The Theory That Cannot Fail: Curtis Yarvin's Autocracy Got Its Trial Run

In August 2026, the Government Accountability Office published its audit of DOGE’s “Wall of Receipts” — the website where Elon Musk’s Department of Government Efficiency claimed to document billions in taxpayer savings. The watchdog examined $110.3 billion in reported cuts to contracts, grants, and leases. It could not verify the methodology behind 96 percent of claimed grant savings. Of 13,476 contracts DOGE said it had terminated, no termination action was taken on 2,503. $27.4 billion in claimed savings was never executed. And 108 of the 264 leases listed as cuts were already in the process of being phased out before DOGE existed.

DOGE had promised $2 trillion. It claimed $110 billion. The government’s own auditor could not substantiate most of that.

This was not a rounding error. It was the predictable output of a theory that was never really about efficiency.

He Calls It Monarchy. He Also Calls It Dictatorship.

Curtis Yarvin is often described as a monarchist. He uses the word himself. But in a December 2024 talk titled “Monarchism and fascism today,” he wrote that his project is “assuring the ‘benevolence’ of the ‘dictator,’ or as I would say the accountability of the sovereign CEO.” The New York Times, in a January 2025 profile, put it more bluntly: “monarchy” is “basically his friendlier term for a dictator.”

His actual model is called neocameralism — the state as a joint-stock corporation, with residents as customers and a CEO-king at the top. There is no hereditary legitimacy here, no divine right, no throne passed from parent to child. The borrowed aesthetic is not Versailles. It is a Delaware C-corp with a crown emoji.

A traditional monarchist might argue that legitimacy flows from blood, God, or centuries of custom. Yarvin argues that legitimacy flows from executive competence — from a sovereign who can cut through the “Cathedral” of media, academia, and bureaucracy and govern the way a startup CEO would. The monarchy label is window dressing. What he is selling is autocracy with a term sheet.

RAGE Became DOGE, and Efficiency Was Never the Point

Yarvin coined the acronym RAGE — “Retire All Government Employees” — more than a decade before DOGE existed. In November 2024, The Verge drew the line between RAGE and Musk’s planned Department of Government Efficiency before either man had taken office. By May 2025, the Washington Post reported that two DOGE advisers said Yarvin had offered “the most crisp articulation” of what the effort was trying to achieve.

The public pitch was savings. Musk promised $2 trillion. DOGE built a “Wall of Receipts” and claimed $110 billion. The GAO audit found the numbers unreliable, inflated, and in many cases simply wrong.

But Yarvin had already told us what the metric actually was — in March 2025, before the savings failed to materialize. In “Barbarians and mandarins,” he wrote:

“Moreover, cost reduction is not the proper metric for the performance of DOGE. The inefficiency of the executive branch is a symptom. The disease causing this symptom is: the libs… The real KPI of DOGE is its ability to take power from the libs, then keep it. If, by spending a bit more money, we destroy more liberal power and generate more post-liberal power, why not?”

Efficiency was the marketing. Power transfer was the product. When the savings did not check out, he had already moved the goalposts.

Why We Built the Redundancy He Wants Deleted

Yarvin’s central complaint about government is that it is inefficient — bloated with overlapping agencies, duplicate programs, and civil servants who slow down the CEO’s vision. He is not wrong that government can be slow. He is wrong about what the slowness is for.

In 1969, the public administration scholar Martin Landau published “Redundancy, Rationality, and the Problem of Duplication and Overlap” in Public Administration Review. The essay opens with Landau on a plane making an emergency landing after a rudder failure. The pilot explained that commercial airliners are deliberately redundant systems — multiple independent mechanisms to perform critical functions, so that one failure does not bring down the aircraft. Duplication and overlap, Landau argued, are not waste. They are what buy reliability.

Government is not a startup. When a startup fails, the founders dissolve the entity, discharge what debts they can, and try again. That is entrepreneurship. When a government fails, people’s lives collapse — benefits stop, inspections cease, tax refunds stall, weather warnings go unissued. There is no second round.

The asymmetry matters. Yarvin wants to run the state like a company that can move fast and break things. But the things being broken are load-bearing.

The 2026 receipts are concrete. The IRS cut its workforce by roughly 31,000 employees — about 28 percent — and the results showed up immediately in the filing season. The Taxpayer Advocate Service reported that average wait times to reach a customer service representative rose from eight minutes to fourteen — roughly a million extra taxpayer-hours spent on hold. The GAO found that fully staffed Taxpayer Assistance Centers fell from 102 to 42, paper refund processing stretched from 13 days to 36, and scanning systems were unavailable at the start of filing season because experienced IT acquisition staff had been cut. At the Social Security Administration, staffing dropped 13 percent. At NOAA, which houses the National Weather Service, staffing fell 19 percent, with reports that offices struggled to maintain 24/7 severe-weather monitoring.

Speed is not efficiency. A plane with one rudder is not faster. It is a plane that cannot survive a rudder failure.

I Want to Be Fair: He Does Have an Answer

Yarvin is not stupid, and he has heard the objection that concentrated power gets abused. His answer is an engineering solution: an anonymous board of trustees holding cryptographic keys that can remove the sovereign if he becomes a liability.

In the same December 2024 talk, he wrote that “No one knows who the trustees are, not even the king, not even each other. So not even the king can pressure them, sabotaging his own accountability.” He compares this to a corporate board that meets four times a year and can only fire the CEO — not manage the company day to day.

It is a clever design on paper. But the corporate analogy deletes its own preconditions.

A firm behaves because courts above it enforce contracts, because regulators can intervene, and because customers can leave at relatively low cost. A sovereign corporation has no court above it. It controls the police that would enforce any attempt at exit. And it rules people for whom leaving means abandoning language, family, community, and home. Exit disciplines only the mobile — not the poor, the old, or the rooted.

Yarvin’s accountability mechanism depends on a level of institutional trust and technological sophistication that has never existed in any actual autocracy. Meanwhile, the part of his program that did get implemented — gutting the civil service, concentrating executive power, treating government as a restructuring target — required no anonymous trustees and no cryptographic deadman switches. It required only a president willing to sign orders and a Congress willing to look away.

A Business, Not a Charity

Yarvin’s longer-term vision, laid out in his 2008 essay Patchwork, replaces nation-states with thousands of sovereign corporate mini-countries. Residents who dislike their government can leave. The design, he writes, is “all ’exit,’ no ‘voice.’”

His second operating principle: “a realm is a business, not a charity. Its goal is to maximize its discounted return on investment.”

Then he asks the question his own design cannot answer. What happens to residents no realm has a profit motive to accept? He calls them “bezonians” and writes:

“If our design does not provide for the existence of a large number of human beings whose existence anywhere is not only unprofitable, but in fact a straight-up loss, to that realm, it is simply inconsistent with reality. The design faces an existential challenge.”

That is the moral-compass problem, named by its own author and filed as an open engineering ticket. A system whose only feedback signal is return on investment has no way to value a person who cannot generate one. That is not a bug he overlooked. It is the specification.

Three bodies of scholarship make the structural problem clear:

Exit cannot escape public goods. In the chapter “Exit, Voice and Technocracy,” the authors note that “if people experience public bads such as low air quality, there is not necessarily an exit option which allows them to escape.” The very non-excludability that makes public goods hard for markets to provide is what makes them impossible to exit. You cannot shop your way out of a poisoned atmosphere.

Markets fail, and economists have known this for decades. Kenneth Arrow’s 1963 paper “Uncertainty and the Welfare Economics of Medical Care” showed that medical care deviates from competitive-market assumptions because of uncertainty, information asymmetry, and the nonmarketability of risk. Arrow did not argue that every market failure automatically justifies government intervention — he was careful about that, and we should be too. But he demonstrated that the premise “markets solve everything” is not a description of reality. It is a simplifying assumption that breaks down precisely where human stakes are highest.

Firms are already private governments. Elizabeth Anderson, in Private Government (2017), defines private government as rule by authorities with “arbitrary, unaccountable power over those it governs” — and specifically rejects the claim that workers’ freedom to quit makes employer autocracy acceptable. For most people, the alternative to one dictatorial workplace is another one. Yarvin’s proposal is to make the state work the way Anderson shows the workplace already does.

There is also an internal inconsistency in Yarvin’s own evolution. The Patchwork-era argument held that competition between sovereign realms would discipline bad governance — firms improve because rivals exist. The later Yarvin grounds good government in the CEO-king’s absolute authority instead. He kept the power and dropped the mechanism that was supposed to make it accountable.

Karl Polanyi, writing in The Great Transformation (1944), put it differently: treating land, labor, and money as commodities — when they were not produced for sale — “means to subordinate the substance of society itself to the laws of the market.” Markets do not have moral compasses. They have price signals. A price signal cannot tell you that a disabled child deserves care, that a rural county deserves weather warnings, or that a veteran deserves the benefits they were promised.

What follows is prediction, not evidence. In my view, if the only scoreboard is profit, the result is a crueler and more unequal society — and what we have seen so far is the opening of that process, not its full extent. The IRS delays, the SSA staffing cuts, the hollowed-out weather service: these are not the endpoint of Yarvinism. They are what happens when you apply the theory halfway, with checks and balances still partially intact. I do not think the full version would be better.

The Efficiency Premise Is Empirically False

Even on its own terms — that autocracy governs more efficiently — the evidence is weak.

Acemoglu, Naidu, Restrepo, and Robinson, in “Democracy Does Cause Growth” (Journal of Political Economy, 2019), found that democratization raises GDP per capita by roughly 20 percent over 25 years. A survey in the Journal of Democracy notes that authoritarian regimes have been overrepresented among both the fastest- and slowest-growing economies in every decade since the 1960s — the “autocratic gamble,” where growth varies more widely and, on average, runs a deficit compared to democracies. Democratic breakdown costs 1.5 to 2 percent of per capita income on average, rising to 4 to 6 percent after two decades of autocratic rule.

The honest concession: autocracies can act faster. Speed is not the same as governing better. A demolition crew works faster than an architect. That does not make it the better choice for building a house people have to live in.

What He Actually Said, and Why It Ends the Argument

In December 2025, Yarvin published “The situation and the solution” on his Substack. He did not disavow the project. He mourned its incompleteness.

“The story now seems clear: the second Trump administration is a tragedy,” he wrote. “I hate to say it, but it looks like the administration has already lost.”

His complaint is not that Trump went too far. It is that Trump did not go far enough:

“Last winter’s shock-and-awe Trump is vanished now. The administration is simply too well-integrated with the permanent government. This marriage is terrible, but it is still a marriage.”

And:

“Trump does not actually want full power. Rather, he is afraid of it. And it is not just him. He is actually far less afraid than the people around him.”

Read that carefully. Yarvin is not embarrassed by the authoritarianism. He is disappointed that it stalled. The tragic flaw, in his telling, is insufficient will — not insufficient caution.

Combined with his March 2025 hedge that “cost reduction is not the proper metric,” every possible outcome now has a pre-assigned explanation that vindicates the theory. Savings fail to materialize? Wrong metric. Checks and balances hold? Insufficient commitment. The administration integrates with the civil service instead of dismantling it? Tragic flaw. This is the structure of “real communism has never been tried.” A theory that no outcome can embarrass is not a theory. It is a faith.

And he was never a neutral observer warning of dangers. In December 2024, the Guardian headlined its profile: “He’s anti-democracy and pro-Trump.” Time reported that Yarvin had argued a hypothetical Trump administration should terminate federal workers, impound government coffers, ignore courts that blocked unconstitutional orders, and curtail the free press and universities “no later than April after the inauguration.” He attended Trump’s second inauguration as an informal guest of honor. He was not sounding an alarm. He was writing the blueprint.

The Litmus Test

Yarvin’s program was partially implemented. A fair test asks what happened.

Executive power did move. Capitol Trace tracked roughly 271 executive orders in the first 18 months of the second Trump term — about 15.1 per month, against a historical average near 2.9, and more orders in eighteen months than the 220 issued across his entire first term. Against 45 bills signed into law. The Yale Law Journal described Executive Order 14,215 — “Ensuring Accountability for All Agencies” — as reflecting “the triumph of a legal movement to reestablish a unitary executive.”

Congress’s power of the purse was tested. A PS: Political Science & Politics article documented that by summer 2025, the administration had frozen or canceled more than $400 billion in congressionally appropriated funds — and that where Nixon’s $18 billion impoundment drew overwhelming bipartisan repudiation, the 119th Congress responded with “little more than a whimper.” Senator Chuck Grassley, a Republican, told acting OMB director Russ Vought: “Congress has appropriated money, and you don’t have the authority to impound it.”

The input was delivered. What was the output?

Approval sits at 33 to 34 percent (Economist/YouGov, August 7–10, 2026). An FT/Focaldata poll found 53 percent of registered voters — including nearly a quarter of Republicans — say they are financially worse off than on January 19, 2025. Angus Reid found that 47 percent could name no accomplishment at all, and 36 percent named “his entire presidency” as the signature failure.

I want to be precise about what this does and does not prove. Tariffs, the Iran war, and broader macroeconomic forces are the larger drivers of cost-of-living frustration. I am not laying grocery prices at Yarvin’s feet. The narrower claim is stronger and survives scrutiny: the state-capacity destruction documented by the GAO, the IRS, the SSA, and NOAA is directly attributable to the RAGE-descended program of workforce cuts and executive consolidation — and that is the part that was supposed to make government work better.

The efficiency pitch failed on its own terms. The power-transfer pitch succeeded just enough to do real damage. And the theorist who designed it looked at the wreckage and said: not autocratic enough.

The Critique and the Cure

Yarvin begins from a real observation: American institutions can be unresponsive, self-protecting, and insulated from accountability. He is not wrong that the administrative state has its own interests. He is not wrong that Congress often fails to exercise its constitutional prerogatives. He is not wrong that Silicon Valley’s obsession with “moving fast” found a receptive audience in a political culture tired of gridlock.

But his cure is the disease at a higher dose. He begins by observing that elites rule without accountability and ends by designing a regime where accountability is abolished on purpose — replaced by an anonymous board of trustees and a cryptographic deadman switch that has never been built, tested, or deployed anywhere on earth.

What we got instead was the partial version: the workforce cuts without the benevolent CEO, the executive orders without the corporate discipline, the power transfer without the accountability mechanism. And even that partial version degraded the services people depend on, inflated its own savings claims, and left the president less popular than when he started.

Yarvin’s response was not to reconsider the theory. It was to insist the experiment had not gone far enough. That is not how science works. It is how ideology works.

A theory that cannot fail is not a theory. It is a promise that someone else will always be responsible for the consequences — and a guarantee that when the consequences arrive, they will be reclassified as insufficient effort rather than evidence. We are living in the evidence. The checks held just enough to prevent the worst version. They did not hold enough to prevent damage. And the man who wrote the blueprint is already explaining why the damage means we should try harder.