They Knew and They Did Nothing
A federal worker's account of the system both parties built to fail
I wrote this as a follow-on commentary after watching Matt Robison ’s and Melissa Amour’s interesting video. Find it at:
I spent many decades in federal service watching safeguards slowly gutted. Not from the top. I was a national security analyst, a supervisor, a mid-level manager. I witnessed the machinery of government from close enough to smell the gear oil, but nowhere near the controls. What follows comes from that vantage point.
What I think I saw, over all those years, was a system that ran on trust. Institutional trust. The kind that says we’ve all agreed, more or less, that there are rules, and that the rules apply to everyone including the people in charge. What the Trump administration revealed (I say revealed rather than created because I believe the vulnerability was always there) is that a remarkable amount of what we called checks and balances was in fact just good manners. Shared behavioral conventions with no real enforcement behind them. Norms dressed up as guardrails, painted on the wall, if you will, to look protective.
The inspector general, for instance. The idea that a president couldn’t simply fire the people whose job is to watch him wasn’t a law with teeth. It was an understanding. A gentleman’s agreement. And it turned out that when the gentleman left, the agreement did as well.
Did Congress know?
Of course they did. Hell, if I could see it so could everyone else.
Every member of Congress who has served any meaningful length of time understands how the system actually operates. They’ve watched administrations of both parties push against norms. They’ve seen inspectors general get pressured and slow-walked and ignored. They knew the guardrails were painted on. This wasn’t ignorance. It was a choice, made deliberately, by people who understood exactly what they were leaving in place.
And they left it that way on purpose. Both parties. I want to be specific about that because the easy version of this argument blames one side, and the easy version is wrong. Democrats weren’t rushing to constrain emergency powers when Obama was using them. Republicans showed no interest in limiting executive authority when Bush was expanding it after September 11th. Each party looked at the presidency as a prize they expected to win again, and neither was willing to tie the hands of their future selves.
A politician’s first calculation is always about themselves. Not maliciously, not in some mustache-twirling way. Just structurally. It’s the nature of the job. Codifying everything, making consequences automatic, removing discretion, closing the wiggle room, sounds like good government in the abstract. In practice it means accepting constraints that might one day apply to you or to someone you support. The senator who could have pushed to codify inspector general independence in 2005 or 2015 looked at that proposal and thought, somewhere in the back of their mind: but what if my president needs to remove an IG who’s causing problems? The norm served them well enough. Why make it a law that binds everyone forever, including us?
The outrage being expressed now by many current and former lawmakers rings a little hollow. They had years, decades in some cases, to fix this. They knew it was broken. They chose flexibility over the safeguard because the flexibility might one day be theirs to use. They built the system the same way people leave emergency exits with the alarm off. Not because they intentionally planned to misuse them, but because they wanted the option at need. The Trump administration walked through every unlocked door in the building.
That’s not a partisan observation. It’s an indictment of a political class that valued its own room to maneuver over the integrity of the institution it was sworn to protect. Both parties built this. Both parties are indicted.
Why reform is hard, before we talk about reform.
I want to say something uncomfortable before getting to what I think should change, because skipping it would make everything that follows sound more optimistic than it deserves to be.
No structural reform survives total institutional capture. If the president controls Congress, the courts, and the enforcement mechanisms simultaneously, the constitutional system has already failed and no office or statute anywhere fixes that. If Speaker Johnson controls the House and the Supreme Court has demonstrated comfort with expansive executive authority, an independent watchdog drawing its power from Congress and the courts is a watchdog on a leash held by the people it’s supposed to watch. That’s not cynicism. That’s really just the math.
So why propose reforms at all?
Because total capture is a moment, not a permanent condition. What these reforms are designed to do is raise the cost of getting there, document what happens while you’re there, and preserve enough of the record that accountability becomes possible when the moment passes. They’re not a wall. They’re a cost, a record, and a foundation. That’s worth building even when it isn’t sufficient on its own.
With that said plainly, here’s what I think needs to change.
The enforcement problem nobody talks about
The whole constitutional architecture assumes a neutral executor. The DOJ enforces federal law. The FBI investigates. The executive branch implements court orders. But all of those functions report to the same person the law is supposed to constrain. When that person corrupts the enforcement mechanism, the circuit breaks. There’s no automatic backup. The founders didn’t build one, either because they didn’t believe a president would actually do this, or because they trusted Congress would stop him if he tried.
Congress didn’t stop him.
The Supreme Court has no army. Andrew Jackson reportedly understood this, which is why he felt comfortable ignoring John Marshall. The Court rules. Implementation depends entirely on the executive branch acting in good faith. What we watched instead was what legal observers started calling legalistic noncompliance: using legal language and procedure to perform compliance while defying the substance of court orders entirely. The administration would file the right paperwork, mouth the right words, and continue doing exactly what the court told it to stop. Pursuing contempt against a sitting president routes back through the DOJ, the same DOJ the president controls. The snake eats its own tail, and the rule of law sits on the sidelines.
So who enforces the law when the chief law enforcement agency is the problem?
Under the current structure, nobody. There’s no automatic transfer of authority. No independent body with statutory power to step in. No trip wire that fires when the person responsible for executing the law becomes the person subverting it. The founders left that gap because they assumed institutional shame, congress would grow a spine, or electoral consequence would fill it. Two of those three have proven unreliable. The third comes too late to matter in real time.
This isn’t a technical failure. It’s a design failure that everyone could see and nobody fixed because fixing it meant giving up something they might want later.
An office that doesn’t answer to the president
What I think is needed, and what I haven’t seen proposed with enough specificity anywhere in the current conversation, is a permanent, independent constitutional compliance office. Not housed in the DOJ. Not subject to executive removal. Funded directly by Congress, reporting directly to Congress, with a mandate that can’t be suspended or dissolved by the person it exists to watch.
Its job would be narrow and specific: monitor compliance with federal court orders across the executive branch, investigate credible claims of noncompliance, and refer violations to Congress and to the judiciary with a documented evidentiary record. It wouldn’t prosecute. It wouldn’t legislate. It would watch, document, and report, and its findings would carry legal weight that triggers automatic congressional review rather than waiting for the majority party to decide whether it’s politically convenient to care.
The constitutional challenge will come. Article II gives the president broad authority over the executive branch, and an office that can’t be removed by the president will face serious legal attack. Good. Design it hardened against that challenge from the start. Anchor it in Congress’s oversight authority under Article I. Give it civil rather than criminal jurisdiction to avoid the separation of powers problems that plagued the old independent counsel statute. Let the courts rule on it with the full record of what happens when no such office exists. That record is now extensive.
The special counsel model is broken for the same underlying reason. A special counsel who serves at the pleasure of the attorney general isn’t independent by any honest definition. The old independent counsel statute had real problems. Ken Starr made sure everyone remembered them. But the core idea was right: an investigation triggered automatically by specific statutory conditions, with a reporting line to Congress rather than the White House, insulated from removal except by congressional vote. The lessons of the 1990s aren’t a reason to abandon the concept. They’re a design brief for getting it right.
What happens to the record during capture
This is a question that doesn’t get asked nearly enough; it needs to be understood before implementation. If this office exists and the executive moves to shut it down, what stops them from erasing everything it built?
In the current structure, honestly, not much. Funding gets cut if Congress is captured. Staff gets reassigned or pressured out. Records get classified or buried in litigation. We’ve watched this happen in real time with less controversial institutions. An independent compliance office wouldn’t be immune.
But records are harder to erase than institutions, and that distinction is the whole design principle.
The office’s findings, if properly structured by statute, would flow simultaneously to multiple repositories the moment they’re produced. Both chambers of Congress, including the minority. The federal judiciary. The Government Accountability Office. The Library of Congress. And by statutory requirement, the public record, published in real time rather than archived internally where they can be quietly buried. Eliminating the office doesn’t eliminate what it already documented, any more than firing an inspector general erases reports already filed. The corruption of the institution and the corruption of the record are two separate operations, each requiring its own visible act of bad faith, each documentable in its own right.
You can shut the office down. Doing so in the middle of an active investigation is itself evidence. Attempting to scrub the public record afterward is a second act of evidence. Pressuring the Library of Congress, the GAO, and minority leadership simultaneously, across party lines, in full public view, is a third. Each act of erasure is harder and more visible than the one before. That doesn’t make erasure impossible. It transforms what might have been a single quiet act of executive discretion into a prolonged, multi-front, publicly documented assault on the record itself. That’s a different political and legal animal entirely.
Then there’s dispersal. If the office’s records are distributed by statute to state attorneys general across all fifty states, the act of erasure becomes a genuinely different order of difficulty. You’d have to simultaneously capture or intimidate dozens of state-level officials across both parties, in states with their own constitutions, their own courts, and their own electorates, many of whom would have every political incentive to preserve and publicize exactly what the federal executive was trying to erase. Red state attorneys general during a Democratic capture. Blue state attorneys general during a Republican one. The dispersal doesn’t have to be perfectly resistant. It has to be resistant enough that the cost of erasure exceeds the benefit, or visible enough that the attempt becomes the story.
This is the most important structural idea in the piece. The founding design protected rights through separated powers: give different institutions competing authority and let them fight it out. The dispersal model applies that same logic to the documentary record. Don’t protect it in one place. Spread it across enough competing jurisdictions that no single act of political will can reach all of it at once.
Other things that need fixing
The independent compliance office matters most, but it doesn’t stand alone.
The inspector general statute needs actual teeth. Firing an IG should require congressional approval, or trigger automatic judicial review on a fast timeline that doesn’t depend on the DOJ to bring the case. The 2022 strengthening of that statute was a genuine step forward. The administration treated it as decorative. A law that gets treated as decorative needs consequences that attach automatically, not ones that wait for someone to decide to pursue them.
The impoundment problem is real and under-discussed. The president doesn’t get to decide not to spend money Congress has authorized. The Impoundment Control Act of 1974 was supposed to settle this permanently. It didn’t. It needs sharper language, automatic judicial triggers, and penalties that don’t require the executive to enforce them against itself. Any reform requiring the wrongdoer to police their own wrongdoing isn’t a check. It’s a damn suggestion.
The pardon power needs constraint. The founders designed it as a mercy mechanism. It’s been used as a corruption shield, a witness-silencing tool, and a blanket immunity operation for political allies. Pre-emptive pardons of people not yet charged should be barred by statute. Pardons in cases where the president has a direct personal interest, meaning cases touching on the president’s own conduct or associates, should require a review process that operates outside the executive branch entirely. This will be challenged in court. Good. Let the full record of how the power has been used be part of that argument.
Emergency declarations need automatic sunset provisions requiring affirmative congressional renewal, not just a joint resolution the president can veto. The current structure lets a president declare an emergency, access sweeping statutory powers, and sustain them indefinitely as long as a third of one chamber stays loyal. That’s not emergency authority. That’s the legislative process wearing a costume.
Schedule F deserves more attention than it gets. The executive order reclassifying tens of thousands of career federal employees as at-will political appointees was revoked by Biden on day one and reinstated by Trump on day one. The professional civil service, the people who actually know where the files are, how the systems work, and what the law requires, can be dismantled before lunch by someone with a pen and no interest in what gets lost. These aren’t political operatives. They’re the accumulated institutional knowledge of a government that took two and a half centuries to build. Their independence needs to be in statute, with real removal protections and real penalties for circumventing them, not in norms that dissolve the moment someone decides they’re inconvenient.
The emoluments clause prohibits a president from receiving financial benefits from foreign governments. It turns out the prohibition is essentially unenforceable without a willing DOJ and a Congress prepared to act. The clause needs a private right of action, the ability for Congress or citizens to sue directly without routing the complaint through the executive branch. A constitutional right with no enforcement mechanism isn’t a right. It’s a sentiment.
The honest accounting
The system didn’t fail because the laws were wrong. It failed because too much depended on voluntary compliance, and the people with the power to fix that preferred the wiggle room. They looked at themselves first, then at the offices they might one day fill, and decided that an inescapable consequence for abusing power was a worse outcome than the risk of someone else abusing it. That calculation was cynical. It was also, by the narrow logic of political self-interest, rational. And it left the rest of us inside a structure that worked fine until someone decided it didn’t have to.
The reforms I’ve described are achievable. Every one of them has a legal and legislative path. What I’m genuinely less sure about, and I say this as someone who spent a career watching political will climb and collapse, is whether the people who’d need to enact these changes have learned anything from watching the system fail. Or whether they’re already looking ahead to the next election, deciding quietly that they’d rather keep the doors unlocked.
Because here’s what I know from my perch. Political will is a renewable resource only when the cost of inaction becomes undeniable. We may not be there yet. But we’re closer than we’ve ever been, and the question of whether that’s close enough is one that history, not I, will answer.
Friends, if my writing has been worth your time, buy me a one-time Ko-Fi below. It goes a long way toward keeping the writing free for everyone.
Further Reading: