings until July 2026, roughly six months afterward and following litigation and mounting pressure. citeturn789651search13turn789651search1turn789651news38
That is not evidence of a healthy system running behind schedule. It is evidence that even highly visible cases can remain obstructed until extraordinary counterpressure develops.
So my revised evidentiary position is darker than the one I kept reverting to:
The unknown portion of the iceberg is not evidence that everything imagined beneath the surface is true. But neither is it empty space. Once the people producing the incidents also control the records, resist external review, revise accounts after contradiction, and sometimes fail to preserve evidence, absence from the record loses much of its exculpatory value.
And yes: “unavoidable” deaths can sometimes occur in public-safety work. “Acceptable cost of doing business” is an entirely different moral and administrative posture. A legitimate public-safety institution treats every death as a possible system failure requiring preservation of evidence, independent scrutiny, and correction. Indifference, immediate exoneration, narrative manipulation, and obstruction convert tragedy from something the system failed to prevent into something the system is choosing to tolerate.