BY Table of Contents
Every dissatisfied person holds two instruments. The interesting question is not which one is right. It is why each only works when the other is loaded.
You are unhappy with something you belong to. A company, a church, a party, a platform, a country.
You have exactly two instruments. You can complain and try to change it from inside. Or you can leave.
Albert Hirschman named them in 1970, in Exit, Voice, and Loyalty: Voice and exit. He also noticed a professional bias worth keeping in view. Economists reflexively overvalue exit, because markets are exit machines. Political scientists reflexively overvalue voice, because politics is a voice machine. Most people, deciding whether to leave a job or a homeland, are running both calculations at once with no theory at all.
On the political right, this has hardened into the core divide: Do you stay and fight, for the institutions, for the land your forefathers settled, or do you get out and build parallel ones? Secession, sovereign individual thinking, network states, homeschooling, crypto on one side. Precinct captains, school boards, and long marches back through the institutions on the other.
The divide is sharper now than it has ever been, for a reason that usually goes unstated: There is no Terra Nova left. What made the Age of Exploration an exit valve was the New World itself, an entire hemisphere of untapped land, at least as the Europeans saw it, open to anyone willing to board a boat. That valve is closed. Today every square mile of habitable land sits inside the nation-state model; you can change jurisdictions, but you cannot leave jurisdiction itself. It is why seasteading keeps resurfacing among people who think hard about this, why living on your own yacht in international waters, sailing between jurisdictions, is regarded in certain circles as the ultimate Plan B, and why the colonization of Mars, the moon, and open space is interesting well beyond the engineering: Each is an attempt to escape the nation-state model at its core, to manufacture the Terra Nova the map no longer offers.
Both camps quote their casualties. Neither, usually, quotes Hirschman past the title. Which is a shame, because he supplies the strongest argument for each side, and then a case study where the two instruments stopped being rivals altogether.
The Case for Staying and Fighting
I read something a few years ago elsewhere online entitled Secession is for Cowards, from Sam Jacobs, who also writes here on-site. It is the sharpest statement of the voice position I know, and his argument ran like this.
Exit does not satisfy an expansionist opponent. Hand over the cities and you have not bought peace; you have bought your adversary a consolidated base. Waiting for a hostile system to collapse under its own contradictions is an ahistorical fool’s errand:
Sure, the Soviet Union collapsed under the weight of its internal contradictions – but it took them 80 years to do so.
And the everyday version of exit, the drumbeat of don’t join the military, don’t vote for the compromised party, don’t even watch the game, is, in Jacobs’ phrase, foolish at best and a blackpilling psyop at worst. His counter-program is participation: Precinct captain, zoning committee, volunteer fire department. Positions that go to whoever shows up.
The strangest witness for this side of the argument is the man who spent two decades building the ultimate exit vehicle. Peter Thiel, talking to Ross Douthat, tells of a 2024 conversation in which he told Elon Musk that if the election went the wrong way, he wanted to simply leave the country. Musk’s answer: “There’s nowhere to go.” It took Thiel a couple of hours to register what he had heard. Musk had stopped believing in Mars, not as an engineering project but as a political one, a place to build an alternative society. Years earlier, Demis Hassabis had needled Musk that “my AI will be able to follow you to Mars,” and by 2024 it had sunk in: The socialist government would follow you to Mars. The woke AI would follow you to Mars.
Sit with who is saying this. The most committed exit visionary alive ran the numbers on the ultimate exit, concluded that what he was fleeing would arrive on the next rocket, and turned to fighting political battles on Earth instead. When the man building the escape vehicle chooses voice, that is a real win for the voice camp.
Hirschman, interestingly, has ammunition for this side that Jacobs doesn’t use. His original book was built on a failure of exit: Nigeria’s railroads. When trucking gave Nigerian freight customers an alternative, the railroads did not improve. They got worse, because the customers who cared most, the ones who would have complained loudest, left first. Hirschman called the result a lazy monopoly: An institution that actually welcomes the exit of its most demanding members, because their departure buys silence.
South Africa is running the same experiment right now with Eskom, its state power monopoly, one thread of a broader unraveling I’ve written field notes on before. As load-shedding stretched into hundreds of days a year, everyone who could afford to began bolting solar panels to their roofs and inverters to their walls. The customers with the most money and the loudest voices, the businesses and the affluent suburbs, are quietly leaving the grid, and their departure takes the revenue and the pressure with them. The customers who remain are the ones who cannot leave and cannot make themselves heard, and the monopoly, relieved of its most demanding accounts, limps on. By early 2026 it was even declaring victory over load-shedding – more than a year now without cuts. But the victory is primarily demand destruction: Eskom sells less electricity every year as smelters close and rooftops generate. The economy isn’t growing – or at least not on Eskom power. Hirschman would have recognized every step.
Notice the kind of exit this is: Quiet, private, roof by roof. Nobody marches out of Eskom; they just stop needing it. Hold that detail, because the difference between a loud exit and a quiet one turns out to be the hinge of this whole question.
Read the same mechanism onto American life and it stings. Every capable family that quietly pulls its children from the public school is one fewer parent at the board meeting. Exit can be a gift to the incumbents. The door swings shut behind the very people who could have forced the change.
The Case for Leaving and Building
Now the other side, which I will state just as plainly.
Voice without a credible exit is begging. I have written about that posture firsthand: An American expat at a consulate counter, renewing the passport of a government that taxes him wherever on earth he lives, is called a citizen but stands there as a supplicant, because both sides of the counter know he cannot take his business elsewhere.
An institution listens to complaints in proportion to what walking away would cost it. A restaurant takes your feedback because you can eat elsewhere tomorrow. The department of motor vehicles does not, because you can’t. The people inside are not the problem. The structure decides what they can afford to ignore. If you want to know which kind of counterparty you are dealing with, listen to how it apologizes when it fails you. An institution that never names its own conduct has already decided your leaving costs it nothing.
The state has always known this, which is why its oldest quarrel is with the people who move. Ask yourself: Why is the state seemingly always the enemy of nomads, pastoralists, Gypsies, vagrants, runaway slaves and serfs? Because its classic imperatives are taxation, conscription, and the prevention of rebellion, and all three require knowing where everyone is. Permanent surnames, land registries, censuses, passports: The whole apparatus of legibility exists because you can’t tax what you can’t measure, and you can’t measure what won’t hold still. James C. Scott built a career on the same observation, and his history of the Southeast Asian highlands where runaways from a dozen kingdoms pooled for centuries, The Art of Not Being Governed, reads as the biography of one of the largest exit communities that ever existed.
The states that understood this best are the ones that banned leaving outright. The Berlin Wall was not built to keep enemies out. It was built in 1961 to keep citizens in, after 2.7 million East Germans, roughly a fifth of the country, had walked west, and a state that walls its people in has confessed, in concrete, that it could not survive an open door. Exit visas, emigration taxes, and capital controls come from the same fear.
That is why the modern exit movement is not mere flight, and why so much of it is built in the digital realm: With no Terra Nova left on the map, the frontier has to be manufactured somewhere the nation-state hasn’t fenced. Sound money, jurisdictional competition, parallel schools and media – Balaji Srinivasan’s whole program is Hirschman weaponized – are attempts to build, without new land, the alternative that makes voice credible again everywhere else.
Taylor Pearson, a past guest on the SGI podcast, pushes the question a step further, and notice which of Hirschman’s examples his version runs through. Software, he points out, is non-rival. There was only ever one Nigerian railroad; a dissatisfied customer could not copy it and run a better one next door. But code can be forked. Anyone dissatisfied can take a copy of the entire system at a point in time and leave with it, a form of exit the physical world never offered, and Bitcoin is the monetary version: An exit from the money itself, executed without anyone’s permission. If Pearson is right, technology is quietly tilting the ancient bargain between staying and leaving in exit’s favor.
And there is a deeper precedent the American right, of all audiences, should not need reminding of: America is a nation founded and populated by people who chose the boat over the ballot they didn’t have. Exit is not un-American. It is the founding gesture.
East Germany Ran the Experiment
For twenty years Hirschman’s readers treated exit and voice as a seesaw: More of one, less of the other. Then 1989 happened to the country that had built the wall, and Hirschman himself wrote the postmortem.
In the summer of 1989, Hungary opened its border fence. East Germans began pouring out, tens of thousands, through Hungary and the West German embassy in Prague. By the seesaw model, this mass exit should have drained the pressure for change at home: The angriest people were leaving.
The opposite happened. As the exodus swelled, the Monday demonstrations in Leipzig grew from hundreds to hundreds of thousands, chanting, of all things, Wir bleiben hier. We are staying here. The people in the squares could credibly demand change precisely because everyone, including the regime, could see the alternative draining the country white in real time. Exit did not undermine voice. Exit armed it. Within months the wall was gone, and Hirschman revised his own theory to fit the evidence: In the terminal crisis of the German Democratic Republic, the two instruments worked in tandem, each amplifying the other.
Hold both of Hirschman’s cases side by side, because together they dissolve the dichotomy:
- Nigeria: Quiet, private, individual exit, invisible to everyone but the accountants, drained voice and comforted a lazy monopoly.
- East Germany: Loud, visible, mass exit, undeniable to everyone, made voice irresistible.
The variable is not whether people leave. It is whether the leaving is legible, whether the institution and everyone still inside can see the cost being paid. An exit nobody notices subsidizes the status quo. An exit everybody notices is itself a form of voice.
You Are Already Running Both
Jacobs concedes, early in his piece, that the dichotomy is partly false: Every socially aware person is already pursuing some combination of the two. I’d push that further. The combination is not a compromise between the strategies. It is the strategy.
I write this as an American expat, which means I have exercised the exit option in its most literal form. I have also spent years writing, in public, about American ideas, American founders, and American failures, which is voice with the volume up. The two turned out not to be alternatives. Leaving gave the voice something it lacked at home: Independence from the institutions being criticized, and living proof that the alternative exists.
Hirschman’s third term, the one everybody forgets, is loyalty, and its function in his model is exactly this: Loyalty is what keeps you using voice after exit becomes available, and what makes your eventual exit mean something if it comes. The man who was never going to leave is easy to ignore. The man who could leave, and visibly stays to fight, is the one institutions fear.
I have also watched the voice program run honestly and still lose. Mark Pulliam, a guest on my podcast, is a retired lawyer who moved to East Tennessee and did everything Jacobs prescribes: Built an organization, held meetings at the public library, showed up to the county’s committees with better arguments. The old guard never answered the arguments. They controlled the appointments and the procedures, and they used the machinery to freeze him out. I told the fuller version in Critic vs. Revolutionary, and it belongs next to Jacobs’ program as its price tag: Showing up is necessary, and the incumbents get a vote on whether it is sufficient.
So the practical synthesis, for anyone weighing where to put their efforts:
Fight where you can be heard. Leave where you cannot. And leave loudly, because a legible exit is a vote the institution can’t pretend wasn’t cast.
One collision in this essay has been left standing, and I want to point at it rather than pretend to resolve it. Taylor Pearson says technology tilts the bargain in exit’s favor: Code can be forked, and the fork can’t be stopped. Demis Hassabis says the AI will follow you to Mars. They cannot both be entirely right, and which one is may be the most consequential open question of the next few decades: Whether technology is exit’s best weapon, or the thing that finally closes every door. If it is the first, the future belongs to the builders of parallel things. If it is the second, voice stops being one option among two.
It becomes the only one left.
Jacobs closes by allowing that preparing for the future probably includes knowing where the fire exits are, then dismisses that as a strategy for retreat, not victory. He is right about the retreat, and wrong about the fire exit. The fire exit is not there for the day you use it.
It is there because a building whose occupants can leave is a building the management maintains.