A Third View of the State

Political philosophy usually hands you two options for what the state actually is, and asks you to pick a side. The liberal tradition, running from Locke through the American founding to most modern minarchists, treats the state as a neutral arbiter: a night watchman, an umpire, a mechanism people consent to precisely because it protects rights that existed before it did and stops the moment it stops doing that job well. The other tradition, running from Rousseau through a line of Enlightenment perfectibilists to Marx and every twentieth-century movement that carried his name, treats the same institution as something categorically different: not a neutral referee at all, but an organized weapon, built and held by whichever class currently owns the means of production, wielded to keep the other class in line behind a fiction of impartial law.

Pick the umpire and you get classical liberalism. Pick the weapon and you get a call to seize it, or to abolish it once the class it serves has been abolished too — Marx just being the vision's most consequential name for it, not its author.

Neither description survives contact with this site's own record, and it's worth saying plainly why, because the answer isn't "a bit of both" in the mushy centrist sense. It's a third, older, more specific account — one that keeps the Marxist diagnosis of how the state actually behaves while rejecting its remedy, and that explains, better than either standard option does, what a movement built around social liberty is actually for.

Why the umpire story keeps failing its own test

Why Neither Party Wants Social Liberty already catalogued the evidence against a genuinely neutral state, essay by essay, without needing to name the philosophical claim it was undermining. Occupational licensing spread from a narrow mid-century slice of the workforce to roughly a quarter of it today, mostly justified as consumer protection and mostly defended by the professional associations it happens to protect from competition. Control of Human Actions Through Land found the same shape in zoning: a narrow public-health tool that calcified into a permanent veto over what a landowner can do with their own property, defended by an unusually durable coalition of exactly the incumbents it benefits. A Meadow of Sacred Cows found it again in the world's shipping lanes, where American taxpayers fund the naval protection an entire global industry receives for free, a policy nobody would defend on the merits if it were proposed fresh today and that persists precisely because the beneficiary is concentrated and organized while the cost-bearer is diffuse and isn't.

None of those are examples of a neutral umpire making a bad call. They're examples of the umpire consistently, structurally, favoring one side — not because any individual referee is corrupt, but because the game itself has been shaped, arrangement by arrangement, by whoever had the standing to shape it before anyone else showed up to object. A state that behaved the way Locke's social contract describes would produce policy nobody could trace back to a specific, nameable, unearned advantage. This state produces exactly that, over and over, in every domain this site has looked at closely enough to check.

Why the weapon story gets the mechanism right and the remedy wrong

Marx and Engels gave the weapon its most politically consequential name, but they didn't invent the vision it belongs to, and the essay is worth pausing on the older lineage before it goes any further, because the remedy this essay is about to reject wasn't a nineteenth-century economic innovation — it was the latest and most systematized version of an assumption about human nature that had already been circulating in European political thought for a full century before Marx picked it up.

Jean-Jacques Rousseau supplied the founding move: that man is born free and good, and it's the corrupting effect of unjust institutions, not anything intrinsic to human nature, that produces the vice and conflict a society then has to manage — which means the fix was never to constrain a permanently flawed creature, it was to design the institutions correctly and let the underlying goodness re-emerge. William Godwin's Political Justice (1793) pushed the same premise to its logical edge a generation later, arguing that once government's corrupting apparatus was stripped away entirely, reasoned human beings would simply cooperate without needing to be coerced into it. The Marquis de Condorcet, writing his Sketch for a Historical Picture of the Progress of the Human Mind while under sentence of death during the Terror his own revolution had produced, held to the identical premise that reason's steady advance would keep dissolving injustice indefinitely, an optimism his own circumstances were actively disproving as he wrote it down. Marx's actual contribution wasn't the vision — it was locating a specific, allegedly scientific mechanism (the ownership of the means of production) that this century-old assumption about corrigible human nature could be pinned to, and a specific historical agent (the proletariat) that could carry the redesign out.

That's what makes the weapon story worth taking seriously and rejecting for the same reason at once: the state-as-instrument diagnosis is closer to what the evidence in the previous section actually shows than Locke's neutral umpire is, a century and a half before "regulatory capture" entered an economics textbook.

But the remedy inherited the Rousseau-Godwin-Condorcet premise along with the mechanism — abolish the specific arrangement (private property in capital) standing between corrupted institutions and humanity's underlying goodness, and the state, with no class left to serve, withers away on its own, the same way Godwin expected coercion to dissolve once government's apparatus was removed.

That remedy has a hundred years of evidence against it now, and the evidence doesn't point where the theory predicted. Every state that has actually abolished private ownership of capital produced not the state's withering but its intensification — a party apparatus, a nomenklatura, a planning ministry, holding exactly the same "who gets the surplus" power the capitalist class supposedly held, with less of the diffuse, adversarial pushback a market economy's dispersed ownership had at least made possible.

The weapon didn't disappear once the specific class Marx named was removed from the fight. It changed hands and kept firing. That's not a minor implementation failure the theory can shrug off, and it isn't only Marx's specific mechanism that the failure indicts — it's the whole Rousseauian premise underneath it, that removing a corrupting arrangement is sufficient to let a non-extractive human nature take over on its own.

The thing actually driving the capture wasn't capitalism specifically, and it wasn't a correctable defect in institutional design either. It was something older and narrower that capitalism happens to carry, but didn't invent and doesn't require.

The account that actually predicts this

Franz Oppenheimer supplied that older, narrower account in 1908, in a book called simply Der Staat — and it's a genealogy of the state, not an economic system, which is exactly why it survives the failure the Marxist prediction didn't. Oppenheimer's central distinction is between two ways any group can acquire wealth: the "economic means," producing something and trading it voluntarily for something else of value, and the "political means," taking what someone else has already produced, by force or by a standing threat of it, without producing anything in return. The state, on Oppenheimer's account, is the political means organized, regularized, and eventually legitimized — historically, the mechanism by which a conquering group (frequently pastoral, mobile, and militarily organized) subdued a settled agricultural population and converted what had been outright plunder into a durable, heritable claim on that population's land and its output going forward.

Albert Jay Nock's 1935 essay "Our Enemy, the State" carried the same distinction into an American vocabulary a generation later, and it is, not coincidentally, the same distinction Who Pays? and every essay on this site tracing an unpriced cost back to its beneficiary have been re-deriving from scratch, essay by essay, without always citing the source.

The reason this account predicts what actually happened to twentieth-century communism, where Marx's account didn't, is that it never located the mechanism in capitalism to begin with. Political means — capturing value someone else produced, by controlling access to what they need rather than by producing anything yourself — doesn't require private ownership of a factory. It requires control of something scarce that other people depend on, and land is the oldest, least escapable version of that: everyone needs somewhere to stand, farm, or build, land can't be manufactured to meet demand the way a factory's output can, and whoever controls access to it can extract a toll from everyone who doesn't, indefinitely, whether the controlling party calls itself a monarch, a landed aristocracy, a capitalist rentier class, or a central planning ministry allocating housing.

Abolish the capitalist and the mechanism doesn't disappear, because the mechanism was never about capitalism. It was about who controls the scarce, unproducible thing everyone else has to pay to access.

Where this connects to a name

Oppenheimer wasn't working in isolation, and the connection matters more than a footnote usually gets credit for. He wrote explicitly in dialogue with Henry George, whose 1879 Progress and Poverty had already identified land monopoly — not capital, not labor, but the specific unearned rent that accrues to whoever holds title to a valuable location, regardless of what they did to earn that value — as the mechanism siphoning the gains of general economic progress into a narrow, landholding class even as productivity and population grew.

Single Tax for the Living World and Transitioning to LVT have already made this site's version of George's case at length: the value of a parcel of land comes overwhelmingly from what surrounds it — a community's investment, a region's infrastructure, an ecosystem's services — not from anything its titleholder personally added, which is exactly why capturing that value publicly isn't expropriation in the way taxing labor or capital is.

Put Oppenheimer and George side by side and the third view of the state comes into focus. The state didn't begin as Locke's contract among already-rights-bearing individuals, and it didn't begin, either, as Marx's instrument of a specifically capitalist class. It began, and has mostly continued, as the political means organized around the oldest and most durable form of unproducible scarcity there is — control of land and what it yields — legitimized over centuries into something that looks, from the inside, like a neutral system of law rather than what The Rule of the Unquestioned already named for other reified categories: an arrangement that arrived looking like a fact of nature because nobody currently benefiting from it has any reason to point out that it was a choice.

Two temperaments, not two theories

None of the evidence above is new, and that's worth sitting with before moving on to strategy. The Soviet counterexample has been public record for decades. Regulatory capture has been a named, measured phenomenon in economics since the 1970s. Oppenheimer's account has been sitting in print since 1908.

And yet the umpire and the weapon are still, overwhelmingly, the two options on offer whenever this argument gets rehearsed, which means the dichotomy isn't surviving on the evidence. It's surviving on something the evidence doesn't touch.

Thomas Sowell gave that something its clearest name in A Conflict of Visions (1987): not two competing theories about how the state happens to behave, but two incompatible starting assumptions about human nature itself, held prior to and mostly immune from whatever evidence gets marshaled afterward.

The "constrained vision" treats human capacity for reason and virtue as permanently limited, sees existing institutions as encoded solutions to problems no individual mind can fully re-derive from scratch, and trusts a system of checks more than it trusts any occupant of power, however well-intentioned. The "unconstrained vision" treats those same limits as contingent and correctable, sees existing institutions as the residue of past injustice rather than accumulated wisdom, and trusts the right people, armed with the right analysis, to redesign the arrangement correctly if only they're finally given the chance.

Sowell traces that second vision back well past Marx, through Godwin's confidence that reason alone could dissolve the need for coercion and Condorcet's faith in the mind's inevitable progress, to Rousseau's founding claim that corrupted institutions, not human nature, are the actual source of injustice. Marx supplied that century-old vision its most durable twentieth-century vehicle, not its origin — which is exactly why refuting Marx's specific economics has never been sufficient to dislodge it. The vision survives its most famous spokesman's failure the same way it survived Condorcet's own revolution eating him, because the commitment was never really to the specific mechanism any one thinker proposed.

Locke's umpire is what the constrained vision produces when it turns to political theory. The weapon, ready to be seized and redirected by whichever heir to Rousseau's premise currently holds the analysis, is what the unconstrained vision produces when it turns to the same question. Sowell's claim is narrower and more useful than "people disagree": it's that a person's vision, once settled, functions less like a conclusion and more like the lens the next fact gets read through, which is exactly why the Soviet evidence doesn't dislodge the unconstrained side — a mind already committed to that vision reads the same failure as proof the redesign wasn't thorough enough, not as evidence the whole project rests on a mistaken premise, because the vision itself supplies what counts as a disconfirming result before any particular fact ever shows up to be one.

Jonathan Haidt's The Righteous Mind (2012) supplies the mechanism underneath Sowell's pattern, working from moral psychology rather than intellectual history: intuition arrives first, fast and automatic, and reasoning arrives second, mostly employed as counsel for a verdict the mind has already reached rather than as a neutral judge weighing evidence for the first time. That ordering is why an appeal to the Soviet Union's actual economic history, however well documented, tends to bounce off a committed unconstrained-vision reader the same way an appeal to redlining's actual mechanics tends to bounce off a committed constrained-vision reader who's already decided the state's core function is protective rather than extractive.

Neither reader is being dishonest. Both are doing exactly what Haidt's research says minds normally do — reasoning in service of an intuition already in place, not the other way around — which means the dichotomy this essay opened with isn't a gap in anyone's information. It's a gap in what each side's mind is even built to treat as a live question.

That's what makes the Oppenheimer-George account harder to hold than either of the two it's replacing, not easier, and it's worth being honest about the cost rather than presenting the third view as though it were simply the version with better footnotes. The constrained vision offers the comfort of a system that mostly just needs protecting as it stands. The unconstrained vision offers the comfort of a villain, nameable and finally within reach, whose removal resolves the whole problem in one clean motion.

The third view offers neither kind of comfort. It says the mechanism long predates the current occupant of the villain's chair, will outlast whoever's removed from it next, and requires the same specific, unglamorous work — pricing a rent, closing a loophole, sunsetting a rule — indefinitely, with no final chapter where the work is finished and everyone can stand down.

A mind built the way Haidt describes doesn't reach for that account first, because it doesn't resolve the intuition the way either umpire or weapon does. It just keeps asking the same narrow question forever, which is a much harder thing to organize a movement's psychology around than either "protect what we have" or "seize what they took."

What this means for the caucus, if this is the framework

If this diagnosis is right, it settles a strategic question this site has been circling without naming directly: what is a social liberty caucus actually for, and what should it try to become? And the psychological asymmetry above answers a follow-up question before it gets asked: given that neither temperament reliably updates on evidence, the caucus was never going to win by being the side with the better argument. It has to build fixes that work whether or not anyone's underlying vision ever moves.

It rules out the first strategy a movement usually reaches for by default, which is to treat the state as Locke's umpire and simply try to steer it toward better calls — better regulations, better-credentialed administrators, a more enlightened set of hands on the existing machinery. Why Neither Party Wants Social Liberty already showed why that strategy is a dead end for this specific movement: the administrative capacity a "better umpire" strategy depends on is the same mechanism, run by the more sympathetic coalition, that produces licensing capture and zoning capture when the less sympathetic coalition holds it. An umpire that behaves neutrally only when staffed by people who already agree with you was never actually neutral. It was just currently pointed the right direction, which is not a structural fix and was never going to survive a change of administration.

It rules out, with equal force, the second strategy — organizing the caucus as a faction whose job is to seize the existing machinery of political means and redirect it toward a more deserving beneficiary. That's the Marxist error, restated at whatever scale a caucus operates at: if political means is the mechanism and the goal is simply to put a better class, or a better coalition, in charge of wielding it, the caucus has adopted the exact tool its own diagnosis says can't be trusted to stay pointed the right way once the current occupants move on. A caucus that wins by capturing the state's existing discretionary power has built nothing that survives its own eventual loss of that power — it has just added one more name to the long list of factions who held the weapon for a while.

What the Oppenheimer-George account licenses is narrower and harder, and it's the strategy this site's other essays have already been describing in pieces without naming the unifying logic. The caucus's job isn't to hold the state. It's to permanently close off the specific mechanism of political means wherever it can be found and priced — land rent captured through a land value tax rather than left to accrue to whoever holds title, severance taxes on extracted resource value that was never anyone's labor to begin with, sunset clauses and revisability requirements of the kind Daylight...or Die and A Meadow of Sacred Cows already argued for, structural corrections that keep working regardless of which coalition happens to be in office in a given decade, because they don't depend on that coalition's continued goodwill to keep functioning.

That's also why Growing a Movement as a Design Problem's regional-cell strategy and Why Neither Party Wants Social Liberty's cross-coalition organizing aren't just tactically convenient, on this account — they're the correct shape for a movement whose actual goal isn't to win and hold the state, but to find, region by region and bill by bill, whichever internal faction of whichever party is already closest to pricing a specific instance of political means correctly, and build the structural fix with them before the coalition that benefits from leaving it unpriced reorganizes to stop it.

A caucus organized this way never needs a permanent majority, which is fortunate, because a permanent majority is itself just political means wearing a friendlier flag — a nameable, concentrated group that has arranged never to have to defend its position on the merits again. The goal was never to become the winning side of Marx's fight. It was to make sure the specific mechanism both traditions correctly diagnosed, however differently they explained its origin, has nowhere left to hide once the bill finally gets read aloud in front of the people who've been paying it.

Two doors into the same fix

Building fixes that don't depend on anyone's vision moving is the right structural answer, but it isn't a reason to stop talking to people, and Haidt's own mechanism actually points toward how the conversation should run: not an attempt to swap out someone's vision for a better one, which the psychology above says mostly fails, but an appeal made in the vocabulary the existing vision already trusts, aimed at the specific case where that vision's own standards are being violated.

A constrained-vision reader doesn't need to be talked out of trusting institutions, checks, and the rule of law over any individual's good intentions — that instinct is the whole reason We the People and Splitting the Atom of Sovereignty argue for keeping checks genuinely real in the first place, and a caucus built on this framework needs that instinct as an ally, not a target. What it needs is the evidence Why Neither Party Wants Social Liberty already assembled, put directly in front of exactly this reader: occupational licensing, zoning, the shipping-lane subsidy are not the neutral rule of law functioning as advertised. They're the rule of law having been quietly captured by the concentrated interest closest to writing it, which is precisely the failure mode a constrained vision is supposed to be most alert to. The ask isn't "trust the state less in general." It's "apply the skepticism you already have consistently, including to the arrangements your own coalition currently benefits from."

An unconstrained-vision reader doesn't need to be talked out of noticing that a concentrated group is extracting value from a diffuse one without earning it. What it needs is the Soviet evidence, put just as directly: the villain was never capitalism specifically, it was control of an unproducible scarcity, and every attempt to fix that by seizing the machinery rather than pricing the scarcity has reproduced the same extraction under new management. Transitioning to LVT and Single Tax for the Living World aren't a smaller, compromise version of the redistributive goal this reader already holds. They're a version of it that survives the next election, because it targets the mechanism instead of the current occupant of the chair.

Neither conversation asks its listener to become the other kind of reader, and neither needs to. Growing a Movement as a Design Problem already found the abundance wing inside the Democratic coalition and the fusionist wing inside the Republican one arguing, from their own separate premises, for pieces of exactly this fix. A caucus that understands why those two arguments never converge into one shared vision can still build the same sunset clause, the same land value tax, the same pricing mechanism with both of them — not by resolving Sowell's dichotomy, which isn't this caucus's job to resolve, but by finding, in each vision's own terms, the version of the argument that vision was already equipped to hear.

---

Sources: John Locke, Second Treatise of Government (1689); Jean-Jacques Rousseau, Discourse on Inequality (1755) and The Social Contract (1762); William Godwin, Enquiry Concerning Political Justice (1793); Nicolas de Condorcet, Sketch for a Historical Picture of the Progress of the Human Mind (written 1793–94, published 1795); Karl Marx and Friedrich Engels, The Origin of the Family, Private Property and the State (Engels, 1884, drawing on Marx's notes); Franz Oppenheimer, Der Staat (1908), translated as The State: Its History and Development Viewed Sociologically (1914); Albert Jay Nock, Our Enemy, the State (1935); Henry George, Progress and Poverty (1879); Morris M. Kleiner, research on occupational licensing coverage growth in the U.S. workforce, cited previously in Why Neither Party Wants Social Liberty; Thomas Sowell, A Conflict of Visions: Ideological Origins of Political Struggles (1987); Jonathan Haidt, The Righteous Mind: Why Good People Are Divided by Politics and Religion (2012); We the People; Splitting the Atom of Sovereignty; Growing a Movement as a Design Problem; Transitioning to LVT; Single Tax for the Living World.