The ‘procedure fetish’: why the rules meant to make government work are breaking it
“It’s one more bottleneck in a string of bottlenecks that includes judicial review, the extraordinary rigors of the notice and comment process, endless public feedback … that makes it extremely difficult to do anything of any consequence.” ~ Nicholas Bagley, a law professor at the University of Michigan and one of the most important thinkers on why American government has become so mired in process. Here is the discussion in context.
A few years ago, Los Alamos County identified a safety hazard on Central Avenue. Parked cars were blocking drivers’ sightlines at crosswalks, preventing them from seeing pedestrians stepping into the street. The fix has a name: daylighting. You remove the parking space immediately before the crosswalk. The cost is essentially zero. The safety benefit is well-documented. The County has known about the problem and the solution for years.
At the May 12, 2026, County Council meeting, Councilor Suzie Havemann gave voice to what many residents had been thinking. She couldn’t understand why the County was spending years studying the problem instead of acting—why not simply talk to the affected businesses? “We’re talking like four businesses,” she said. “[So] go talk to Laura Crucet at Sugar & Cream and see how she feels about it, and then let’s get on with it. Because it’s just an accident waiting to happen, and we’ve known this for a long time.”
Days after Havemann spoke, traffic cones appeared in those spots. And a day after that, the parking spots had been slathered in orange “you can’t park here” paint. The difference was dramatic—see the before-and-after photos at the end of this piece. I sat on a park bench and watched the crosswalk by the post office on a busy weekday afternoon: drivers stopped for every pedestrian. The County proved that with some prompting, it could move swiftly. But why crawl along for years only to sprint at the finish line? What took so long?
It’s not just Los Alamos County that studies things for so long that literally entire generations can pass between idea and action. The reason our government is so kludgy is the same reason all levels of government—from local to state to national—fail to deliver what residents need. It’s the reason why nothing seems to work anymore.
It’s what University of Michigan law professor Nicholas Bagley calls the “procedure fetish.”
The built-in bias toward doing nothing
In a 2019 Michigan Law Review article titled “The Procedure Fetish,” Bagley argues that modern American government has layered procedural requirements on itself to the point of paralysis. Before a federal agency can issue a major rule, it must notify the public, accept and respond to comments, conduct a cost-benefit analysis, survive White House budget review, and then defend the final product against legal challenge—a process that can take years and still end in a judge throwing it out. States and localities pile on their own versions.
This process is meant to protect the public by requiring agencies to hear from ordinary people, not just industry insiders—and ensuring that, before the government acts, it has carefully considered what it’s doing.
Bagley (who is, recall, a professor of law) disputes the premise that the public comes out a winner in this system. He writes:
[I]t is simply wrong to assume that more procedures will discourage capture because those procedures aim to foster deliberation, transparency, and rationality. The reverse will usually be true. Exploiting a procedural opportunity takes time, attention, and resources. The same interest groups that are the villains of the capture narrative can deploy their relative organizational advantages to pull procedural levers with more frequency and greater expertise than groups representing the public interest.
In practice, proceduralism raises the cost of action but not of inaction. A government agency that wants to do something—issue a rule, build a thing, enforce a standard, make a change—must navigate the entire obstacle course. An agency that sits on its hands faces no burden at all.
Liability overwhelms all other concerns
At the same May 12 Council meeting on Road Safety, Public Works Director Eric Martinez explained why his department is careful before installing even a plastic bollard in the road.
When you do these things, we do them very deliberately, following national guidelines. …Even a plastic marker in the middle of the road. There’s standards for the application of things like that. And a lot of it does have to do with liability. “Well, it was there yesterday, but not today,” And somebody hit it. Our crews weren’t able to get to it and reinstall it, and so now the County may be liable because it was there before [and now it’s not and it] would have prevented, or maybe prevented, an accident. So things that are hard to maintain can put us into a bit of a liability situation. And even things that we may install that may not be within standards are hard to defend. And I’m going to look at our attorneys, because that’s something we have to be just very conscious of.
In other words, people can get hit and injured/killed in a crosswalk, and the County isn’t liable for that, because it’s the driver’s or the pedestrian’s fault. (“Look both ways” and other safety PSAs reinforce this message.) But if the government has provided lifesaving infrastructure, then the maintenance of that infrastructure could become a liability. Because it works. The very thing that saves lives becomes a hassle and a headache for them to maintain. Put another way: the absence of safety infrastructure is legally safer than its presence.
It’s not that the Public Works Director is making things up. Liability is real. Lawsuits can drain municipal budgets. Standards exist for reasons. And nobody wants to get yelled at—either by their manager, the attorneys, or the public.
“People who go into public service... spend their days frightened,” said Marc Dunkelman, on a recent appearance on the Pod Save America podcast on the topic of power and proceduralism. “They’re frightened that they’re going to get their boss in trouble... they’re frightened that they’re going to make a decision that’s going to get them subject to a [lawsuit]. I think we want to re‑empower people who go into public service to be able to make those sorts of changes.”
Of course we want our government to be appropriately cautious, but notice what “caution” often points toward. What is being protected? Lives? Or backsides? One way to tell is to count how often “liability” comes up in discussions about safe streets. The apparatus appears geared at protecting the government from lawsuits, not at protecting pedestrians from cars. Even though people are dying or being injured right now, any new idea to protect residents is met with endless catastrophizing about legal exposure. Departments are stuck in analysis paralysis. So local governments wait, study, consult, document, and defer—because doing something carries legal risk, and doing nothing does not.
Multiply the Los Alamos example across all levels of government, everywhere in the country. The procedures meant to ensure careful action end up ensuring no action. “If America has a procedure problem,” Bagley writes, “it may be because it has a lawyer problem.”
From ‘protect the people from government’ to ‘protect the government from people’
The procedural turn in American governance traces back to two waves of distrust, both reactions to the expansion of federal power during and after the New Deal.
The first came in the 1940s, when business groups alarmed by FDR’s administrative state pushed for the Administrative Procedure Act—a due-process fix requiring agencies to provide notice, explain their reasoning, and give affected parties a chance to weigh in. The second arrived in the 1960s and 70s, when Vietnam, civil rights, and Watergate eroded the left’s faith in government as a vehicle for progress. Their primary mechanism for forcing good behavior from government was the lawsuit.
But litigation changes behavior in ways that few seemed to anticipate. Jennifer Pahlka, founder of Code for America and former U.S. Deputy Chief Technology Officer, has spoken about this problem: “Every time we sue, we make the government more risk-averse,” she said in this conversation on the Ezra Klein podcast about government failures. “There’s a lot of adversarialism out there, and the natural result of that is going to be a system in which you defend your judgments by using no judgment.” The result:
A culture emerges in which the goal is for decisions to be the outcomes of processes, not people. You want to be able to defend any decision from criticism by demonstrating strict adherence to a process in which no judgment can be questioned, because no judgment was used. ~ Jennifer Pahlka, The Cascade of Rigidity
The perverse outcome is that perfectly nice people find themselves thinking things like: OK, sure, someone died, but look—I followed every rule. It’s not MY fault. And in a narrow sense, they’re right. As Pahlka says in her conversation with Klein, “It’s quite true, in people’s careers, that they are not punished for something not working as long as they can say, ‘I did what the procedure prescribed.’” To this day, despite decades of evidence to the contrary, the pro-litigation camp holds on to its faith that lawsuits will make the government behave better. They don’t grapple with the fact that lawsuits have produced a defensive-crouch culture that does not serve the public interest.
Lawyers have an important job, but it’s not running the government. Or it shouldn’t be. As legal risk-aversion increasingly drives American decision-making, we are rapidly ceding ground to nations that have made building things their central competency.
What builds trust in government? Process? Or outcomes?
The official defense of all this procedure is that the process itself makes government legitimate—that citizens can trust a government that follows the rulebook carefully.
Bagley says this defense pretty obviously doesn’t hold water. “Legitimacy is not solely—not even primarily—a product of the procedures that agencies follow,” he writes. “Legitimacy arises more generally from the perception that government is capable, informed, prompt, responsive, and fair.”
It’s very difficult to argue that the procedure fetish is necessary to gain public trust when we’re drowning in process, and nobody trusts the government. A government that follows every procedural rule but cannot act—cannot build, cannot install a simple plastic bollard, cannot respond to a documented safety hazard at anything approaching normal human speed—is untrustworthy. And when people decide that government is untrustworthy, the ground becomes fertile for disillusionment and demagoguery.
At the May 12 meeting, Suzie Havemann was not calling for the government to abandon the process entirely. “I think we all have said that public participation and getting input is super important,” she said. “And we value it as a governing body… but sometimes it causes delays, and then it’s frustrating.”
A handful of parking spots, a few businesses, a known hazard, an obvious fix. Go talk to the business owners, she said, and let’s get on with it.
They got on with it.
What a responsive government looks like
Within days of that meeting, the County removed the parking spaces that blocked sightlines at the most heavily pedestrianized crosswalks on Central Avenue. I went back with a camera.
Pedestrian activity was heavy, but every driver I saw stopped to let pedestrians cross. The removal of parking spaces did exactly what the research shows will happen: it made all road users visible to each other.
This solution couldn’t be much simpler. It essentially costs nothing. The hazard had been identified years ago. What changed was that first, the public got more insistent about its desire to see safety interventions along this corridor. Second, an elected official said, on the record, that the waiting had to stop. This one-two punch paid off—not just for road users, but for the County’s standing. When the County put out a press release on the change, the comments on social media were mostly positive—which, for social media, is a small miracle. “As both a daily driver and pedestrian in that area, I’m so grateful for this!” wrote one resident. “Smartest decision ever to keep pedestrians safe!” wrote another.1
That is what Bagley means when he writes that public trust rises when government is seen as “capable, informed, prompt, responsive, and fair.” Legitimacy doesn’t come from rigidly following procedure, says Bagley—it comes from delivery. It comes from results. It comes from responsiveness.
The people calling for a capable government are not the same as the people calling for a smaller one. Anti-statists want to dismantle the government entirely so that concentrated private power faces no check. That is the opposite of what Bagley, Pahlka, and Dunkelman are arguing for. A government too risk-averse and too tangled in procedure to act is not a government that protects anyone, least of all the people who need protection most.
Galen Rosenberg’s “Look both ways! Don't wanna die in a crosswalk” makes a complementary argument from the ground level: as an American currently living in South Korea, he sees the connection between infrastructure design and public safety—and draws a contrast with Seoul, where investment in design, not PSA campaigns, cut pedestrian fatalities by 60% in less than two decades. I highly recommend his essay.
Great article. I have tons of cognitive dissonance about this because like many liberals, I believe procedure is important. I started questioning NEPA around 2015, and I reread the statute with fresh eyes.
In short, in a bureaucratic world process is safe whereas an end product is risky. That's because price of failure is much bigger than the reward for success. This is the inverse of most businesses where the expectation is to get something done.
It won't change.