The seduction of private law

The most dangerous arguments are not the foolish ones. Foolish arguments announce themselves and are dismissed. The dangerous ones are the arguments that have been polished until only the attractive parts remain, so that what is left looks like a proof when it is in fact an incentive story with half the incentives deleted.

The private-law thesis associated with Morris and Linda Tannehill and with Hans-Hermann Hoppe is exactly such an argument. It does not say, in the adolescent way, that there should be no rules, no courts, no enforcement, and no protection. It says something more sophisticated, and therefore more seductive: that law can be supplied as a market service; that insurers and private defence agencies, because they bear the costs of violence, will be economically driven to prevent it; and that property will be honoured more reliably by firms that profit from order than by a state that profits from disorder. The state, on this account, is a territorial monopoly over final judgement, and like all monopolies it overcharges and underperforms. Replace it with competition and the abuses dissolve. Hoppe places “free and unregulated” insurance agencies at the centre of this competitive order; the Tannehills build it from private protection firms and private arbitration. Both rest the system on a single load-bearing claim: insurers dislike loss, violence is loss, therefore insurers will buy peace.

The claim has a genuine surface plausibility, and the honest way to meet it is not to mock it but to take its strongest form seriously and then ask whether the conclusion actually follows from the premise. It does not. The premise is about incentives. The conclusion is about institutions. Between them lies a gap that the thesis never crosses, and the things that fall into that gap — the definition of property, the finality of judgement, the standing of the poor, the concentration of force — are not minor details. They are the whole of what a legal order is.

This essay argues that the Tannehill–Hoppe thesis mistakes a partial incentive for a complete institutional order, and that its most probable failure mode is not the chaos its critics usually predict. Chaos is the easy objection and the wrong one. The harder and more damaging point is that the thesis is most likely to succeed at producing order, and that the order it produces is plutocracy: a stratified regime in which protection tracks wealth, law becomes a tiered subscription product, and coercive capacity concentrates into an insurance-security-arbitration cartel that is, in everything but name, a state — only a state with no public accountability, no equal standing, and no appeal beyond the price list. The thesis promises the abolition of sovereignty. What it actually offers is the sale of sovereignty as an insurance product. It does not abolish coercion. It changes the billing department.

The thesis at its strongest

To attack a weak version of an argument is a waste of everyone’s time, so the thesis deserves to be stated in the form its ablest defenders would recognise.

Begin with the diagnosis of the state, which is the strongest part of the case and largely correct. The state is a compulsory territorial monopolist of final decision-making. Because it faces no competitor in the supply of ultimate adjudication and coercion within its territory, it has the standard monopolist’s incentives: to raise its price (taxation), to lower its quality (bureaucracy, delay, capture), to expand its jurisdiction, and, most corrosively, to act as judge in its own cause whenever the citizen and the state are themselves the disputing parties. This is not a caricature. It is a serious public-choice observation, and the libertarian tradition from Rothbard through Hoppe has pressed it with real force. A monopoly on justice is still a monopoly, and monopolies are not improved by being called governments.

From this diagnosis the thesis derives its prescription. If monopoly is the disease, competition is the cure. In a private-law order, protection agencies compete to supply defence; insurers compete to underwrite the risks of life, property, and conflict; and arbitration firms compete to resolve disputes. No single agency holds a territorial monopoly, so each is disciplined by the possibility that dissatisfied clients will leave.

The insurer then enters as the mechanism that converts competition into peace. Violence destroys insured property, raises claims, lowers asset values, and threatens portfolios. An insurer therefore has, on this account, a direct financial interest in suppressing violence among and against its clients, in coordinating with other insurers to standardise rules and contain conflict, and in disciplining any protection agency whose aggression generates claims. David Friedman’s version sharpens the economic logic: rights-enforcement and dispute-resolution are services with prices, and a competitive market in such services will, he argues, tend to generate efficient legal rules because the parties to a dispute share an interest in cheap resolution and will gravitate to the arbitration arrangements that minimise the joint cost of conflict.

Reputation and contract are then offered as the safeguard against predation. An aggressive agency that preys on non-clients invites retaliation, loses the custom of those who fear it, finds itself denied access to the arbitration networks every agency needs, and becomes uninsurable. The market, in this telling, ostracises the predator more reliably than any regulator could, because the discipline is automatic and self-interested rather than political and corruptible.

The conclusion is that law survives without the state. It becomes private, contractual, polycentric, and competitive. The courthouse does not vanish; it is unbundled, priced, and sold by firms that want repeat business.

This is the claim in its best clothes. It is not “no law.” It is the claim that market incentives can do the work that public sovereignty now does badly, and do it better. Everything that follows is an attempt to show why the work does not in fact get done — why each of the load-bearing moves smuggles in a premise it has not earned.

Incentive is not institution

The first and most fundamental defect is a non sequitur sitting at the centre of the argument. Strip the rhetoric and the inference reads:

Violence creates losses. Insurers dislike losses. Therefore insurers will produce a just legal order.

The conclusion does not follow, because the operative verb in the premise — “dislike losses” — does not mean what the conclusion needs it to mean. An insurer does not dislike losses in the way a moralist dislikes injustice. An insurer dislikes expected payouts in excess of premium income. Its objective function is the management of a balance sheet: to classify risk, price it, exclude it where it cannot be priced, cap liability, deny weak or expensive claims, settle cheaply, monitor insured behaviour, and protect capital reserves against correlated shocks. These are not vices. They are the legitimate operations of a competently run insurer. But they are operations of portfolio optimisation, and portfolio optimisation is not justice. The two coincide in some cases and diverge sharply in others.

They coincide when suppressing a given act of violence is cheaper than paying for its consequences and the victim is a valued client. They diverge precisely where justice is most needed and least profitable: where the claimant is poor and commercially marginal, where the facts are complex and expensive to establish, where the defendant is wealthy and able to impose litigation costs, where the harm is real but the expected recovery is low, and where doing right by the victim would cost more than abandoning him. In all these cases the insurer’s rational course and the just course point in opposite directions, and the thesis gives us no reason to expect the insurer to choose justice over the balance sheet. It gives us, in fact, every reason to expect the opposite, because the firm that consistently chose justice over its accounts would be outcompeted by the firm that did not.

This is the cleanest way to put the error. The thesis proves, at most, that insurers have an incentive to reduce those forms of violence that threaten insured portfolios by more than the cost of suppressing them. That is a real incentive and it will do real work. It is also a far weaker thing than the thesis needs. It is not an order of justice. It is actuarial housekeeping — the management of loss within a book of business — and it stops exactly where the book of business stops. An insurance company is not a guardian of rights. It is a balance sheet with lawyers, and a balance sheet has no opinion about the dignity of those who cannot afford its premiums.

Property must be defined before it can be honoured

The thesis says that protection agencies and insurers will “honour property.” Tucked inside that verb is the assumption that does the most quiet damage to the entire argument: that property already exists, in determinate and settled form, prior to and independent of the institutions that are supposed merely to guard it. On this picture, titles arrive from somewhere already labelled, and all the enforcement apparatus has to do is stand watch over labels that nature or reason has already affixed.

This is false to the point of being the reverse of the truth. Property is not an object with a tag. It is a structured relation among persons with respect to things — a relation specifying who may use, exclude, transfer, encumber, and bequeath, against whom, on what conditions, and with what priority when claims collide. The hard cases that consume real legal systems are almost never the storybook case of A simply stealing B’s cow. They are disputes about which claim counts: contests over title and the validity of the acquisition chain, over priority between competing creditors, over adverse possession, inheritance, fraud, duress, insolvency, easement, nuisance, trespass, intellectual property, corporate control, fiduciary duty, and the externalities one person’s use imposes on another’s enjoyment. Before any of these can be “honoured,” someone must decide what the property right is.

Ronald Coase’s analysis of social cost is decisive here, and it cuts against the thesis from within economics rather than from sentiment. Coase showed that harm is reciprocal: the factory’s smoke harms the laundry, but forbidding the smoke harms the factory, and there is no natural fact of the matter about whose use should yield. With zero transaction costs the initial assignment of rights would not affect the efficient outcome, but transaction costs are never zero, and where they are positive the assignment of rights determines both the outcome and its efficiency. The implication is fatal to the idea that enforcement is logically downstream of an already-given property order: the content of the rights, and the transaction-cost environment in which they are exercised, are doing the real work, and someone must set them. Harold Demsetz’s account of how property rights emerge points the same way. Rights are not primitives; they are institutional responses that develop to internalise externalities when the gains from doing so exceed the costs of definition and enforcement. Property is historically contingent, evolved, and constructed — which is to say that defining it is not a preliminary to the legal order but the substance of it.

The decision about which property theory to enforce is therefore unavoidable, and it is political in the broad and exact sense that it allocates rights, settles priorities, and determines whose expectations the system will protect and whose it will sacrifice. A private agency that proposes to “honour property” cannot do so without first choosing a theory of property, and that choice is a legislative act however it is dressed. Calling the agency private does not make the act non-political. It merely relocates the politics into a firm that answers to its paying clients rather than to a public. The thesis assumes the settled property order it is supposed to derive, and then uses the assumed order to claim that agencies merely enforce property rather than make law. That is question-begging of the most consequential kind.

The circularity: markets presuppose the law they are said to supply

The deepest problem is structural, and it is best stated plainly because its very simplicity is what makes it hard to evade. Markets do not run on air. They run on a dense substrate of legal infrastructure: rules of property, contract, agency, fraud, evidence, title, priority, insolvency, liability, and corporate form, together with procedures for adjudicating disputes about all of these and mechanisms for enforcing the results. Strip that substrate away and there is no market — only a collection of people making assertions at one another.

The fully private-law thesis claims that the market will supply this infrastructure. But the market requires the infrastructure in order to function at all. The result is a circle the thesis never escapes: the market is said to produce law, yet the market presupposes law in order to produce anything. This is not a quibble about sequence. It goes to whether the system can get started.

Consider what each of the thesis’s own mechanisms requires. A contract is not self-enforcing; it is a promise that becomes binding only because some authority will compel performance or its equivalent, and “the market will compel it” only pushes the question back, because the compelling itself requires rules of formation, interpretation, breach, fraud, duress, and remedy that someone must have established. An arbitration award is not a judgement; it is an opinion, advice with no force, unless there exists an enforcement structure standing behind it that can make the loser comply when the loser would rather not. An insurance obligation is a string of words unless there is law determining when the policy is formed, what it covers, when it is breached, and what happens when the insurer alleges fraud or the insured alleges bad faith. A title record is ink unless there is a recognised rule — public authority or accepted network protocol — that gives the record legal effect against third parties who never agreed to it.

The thesis gestures at reputation, arbitration, and inter-insurer coordination as the substrate’s substitutes, and these mechanisms do real work in the easy cases: among repeat players, with visible facts, low stakes, and rough equality of power, private ordering can and does function without a state standing over every transaction. The error is the leap from “private ordering works in these conditions” to “private ordering can constitute the entire legal order for all persons, all disputes, and all stakes.” That is a fallacy of composition — the inference that because a part has a property the whole must have it too — and it fails for the same reason it always fails: the conditions that make private ordering work in the easy cases are precisely the conditions that the hard cases violate. Reputation disciplines those who need to deal again tomorrow; it does not bind the party for whom this dispute is the last move, or the party powerful enough that no one can afford to refuse him tomorrow regardless of his conduct today. The thesis cannot conjure a legal order out of mechanisms that presuppose one, and waving the word “competition” over the gap does not fill it.

The final-authority problem: feud or sovereignty

Every legal order must eventually answer one question, and it is the question on which the private-law thesis breaks. What happens when two parties disagree, each is genuinely convinced of his right, each is backed by a protection agency, and neither will submit?

The standard answer is arbitration. But arbitration presupposes either prior agreement on a forum or an enforcement structure able to impose the forum’s decision, and in the hard case at least one party has every incentive to refuse both. When the stakes are high enough, the facts contested enough, and the parties unequal enough, the losing side — or the side that anticipates losing — does not consent to the process that will rule against it. At that point the order arrives at a fork from which there is no third path.

Either there exists a final authority with the capacity to impose the outcome on the recalcitrant party whether or not he agrees — in which case sovereignty has reappeared under a new logo, and the entire promise to abolish the territorial monopolist of final decision has been quietly broken. Or there is no such authority — in which case the dispute is resolved not by law but by bargaining under threat, by the relative coercive capacity of the two agencies, by escalation, cartel negotiation, or force. The first outcome is the state, rebranded. The second is the feud.

This is not an incidental weakness; it is structural, and it has been recognised even by thinkers sympathetic to the libertarian project. The private-law model oscillates permanently between the two outcomes and can rest at neither. Centralise enforcement enough to bind the recalcitrant, and you have recreated the sovereign you set out to abolish, now without the constitutional constraints, the public accountability, and the formal equality before the law that even a flawed state at least professes. Decline to centralise it, and you have no answer to the party who simply will not comply, which means you have no law in the hard cases — only negotiated coexistence among armed firms, stable when their interests align and violent when they do not. Private law either becomes law, in which case it owes us an account of the sovereignty it claimed to have escaped, or it remains genuinely private, in which case it cannot bind the one who refuses to be bound. There is no version that is both fully private and fully binding, and a legal order that cannot bind in the hard case is not a legal order. It is a market with weapons.

What insurance actually does

Because the thesis rests the whole apparatus on the insurer, the actual economics of insurance are not a side issue; they are the heart of the matter, and they are unkind to the argument. Insurance is not a benign solvent that dissolves conflict into priced and managed risk. It is a specific institution with well-documented structural features, and each of those features, applied to the supply of law, points toward exclusion rather than universality.

The first is adverse selection. Those most likely to suffer a loss are most likely to seek cover, which means the pool an insurer attracts is systematically worse than the population unless the insurer screens it. So insurers screen. They classify applicants, price by risk, cap exposure, attach deductibles and exclusions, and refuse the risks they cannot price profitably. A market in legal protection subject to adverse selection does not extend equal cover to all; it sorts people into risk classes and prices them accordingly, and some classes it declines outright.

The second is moral hazard. The insured party, shielded from the consequences of his conduct, has weaker incentives to avoid loss, so the insurer must monitor and constrain behaviour, attaching conditions, surveillance, and controls to the policy. Translate this into a protection market and the insurer’s rational response to moral hazard is the surveillance and behavioural regulation of its own clients — a private disciplinary apparatus operated not for justice but for loss control.

The third, and the most damaging to the thesis, is correlated risk. Insurance functions on the law of large numbers, which requires that losses be substantially independent or at least diversifiable, so that the many who do not suffer loss can fund the few who do. But the losses that a legal order exists to address are precisely the ones that are not independent. Civil disorder, organised crime, insurrection, war, mass unrest, territorial conflict, financial collapse, and the breakdown of the legal order itself are correlated risks: they strike many policyholders at once, and they are the systemic events against which insurance is structurally weakest. An insurer pricing household burglary is doing ordinary actuarial work. An insurer asked to underwrite the stability of the legal order is being asked to price sovereignty failure, and against correlated catastrophe an insurer has only three options: withdraw cover, raise premiums beyond the reach of ordinary people, or build coercive infrastructure large enough to control the systemic risk directly rather than merely insure against it. The third option is the road to private government, and it is the option a sufficiently large and rational insurer will take, because controlling the risk is cheaper than paying for it.

The fourth is the claims-denial incentive, which the thesis politely ignores. Insurers do not simply pay. They investigate, contest, delay, narrow coverage, dispute causation, allege misrepresentation, and settle for less than the claim where they can. These are not abuses; they are the normal management of a book, and they are sometimes the difference between solvency and ruin. But “the body that decides whether your claim is honoured has a direct financial interest in denying it” is a description of a profound structural conflict, and it is the conflict the thesis proposes to install at the centre of justice. Underwriting is not standing. Claims management is not due process. Risk exclusion is not universal protection. Settlement pressure is not legal equality. The thesis treats the insurer as a neutral keeper of the peace; the institution is in fact a sorter, a denier, and a manager of correlated catastrophe, and none of those functions is justice.

Rights at the price of coverage

Here the plutocracy argument becomes explicit, and it is an institutional argument rather than a sentimental one. The objection is not that rich people will become villains. It is that, when enforcement is a commodity, the enforceable value of a right comes to track the holder’s ability to pay for it — and that this is not an abuse of the system but its design.

In a public legal order, however imperfectly it lives up to the claim, a person’s right is not supposed to depend entirely on his current purchasing power. The principle is that the right inheres in the person and the law stands behind it regardless of his bank balance. The practice falls far short: the rich already obtain better lawyers, faster service, and more favourable outcomes, and the gap between formal and effective access to justice is one of the genuine indictments of existing legal systems. But the private-law thesis does not solve this defect. It constitutionalises it. It takes the contingent, regrettable, fought-over fact that wealth buys better law and makes it the explicit organising principle of the entire order.

Trace the mechanism. In a fully private market for protection, the wealthy buy comprehensive cover, elite representation, rapid intervention, privileged access to the best arbitration, and strong enforcement. Those of moderate means buy thinner cover, slower service, weaker representation, and access to whatever forums their budget reaches. The poor buy little. The destitute buy nothing, and a right that no one can be made to enforce is not, in any operative sense, a right at all — it is a sentiment with no remedy. The poor do not lose their rights by any declaration or decree. They lose them by under-insurance, which is quieter and harder to protest and therefore more durable.

The point sharpens further once we see that in this model the poor are not merely poorer consumers of law; they are worse legal risks, and rational firms treat them accordingly. The poor have fewer assets to seize, which makes them low-value claimants; they have less capacity to fund a counterclaim or endure a long dispute, which makes them weak defendants; they have little reputational leverage and less ability to take their custom elsewhere, which makes them captive. A rational insurer or protection agency therefore has reason to treat the poor as commercially marginal — high-cost, low-recovery, easily abandoned — which means the poor are not simply served worse but rendered progressively invisible to the enforcement system, present in it only as risks to be priced out rather than persons to be protected. When enforceable right is a function of purchasing power, inequality of wealth becomes inequality of law, and inequality of law with the full backing of force is the precise institutional meaning of plutocracy. Legal standing ceases to be civic and becomes actuarial. That is the substitution the thesis performs while presenting itself as a charter of liberty.

The economics of coercion: scale, concentration, and the dominant agency

The thesis depends on a market that stays competitive — on the persistence of many agencies, none dominant, each disciplined by the others. But coercion is not an ordinary good, and a market whose product is force does not behave like a market in sandwiches. It has scale economies that drive it toward concentration, and concentration in the supply of force is not a large firm. It is a sovereign.

The advantages of size in the production of coercion compound in a way they do not for ordinary goods. A larger protection-insurance firm commands better intelligence, more armed personnel, deeper databases, superior surveillance, stronger legal drafting, larger capital reserves against correlated shocks, and greater bargaining leverage over everyone it deals with. Crucially, it can threaten a smaller agency far more credibly than the smaller agency can threaten it, and it can impose its terms on the arbitration networks that every agency needs, refusing to deal with firms that will not accept its rules. Each of these advantages raises the return to being larger, and the firm that is larger is better placed to become larger still. The competitive equilibrium the thesis assumes is not stable, because the product itself rewards the accumulation of the capacity to coerce.

This is not a hostile invention; it is the conclusion reached by the libertarian tradition’s own most careful philosopher. Robert Nozick, setting out to defend the minimal state to anarchists on their own terms, argued that competing protection agencies would not remain plural. Through an invisible-hand process driven by nothing more sinister than each client’s preference for the agency most likely to prevail in a conflict with clients of other agencies, one agency becomes dominant within a territory; it then suppresses, by force or by buying out, the risky private enforcement of independents, and arrives at a de facto monopoly on the legitimate use of force — which is to say, at a state. The significance for the present argument is that even a thinker determined to vindicate liberty against the anarchist concluded that a market in protection does not stay a market. It collapses into a monopoly. Nozick called the result a minimal state and thought it justified; whether or not one accepts his normative conclusion, his structural prediction is the one that matters here, and it is the opposite of the Tannehill–Hoppe assumption that competition persists.

Tyler Cowen pressed the point from inside the economics of the question and reached a conclusion still more uncomfortable for the thesis. The polycentric system requires a network: agencies must recognise one another’s arbitration, honour one another’s judgements, and coordinate enforcement, because without such a network there is no system, only a scatter of firms unable to resolve cross-agency disputes. But a network with the power to admit and exclude is a network with the power to collude. The very coordination that makes private law function — the shared protocols, the mutual recognition, the interlocking arbitration — is the coordination that allows the established agencies to act as a cartel: to fix terms, to exclude entrants, and to discipline mavericks, exactly as the dominant firm would in any other industry, except that here the industry’s product is force. The mechanism that the thesis needs for the system to work is the mechanism that turns the system into a coercive monopoly. Hoppe himself anticipates that private law would tend toward a unification of law through inter-insurer agreement; he presents this as a benign convergence on good rules, but a unified, coordinated, network-enforced body of law backed by concentrated force is not the abolition of the state. It is the state’s functional re-creation by a cartel, with the added defect that the cartel was never even nominally accountable to anyone but its members. The state does not disappear in this story. It is reconstituted as a premium-funded enforcement conglomerate, and the only thing genuinely abolished is the public’s claim on it.

The natural equilibrium: limited-access order and elite rent-sharing

If the supply of coercion concentrates, the next question is what kind of order the concentration produces, and the political-economy literature gives an answer that is more precise and more damning than the usual prediction of collapse. The likely end-state of a private-coercion order is not Hobbesian chaos. It is a stable, ordered hierarchy in which violence is contained by sharing its proceeds among those capable of deploying it — and that hierarchy has a name.

Douglass North, John Wallis, and Barry Weingast argue that the overwhelmingly dominant way human societies have controlled violence is not the rule of law but the limited-access order, or “natural state,” in which powerful actors with the capacity for violence are given privileged access to economic rents — to land, trade, offices, and organisations — on the condition that they refrain from fighting. The logic is that open violence destroys the rents, so the elites who hold them have a shared interest in keeping the peace among themselves, and the order is stabilised precisely by restricting access: by limiting who may form organisations, who may compete, who may enter the privileged circle. The result is genuine order, often durable, but it is not equal, not open, and not liberal. The few coordinate to preserve a profitable peace; the many receive whatever standing the coordination finds convenient to grant them. The rare alternative — the open-access order, with impersonal rule of law, open economic and political competition, and entry on equal terms — is historically exceptional and rests on demanding conditions that do not arise automatically and are not the default toward which societies drift.

This framework maps onto the private-law failure mode with uncomfortable exactness. The actors a private-coercion order would empower — large insurers, dominant security firms, major creditors, landlords, infrastructure owners, and the arbitral institutions that serve them — are exactly the actors who, in the North–Wallis–Weingast account, form the coalition of a natural state. They need no conspiracy and no smoke-filled room; shared incentives are sufficient. Each prefers predictable titles, enforceable debts, low disruption, stable returns, and barriers against the low-cost entrants, defaulting debtors, squatters, strikers, and inconvenient outsiders who threaten the value of what they hold. A coalition of such actors, coordinating to preserve a profitable peace and to restrict access to those who might disturb it, is the textbook structure of a limited-access order. The private-law society does not collapse; it congeals — into precisely the rent-sharing elite settlement that most of human history has produced whenever the capacity for violence was privately held.

Mancur Olson’s analysis explains why concentration is the attractor rather than an accident. Under genuine anarchy, a “roving bandit” who can plunder and move on has no reason to leave his victims enough to prosper, so roving predation is ruinous for everyone. But a “stationary bandit” who monopolises theft within a territory acquires an encompassing interest in the productivity of those he robs: because he will be there to rob them again next year, he has reason to leave them enough to invest, to provide the order that lets them produce, and to rationalise his extraction as regular and predictable taxation rather than random seizure. Olson’s startling implication is that monopolised coercion can be Pareto-superior to competitive predation — which is exactly why a dominant private-protection firm has every incentive to become the stationary bandit: to stabilise its extraction by monopolising violence, to provide order because order is what makes its clients worth protecting, and to call the regular fees premiums and the protection a service. The economic logic that built states out of anarchy in the first place does not pause for the word “private.” It runs in a market for protection exactly as it ran everywhere else, and it runs toward monopoly.

Protection and the racket

There is an empirical version of all this, and it is the version the thesis most needs to avoid, because the historical cases in which protection was genuinely supplied by private firms in the absence of an effective state are not flattering. The clearest record we have of privately produced protection is the record of the mafia, and that is not an accident of vocabulary.

Charles Tilly placed the point at the level of theory. War-making, state-making, protection, and extraction, he argued, sit on a single continuum, and the difference between a government and a protection racket is far less principled than the government would have us believe. Both sell protection; both sometimes manufacture the very threats they then charge to defend against; both extract payment under the implicit understanding that refusal carries a cost. Legitimacy, in Tilly’s deflationary account, is largely retrospective — a story told after the fact by those who won the competition to control violence in a territory. The relevance to the private-law thesis is direct: the thesis proposes to launder the protection–extraction continuum by inserting the word “insurance,” but inserting a word does not change the structure. A firm that says “pay us and we will protect you” and a racket that says “pay us and we will protect you, including from ourselves” differ not in their offer but in the constraints upon them — in whether there is legality, accountability, contestability, exit, and due process limiting the coercion. And in a fully private order, those constraints are supplied, if at all, by the same enforcement ecosystem that profits from coercion. The distinction between protection and racket becomes endogenous to the firms that have the most interest in erasing it.

Diego Gambetta’s study of the Sicilian mafia supplies the empirical flesh and removes any comfort the thesis might draw from the abstractness of the objection. The mafia, Gambetta showed, is best understood not as a criminal aberration but as an industry: an industry supplying private protection and the guarantee of transactions in a society where the state failed to provide reliable enforcement and trust. Where public protection of property and contract was weak, a market for private protection emerged to fill the gap — and what filled it was not a competitive ecology of mutually disciplining firms converging on liberal law. It was a violent, territorial, rent-extracting hierarchy that guarded the property of those who paid, settled disputes on its own terms, suppressed rivals, and treated outsiders as objects rather than clients. Federico Varese’s parallel study of the Russian mafia in the chaotic aftermath of the Soviet collapse found the same pattern: when state enforcement disintegrated and a sudden demand for the protection of new property rights went unmet by any public authority, private protection markets arose — and again they took the form of mafias, not of the benign insurer-arbitrators the thesis imagines. The point must be made carefully, because the careless version is false: not every private insurer is a mafia, and the claim is not that private protection is always criminal. The claim is narrower and harder to dismiss. The thesis owes us a mechanism that prevents a market in protection from becoming a market in protection rackets, and the only mechanism it offers is competition — but coercive competition, the competition to supply force, is exactly the thing that the historical cases show produces rackets rather than restrains them. The one large-scale natural experiment we have in privately produced protection returned the mafia. That is the evidence, and the thesis has no answer to it beyond the assertion that this time the firms would behave differently.

Why reputation and ostracism fail in the hard cases

The thesis’s last line of defence is reputation. Aggressive agencies, the Tannehills argue, would be ostracised — boycotted by clients, denied arbitration, made uninsurable — and so the market would discipline predators without any public authority. The mechanism is real, and the question is not whether it ever works but whether it works in the cases that matter, and there the answer is no.

Reputation disciplines a firm under a specific and demanding set of conditions: many competitors, low switching costs, transparent facts, rough equality of bargaining power, reliable information about conduct, and the absence of any agency dominant enough to ignore the others’ disapproval. These are the conditions of the easy case, and in the easy case the firm did not need much disciplining to begin with. The hard case — the case the legal order exists for — violates every one of them. There the facts are contested and expensive to establish, so the predation may never become common knowledge; switching costs are high, so dissatisfied clients are captive rather than mobile; bargaining power is grossly unequal, so the victims are exactly the parties least able to organise a boycott; and, decisively, the predatory agency is often the dominant one, serving the wealthy clients, controlling the critical infrastructure, and commanding the superior force.

Against a dominant agency, ostracism is theatre. The parties with the capacity to discipline a powerful protection firm — the other large firms, the major asset holders, the creditors — are precisely the parties who benefit from its protection and have no interest in disciplining it, while the parties with the interest — the poor, the marginal, the preyed-upon — are precisely the parties without the leverage. Reputation works against the small and dependent firm that needs everyone’s goodwill to survive; it does not work against the firm large enough that others need its goodwill more than it needs theirs. Exit disciplines a seller only when exit is real, and for the captive client of a dominant coercive firm exit is not real. Under those conditions reputation ceases to be a constraint and becomes a brand — a thing the powerful firm cultivates precisely so that its coercion may be called service. The ostracism argument assumes away the concentration that the economics of coercion produces, and an argument that depends on assuming away the central problem is not an answer to it.

Peace is not liberty

It is worth pausing on a conceptual confusion that runs beneath the whole thesis, because once it is named, much of the argument’s appeal dissolves. The thesis treats the production of peace as though it were the production of liberty, as though a demonstration that insurers prefer order were a demonstration that they would supply a free society. But peace and liberty are not the same thing, and the gap between them is where every illiberal order in history has made its home.

A cartel can produce peace. A feudal manor can produce peace. A dictatorship can produce peace, often a very orderly one. A prison produces peace. A plantation produced peace. A mafia, where it is unchallenged, produces a notably effective peace. Peace — the absence of open violence — is compatible with almost any distribution of power and almost any degree of subordination, and the more total the domination, frequently the deeper the peace. So a proof that a private-law order would tend toward peace, even if it were sound, would establish far less than the thesis supposes. It would establish that the order is quiet, not that it is free.

The questions liberty actually turns on are the ones the thesis does not ask: peace under what rules, by whose consent, with what right of appeal, with what possibility of exit, with what equality of standing before the law, with what protection for those too weak to defend themselves, and with what limits on the very firms that enforce the rules. A plutocratic order answers all of these in a particular and unfree way. It supplies a very specific peace: stable titles for incumbents, low disruption for capital, predictable enforcement for creditors, and quiet subordination for those without bargaining power. That is exactly the peace an insurance-security-arbitration cartel would have every reason to supply, and exactly the peace the historical limited-access order has supplied for most of human history. To prefer this peace is not to prefer liberty. The thesis proves, at the very most, an incentive toward a certain kind of order, and then helps itself to the word “freedom” as though order and freedom were synonyms. They are not, and the difference is the entire subject.

The honest counterargument: non-state governance can work

The critique made so far would be cheap if it amounted to the claim that nothing works without the state, because that claim is false, and the strongest non-statist scholarship demolishes it. Intellectual honesty requires meeting that scholarship squarely rather than hiding behind a strawman, and the most serious version of it is Elinor Ostrom’s.

Ostrom’s work on common-pool resources is the most rigorous demonstration available that communities can govern themselves, resolve disputes, and sustain cooperation over long periods without either privatisation or an external state. Studying irrigation systems, fisheries, forests, and grazing commons that endured for centuries, she identified a set of design principles that long-surviving self-governing institutions tend to share: clearly defined boundaries around the resource and the community; rules matched to local conditions; collective-choice arrangements that let those affected by the rules participate in making them; effective monitoring, often by the users themselves; graduated sanctions that escalate with the severity and repetition of violations; accessible and low-cost conflict-resolution mechanisms; at least minimal recognition by external authorities of the community’s right to organise; and, for larger systems, nested layers of governance. Her achievement was to refute decisively the lazy dichotomy that the only choices are state control or privatisation, and to show that polycentric, self-organised governance is a real and sometimes superior third path. Any honest critic of the private-law thesis must concede this, and it is conceded here without reservation: the state is not the only alternative to the market, and non-state governance is not a fantasy.

But the concession does not rescue the thesis; it locates precisely where the thesis fails, and Ostrom’s own findings are what locate it. The institutions she documents work because of their conditions, and those conditions are the inverse of the conditions a society-wide private-coercion order would face. Ostrom’s commons are bounded: a defined community of identifiable members who interact repeatedly, can observe one another’s conduct, share a stake in the resource’s survival, and possess local legitimacy and the means to monitor and sanction at low cost. They govern a common resource among known neighbours, not the totality of high-stakes disputes among strangers across a whole society. They rely on graduated, communally administered sanctions, not on the deployment of organised coercive force against the recalcitrant powerful. And, tellingly, several of her principles presuppose exactly the thing the private-law thesis denies — minimal recognition by an external authority, and nested governance that includes higher tiers — which is to say that Ostrom’s successful self-governance is typically self-governance within a broader legal order, not a replacement for one. The Tannehill–Hoppe thesis takes the genuine fact that bounded communities can govern shared resources among repeat-dealing neighbours and inflates it into the entirely different claim that a market can supply the complete coercive legal order for an unbounded society of strangers, across every dispute, every wealth level, and every act of violence. That is the fallacy of composition again, in its most consequential form. Private ordering can supplement a legal order; Ostrom shows as much. It does not follow that private ordering can constitute one, and Ostrom’s conditions are precisely the evidence that it cannot scale to do so. The honest reading of the strongest non-statist authority is therefore not a vindication of the thesis but a specification of its limits.

The state is bad; private plutocracy is not the cure

None of this is a defence of the state as it exists, and it is important to say so plainly, because the thesis draws much of its energy from the real and serious failures of public legal orders. Those failures are not in dispute. State legal systems are slow, costly, and bureaucratic; they are captured by organised interests; they are unequal in practice however equal in principle; they are sometimes corrupt and sometimes abusive; police forces can become predatory; regulation can ossify into a shield for incumbents; and the public-choice critique of the state’s incentives is largely correct. The libertarian diagnosis with which this essay began is not a setup to be knocked down. It is true.

But the truth of the diagnosis does not establish the soundness of the cure, and here the thesis commits its most basic methodological error: it compares the real state, with all its documented defects, against an idealised market, with its defects assumed away. That comparison is rigged. The honest comparison is between real public authority, defects and all, and real private coercion, with its defects — the concentration, the cartelisation, the exclusion of the poor, the conflict of interest at the heart of claims, the historical tendency to produce mafias and limited-access orders — fully in view. Set the real against the real, and the private-law order does not emerge as the freer of the two. It emerges as the older and uglier one: private sovereignty for those who can afford it, which is the arrangement the rule of law was painfully constructed to replace.

There is, moreover, a coherent alternative to both horns of the false dichotomy the thesis offers, and it is the tradition of constitutional political economy the thesis ignores. James Buchanan and Gordon Tullock distinguished the constitutional level, at which the rules of the game are chosen, from the operational level, at which actors play within them, and argued that the remedy for the abuse of collective power is not the abolition of public authority but its constraint — the binding of the sovereign by rules agreed at the constitutional level, designed to hold whoever happens to wield power later. The same impulse runs through Acemoglu and Robinson’s distinction between inclusive institutions, which distribute power broadly and open access to opportunity, and extractive institutions, which concentrate power and channel resources to a narrow elite — with the historical record showing that prosperity and freedom track the inclusive form and that concentrated, unaccountable power, whether public or private, tends toward extraction. The remedy for a coercive monopoly that overcharges and abuses is to constrain it: to subject it to constitutional limits, transparency, division of power, independent appeal, due process, equal standing, legal aid, and competition where competition is genuinely possible. The remedy is emphatically not to auction the coercive function to whoever can pay the most for it, because an unconstrained coercive power sold to the highest bidder is not less dangerous than an unconstrained coercive power held by the state. It is more dangerous, because it has shed even the pretence of public accountability and the formal commitment to equality before the law that gives the citizen a standpoint from which to demand better. The cure for bad public power is better-constrained public power. It is not private power, unconstrained and for sale.

Conclusion: the actuarial state

The Tannehill–Hoppe thesis begins from a genuine insight and ends in an illusion. The insight is that incentives matter, that a monopoly on justice is still a monopoly, and that the state’s incentives are often as bad as its defenders are reluctant to admit. The illusion is the inference that because insurers dislike loss, insurers will supply justice — that a single incentive, isolated and idealised, can stand in for an entire civilisation’s worth of institutions.

It cannot, and the reasons compound. Insurers minimise expected liabilities, not injustice, and the two part company exactly where justice costs more than abandonment. Property cannot merely be honoured, because it must first be defined, and defining it is a political act that allocates rights and cannot be performed by a firm without that firm legislating for its clients. The market cannot supply the legal substrate it presupposes, on pain of a circularity it never escapes. Arbitration cannot bind the recalcitrant in the hard case without an enforcement structure that is sovereignty under another name, leaving the order to oscillate forever between feud and a rebranded state. Coercion is not an ordinary good; it has scale economies that drive it toward concentration, as the libertarian tradition’s own philosopher and its own economists were honest enough to conclude. Concentration plus elite coordination is not chaos but a limited-access order — the rent-sharing settlement that most of human history has produced whenever violence was privately held. The one large natural experiment in privately produced protection returned the mafia. And reputation, the thesis’s last defence, disciplines the small and dependent firm while leaving the dominant one untouched, which is the only firm that needed disciplining.

The more plausible end-state of the private-law project is therefore not a society without sovereignty. It is sovereignty without accountability: a private order in which coercion is priced, law is bundled into coverage tiers, and justice is delivered according to the level of protection a person can afford. The poor are not stripped of rights by any decree; they are priced out of them, which is quieter and lasts longer. The wealthy are not granted privileges by any statute; they buy them, which sounds like freedom and functions like aristocracy. This is the actuarial state — a regime that has all the coercive substance of the thing it claimed to abolish and none of the public claims that even a flawed republic concedes to its citizens.

The state is not abolished when the courthouse is moved into the insurance office. The monopoly on final judgement is not dissolved when it is reconstituted as a cartel of firms that answer only to their shareholders. The thesis does not abolish coercion; it changes the billing department. And the order it most likely produces is not liberty against the state. It is the state behind a paywall.

Works Cited

Acemoglu, D., & Robinson, J. A. (2012). Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Crown.

Buchanan, J. M., & Tullock, G. (1962). The Calculus of Consent: Logical Foundations of Constitutional Democracy. University of Michigan Press.

Coase, R. H. (1960). The Problem of Social Cost. Journal of Law and Economics, 3, 1–44.

Cowen, T. (1992). Law as a Public Good: The Economics of Anarchy. Economics and Philosophy, 8(2), 249–267.

Demsetz, H. (1967). Toward a Theory of Property Rights. American Economic Review, 57(2) (Papers and Proceedings), 347–359.

Friedman, D. D. (1973). The Machinery of Freedom: Guide to a Radical Capitalism. Harper & Row.

Gambetta, D. (1993). The Sicilian Mafia: The Business of Private Protection. Harvard University Press.

Hoppe, H.-H. (1999). The Private Production of Defense. Journal of Libertarian Studies, 14(1), 27–52. (Also issued as a monograph by the Ludwig von Mises Institute.)

North, D. C., Wallis, J. J., & Weingast, B. R. (2009). Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History. Cambridge University Press.

Nozick, R. (1974). Anarchy, State, and Utopia. Basic Books.

Olson, M. (1993). Dictatorship, Democracy, and Development. American Political Science Review, 87(3), 567–576.

Olson, M. (2000). Power and Prosperity: Outgrowing Communist and Capitalist Dictatorships. Basic Books.

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

Rothbard, M. N. (1970). Power and Market: Government and the Economy. Institute for Humane Studies.

Rothbard, M. N. (1973). For a New Liberty: The Libertarian Manifesto. Macmillan.

Tannehill, M., & Tannehill, L. (1970). The Market for Liberty. Self-published (later reprinted, Laissez Faire Books / Ludwig von Mises Institute).

Tilly, C. (1985). War Making and State Making as Organized Crime. In P. Evans, D. Rueschemeyer, & T. Skocpol (Eds.), Bringing the State Back In (pp. 169–191). Cambridge University Press.

Varese, F. (2001). The Russian Mafia: Private Protection in a New Market Economy. Oxford University Press.

Williamson, O. E. (1985). The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. Free Press.

Cash is built to forget. A banknote carries no memory of who held it last, no record of how it was earned, no ledger of the hands it passed through. That amnesia is not a defect. It is the point. A currency that remembered everything would not circulate; it would litigate. Every shopkeeper would become a title investigator, every transaction a due-diligence exercise, every tenner a small lawsuit waiting to happen. Money works precisely because it is allowed to travel without a biography.

And yet the law sometimes needs the biography back. It needs to know who owned the notes, who took them, who paid them away, whether they were stolen, mistaken, mixed, or derived from crime, and whether the person now holding them knew any of that. The whole legal architecture of money is a negotiation between these two demands: the demand that cash forget enough of its past to circulate, and the demand that it not forget so much that theft, fraud, and criminal proceeds become legally purified by movement.

This essay is about that negotiation, and about an argument that the negotiation does not stop at the edge of the physical world. The fashionable claim of the last fifteen years — that distributed ledgers create a new kind of money beyond the reach of courts, that “code is law” — is the old cash problem dressed in new clothes. It fails for the same reason the strong version of cash-finality fails. Bearer finality has never been absolute. It is a legal tolerance, extended because ordinary exchange needs it, and that tolerance shrinks as value, opacity, and systemic risk grow. Blockchains do not escape this. They relocate it.

Three claims organise what follows. First, that cash is not a thing but an institution: a legally tolerated degree of bearer finality, not an unlimited right to move value without accountability. Second, that this tolerance is scalar — a five-dollar coffee and a ten-million-dollar suitcase are both “cash” in the trivial sense that the notes remain currency, but they are not the same institution, and the law does not treat them as such. Third, that digital systems inherit this structure wholesale: a smart contract can move a token, but it cannot decide whether the transfer was authorised, lawful, voidable, fraudulent, or subject to a court order. Code may settle the ledger. Law settles the rights.

What cash is, and what it is not

Begin with the ordinary categories: notes and coins, denominated in a sovereign unit of account. Most people think they know what “legal tender” means, and most people are wrong. In England and Wales, Bank of England notes and Royal Mint coins are legal tender, but legal tender does not oblige a shop to accept them. As the Bank of England’s own public guidance explains, the concept is much narrower than the folk understanding: it bears on the discharge of debts, not on whether a retailer must take your cash for a sandwich. A creditor who is properly tendered legal tender in satisfaction of a debt, absent a contrary contractual term, cannot then succeed in a claim for non-payment. That is what legal tender does. It does not make cash compulsory; it makes a proper tender a defence.

Four concepts get collapsed here that ought to be kept apart: money, cash, legal tender, and payment method. A debit card, a cheque, a contactless tap — these are ordinary payment methods, and none of them is legal tender. The distinction is not pedantry. It matters enormously for recovery, because the route by which value can be reclaimed depends entirely on what kind of thing the claimant is chasing: physical notes, a credit balance in a bank account, an electronic payment instruction, a crypto-token, or a contractual obligation to pay. Each travels through a different part of the law.

So what is cash, legally? It is a bearer asset with currency status — property in the hands of the holder, but property that also performs a public monetary function, and whose ordinary property rules are modified precisely because of that function. A watch or a painting carries its title history with it: a thief cannot generally pass good title to stolen goods, and the true owner can usually recover them or sue in conversion. Cash is different. It is meant to pass from hand to hand without inquiry into where it has been.

The pivot case is more than two and a half centuries old. In Miller v Race (1758), a stolen bank note made its way, through an honest chain, into the hands of someone who had given value for it without notice of the theft. Lord Mansfield held that the note was to be treated as currency, and that an action in the nature of conversion would not lie, because there is no property in currency once it has passed. His formulation has never been bettered: where money is stolen, the true owner cannot recover it once it has been “paid away fairly and honestly upon a valuable and bona fide consideration”; but before money has passed in currency, an action may be brought for the money itself. Commerce, Mansfield understood, could not function if every recipient of a banknote had to reconstruct its ownership history. The law therefore made a deliberate sacrifice: it gave up the owner’s claim against the honest recipient in order to keep money moving.

That sacrifice is the load-bearing idea. Cash is not merely a thing. It is a thing whose legal usefulness depends on being treated as final in ordinary circulation.

Cash as tolerated bearer finality

Here is where the argument turns. Miller v Race is usually read as a statement about what cash is. It is better read as a statement about what the law is willing to tolerate. Finality is not a natural property of banknotes; it is a legal concession, granted because the alternative — perpetual recoverability — would destroy the very liquidity that makes cash worth having.

And concessions have conditions. The tolerance is at its strongest in ordinary retail and private exchange, where the social benefit of frictionless payment plainly exceeds the risk of concealed wrongdoing. It weakens — sharply — where the amount is abnormal, the context is suspicious, the transaction is structured to dodge scrutiny, the recipient is a regulated institution, the money is held rather than spent, or the value is linked to crime. The notes do not change. The institutional treatment does.

This is the home of the distinction that runs through the rest of the essay. A five-hundred-dollar cash payment for goods may be cash in the full economic and legal sense: ordinary, final, private, and practically uninvestigated. A ten-million-dollar transfer in physical banknotes is not, in any meaningful sense, “just cash.” It is a concentrated bearer-value event. It raises questions of source, purpose, beneficial ownership, tax, sanctions, laundering, coercion, and public order. The banknotes remain currency. The transaction no longer enjoys the governance tolerance that ordinary cash circulation receives.

Put the principle as a slogan: a small cash payment is presumptively transactional; a very large cash transfer is presumptively evidential. The first is something the law lets pass without asking questions. The second is something the law treats, by default, as a matter requiring explanation.

Figure 1. “Cashness” is scalar. The tolerance the law extends to bearer finality decays as transaction value and opacity rise, while the burden of verification and governance rises to meet it. The banknotes remain legal currency at every point on the curve; what changes is the institutional treatment of the transaction. (Schematic — illustrative of the argument, not an empirical measurement.)

This scalar view is not a moral claim that large sums are wicked. It is an institutional claim about where the law’s tolerance is rationally placed, and the next section explains why the placement makes economic sense.

Small cash solves transaction costs; large opaque cash creates verification costs

Standard economics gives money three functions: a medium of exchange, a store of value, and a connection to the unit of account. Cash performs all three, but it also does things that bank money and card payments do not fully replicate: it settles immediately and with finality, it works offline, it requires almost no operational infrastructure at the point of payment, it transfers by simple delivery, it preserves a measure of privacy, and it keeps working when the power and the networks fail. For small, immediate, ordinary exchange, these are not trivial advantages.

The institutional-economics tradition explains why an arrangement like cash exists at all. Ronald Coase’s insight was that the structure of economic institutions is shaped by transaction costs — the costs of search, negotiation, verification, and enforcement — and that arrangements emerge to economise on them. Douglass North built the broader account: institutions are the “rules of the game” that reduce uncertainty and make exchange possible across distance and anonymity. Oliver Williamson’s work on governance structures pressed the same logic into the question of how transactions are organised to manage the hazards specific to them. Cash, read through this lens, is a transaction-cost technology. It lets low-value payments clear without anyone bearing the cost of investigating provenance — a cost that, for a tenner, would dwarf the value at stake.

But the same features that make cash cheap at small scale make it expensive, and dangerous, at large scale. The advantages do not merely fade as value rises; they invert. Counting, storing, securing, transporting, insuring, and verifying large sums of physical currency is costly and risky. And the privacy that is a benign convenience at the till becomes, in bulk, opacity — the central concern of the systems built to police tax, sanctions, insolvency, and money laundering.

So the economic question driving the whole field is not whether cash is good or bad. It is a question of calibration: how much finality does cash need to function, and how much recoverability does the law need so that it does not become, in effect, a laundering subsidy? Recoverability protects victims and public order. Excessive recoverability destroys liquidity by turning every recipient into a title investigator — exactly the outcome Miller v Race was designed to avoid. The trade-off is real and it is two-sided, and the law’s answer is the scalar tolerance of Figure 1.

State the inversion plainly, because it is the engine of the argument: small cash solves a transaction-cost problem; large anonymous cash creates a verification-cost problem. The first is something society wants to be cheap. The second is something society needs to be legible.

The limits of cash, and why value changes legal character

It follows that the glib claim “cash is cash at every amount” is false in the only sense that matters. Legally, the same banknotes remain currency whatever the sum. Institutionally, the treatment is not the same. There are three limits at which the cashness of a transaction gives way.

The first is transactional. Cash is at its best for low-value, immediate, ordinary exchange. Its usefulness declines with size because the physical burdens of handling it grow faster than the value it carries.

The second is evidential. Cash is hard to trace. For small values that is tolerable, because the administrative cost of tracing would exceed any benefit. For very large values, the difficulty of tracing is not a side effect — it is the precise thing that compliance regimes exist to counteract.

The third is governance. Large anonymous bearer transfers cut against tax systems, sanctions enforcement, insolvency rules, fiduciary duties, court orders, and anti-money-laundering law all at once. At sufficient scale, a bearer transfer is not merely an exchange between two parties; it is an event with consequences for everyone whose interests those regimes protect.

The institutional world already encodes this. Value-keyed rules are everywhere: the United States has long required currency transaction reports above a fixed threshold, and the European Union’s 2024 anti-money-laundering reforms introduce an EU-wide ceiling on large cash payments. These are not claims that large sums are criminal. They are recognitions that large opaque transfers impose systemic risk, and that the law’s response should scale with the risk rather than with the physical form of the instrument. A five-dollar note buying coffee is ordinary cash. A fifty-thousand-dollar payment may require explanation depending on context. A ten-million-dollar transfer — in notes, tokens, stablecoins, or bearer instruments — cannot be treated as ordinary cash circulation, because at that level the transaction implicates governance, not merely exchange.

When cash can be reclaimed: the private-law hierarchy

So when does cash cease to be merely money in circulation and become recoverable property? The private law of recovery is best understood not as a single rule but as a ladder, organised by cause of action, with the claimant’s prospects strongest at the top and weakest at the bottom.

Proprietary recovery sits at the top in principle. If identifiable notes or coins are taken and remain identifiable, the owner may claim the thing itself. But physical cash loses identifiability almost immediately: it is spent, substituted, deposited, mixed, and exchanged, and once it has, the specific-thing claim evaporates.

Following and tracing are the law’s response to that evaporation. To follow is to pursue the same asset into another’s hands; to trace is to identify the value of the original asset as it is exchanged into substitutes. The difficulty is fungibility. Once cash is mixed in a pile, deposited into a bank account, or swapped for another asset, the claim shifts from physical identification to value identification — a far harder exercise. Banque Belge pour l’Etranger v Hambrouck (1921) is the early landmark, allowing a claimant to trace misappropriated money through bank accounts where its identity could still be established. Tracing through accounts, and through mixtures, has been the battleground of this branch of the law ever since.

Unjust enrichment and the action for money had and received form the doctrinal centre, and the decisive modern authority is Lipkin Gorman v Karpnale Ltd (1991). A partner in a firm of solicitors, a compulsive gambler, stole large sums from the firm’s client account and lost most of it at the Playboy Club’s tables. The firm could not recover from the thief, who was insolvent, so it sued the club. The House of Lords held that the club had been enriched at the firm’s expense by the receipt of stolen money, and was liable to make restitution — even though the club had received the money innocently and in good faith. Crucially, the same decision recognised the defence of change of position: the club’s liability was reduced by the winnings it had paid out, because it had, in good faith, changed its position in reliance on the receipt. Lipkin Gorman did two things at once: it confirmed that innocence does not automatically bar a restitutionary claim, and it gave innocent recipients a defence calibrated to what they had actually, detrimentally, done.

Personal claims against the wrongdoer sit alongside all of this and are easily forgotten. The thief or fraudulent payer remains personally liable even where the cash itself cannot be recovered from an innocent third party. The loss of a proprietary route is not the loss of all remedy; it is the loss of one remedy against one person.

Claims against recipients with notice or fault mark the point where the analysis leaves the world of innocent circulation altogether. Where a recipient had knowledge, notice, suspicion, dishonesty, or breached anti-money-laundering obligations, the case moves toward knowing receipt, dishonest assistance, money-laundering liability, or statutory recovery. These are not defences to circulation; they are reasons the protection of circulation never applied.

Figure 3. The recovery hierarchy. Recovery is strongest against the thief and the volunteer, weaker against a recipient with notice, and weakest of all against the bona fide purchaser who gave value without notice — against whom the money “passes in currency” and the claim is defeated. The dividing line is not innocence alone. The strong protection requires innocence plus value plus absence of notice plus ordinary circulation. Bar lengths are schematic.

“Without knowledge”: innocent receipt, and the real shape of the defence

The hardest cases are framed by a single deceptively simple phrase — the recipient who took the money “without knowledge.” The phrase hides at least four different questions, and collapsing them produces most of the confusion in this area.

In private law, lack of knowledge does not, by itself, defeat restitution. As Lipkin Gorman shows, a recipient can be wholly innocent and still be enriched at the claimant’s expense; innocence matters, but it matters at the level of defences — above all, change of position — rather than as a complete answer to the claim.

In property law, the innocent purchaser for value is treated very differently from the volunteer. The volunteer — the donee who gave nothing — is exposed, because there is no countervailing reason to protect a recipient who has parted with no value. The purchaser is protected, because commerce depends on the ability to accept money without reconstructing its past. Value is doing real work here: it is the thing that converts mere innocence into a defence.

In criminal law and anti-money-laundering, the relevant mental states are knowledge, suspicion, or reasonable grounds for suspicion. A person who genuinely does not know and has no reason to suspect that cash is criminal property is in a fundamentally different position from one who knows or suspects and proceeds anyway.

In civil recovery under the Proceeds of Crime Act 2002, the statutory language is decisive, and it crystallises the whole analysis. Property obtained through unlawful conduct is “recoverable property,” and the Act allows it to be followed into the hands of others and traced into substitutes. But section 308 carves out the bona fide purchaser: where a person obtains the property “in good faith, for value and without notice that it was recoverable property,” it may not be followed into that person’s hands, and it ceases to be recoverable. The statute thus reproduces, in modern form, the very compromise of Miller v Race — and it tells us exactly what the strong protection requires. Not innocence alone. Innocence plus value plus absence of notice.

A recent decision sharpens the limits of that protection in a way that matters for the scalar argument. In R (World Uyghur Congress) v National Crime Agency (2024), the Court of Appeal considered whether goods alleged to be the product of forced labour could be cleansed of their taint somewhere along a supply chain by the payment of market value. The Court held that the provision of adequate consideration by one party does not prevent property from being criminal property in the hands of another party who has the requisite knowledge or suspicion; and it rejected the proposition that the chain could be broken “merely by the use of adequate consideration in any of the transactions involved.” The section 308 exception is real, but it is narrow and personal to the qualifying recipient. Taint is not laundered away by the bare fact of payment. This is the doctrinal expression of the scalar idea: the law does not let value pass clean just because money changed hands; it asks who knew what, and gave what, and in what circumstances.

The state’s claim: seizure, forfeiture, and the discipline of proportionality

Private recovery is only half the picture. The state has its own claim on tainted value, and it is a different kind of claim, serving a different end.

Under the Proceeds of Crime Act 2002, the state can pursue value along several distinct tracks. Conviction-based confiscation follows a criminal conviction and is defendant-focused and value-focused: it orders the defendant to pay a recoverable amount, rather than necessarily seizing specific notes. Civil recovery under Part 5 operates in rem, against the property itself, on the civil standard of the balance of probabilities, and crucially without any requirement of a conviction — indeed, civil recovery orders have been made against people acquitted in the criminal courts. Cash seizure and forfeiture allow authorised officers to seize cash on reasonable grounds for suspecting it to be recoverable property or intended for use in unlawful conduct, with detention and forfeiture then subject to court process. Restraint orders, investigative orders, and taxation powers fill out the toolkit.

The point of the public scheme is not to return property to a private claimant. It is to disrupt criminal benefit, to prevent future unlawful use, and to defend the integrity of the monetary system as such. That is why it is structured around property and suspicion rather than around private title and proof of a specific wrong.

And that structure carries an unavoidable governance tension, which is really a rule-of-law tension. Seizure begins with suspicion; forfeiture and recovery require proof and process. If the thresholds are set too low, lawful cash users are burdened and the mere act of holding cash becomes a ground for suspicion — a corrosive outcome for financial privacy and for the presumption that ordinary people may use ordinary money without explaining themselves. If the thresholds are set too high, cash becomes a convenient laundering medium and the integrity goal collapses. The whole legitimacy of the regime depends on whether suspicion, proof, process, and proportionality are kept in proper alignment. This is the public-law face of the same calibration problem the private law faces: enough recoverability to protect the system, not so much that ordinary use is criminalised by default.

Digital cash is not cash because it moves digitally

Now extend the structure to the digital world, where the temptation to reason from labels is strongest and most dangerous. A system is not “cash” merely because its users can send units to one another peer-to-peer. The marketing word “digital cash” decides nothing. What decides things is a set of attributes: finality, control, governance, reversibility, traceability, legal status, settlement dependency, value stability, and recoverability. Test the system against those, not against its branding.

Run the test across five families. Physical cash gives direct possession and genuine bearer transfer. Bank deposits are not bearer anything; they are account-based claims against an institution, mediated and reversible through that institution and the courts. E-money and payment apps are contractual claims operated through regulated intermediaries. Cryptoassets are ledger-based assets whose operation in practice depends on a whole apparatus of protocol rules, validators or miners, software clients, exchanges, wallet providers, developers, courts, and social coordination. Central bank digital currencies, if issued, would be sovereign digital liabilities whose legal nature would depend almost entirely on design choices: account- or token-based, privacy-preserving or surveilled, programmable or not, offline-capable or not, intermediated or direct.

Figure 2. Four layers, six instruments. The same four questions — what form the value takes, what function it performs, who governs its alteration and reversal, and how it is recovered — receive different answers across the payment landscape. The governance row is the analytically decisive one: the darker the cell, the more concentrated the control over alteration, freezing, and reversal. Physical cash is the lightest; custodial stablecoins and design-led CBDCs are the darkest.

English law has already done much of the conceptual work. In AA v Persons Unknown (2019), the Commercial Court held that a cryptoasset such as bitcoin is property, applying Lord Wilberforce’s classic four-part test in National Provincial Bank Ltd v Ainsworth — that a property right must be definable, identifiable by third parties, capable of assumption by third parties, and have some degree of permanence. Bitcoin satisfies that test because it is rivalrous: the holding of a unit by one person prevents another from holding that very unit, a feature that arises not from law or convention but from the way the software works. The Law Commission’s Digital Assets report (2023) reached the considered conclusion that the common law can recognise a distinct, “third category” of personal property — beyond the traditional dichotomy of things in possession and things in action — capable of accommodating crypto-tokens, and recommended legislation to confirm it. That recommendation became the Property (Digital Assets etc) Bill, introduced to Parliament in September 2024, whose operative clause provides that a thing is not prevented from being the object of personal property rights merely because it is neither a thing in possession nor a thing in action.

None of this makes crypto “cash.” It makes crypto property — which is a different and, for our purposes, more important point. A digital object can be transferable and valuable without having the legal or economic status of cash; and conversely, any state design for genuine “digital cash” would have to reproduce cash’s institutional features — final settlement, broad acceptability, resilience, calibrated privacy, and lawful recoverability — rather than merely its peer-to-peer feel. “Digital cash” is not a technical category. It is a legal-governance category.

Blockchains are governance-bound, and code is not law

This brings us to the central modern claim, and to its refutation. The slogan “code is law” holds that a sufficiently decentralised protocol places transactions beyond the reach of courts: what the ledger records is final, full stop. The slogan is false, and the reason it is false is the reason the strong version of cash-finality is false. Blockchains do not abolish governance. They relocate it.

A distributed ledger as used by actual people depends, at every turn, on institutions and choices. Software gets updated; protocols get forked; validators and miners apply policies; exchanges list, delist, and freeze; wallet providers set defaults; node operators choose which implementation to run; issuers of stablecoins can freeze and blacklist; courts make orders; regulators enforce sanctions. Even where a base ledger is technically resistant to unilateral alteration, none of that resistance survives contact with the institutional layer. And the crucial analytical move is this: every one of those choices is governance. A fork is governance. A protocol upgrade is governance. Refusing to upgrade is governance. Exchange delisting is governance. Validator censorship is governance. A stablecoin freeze is governance. A court-ordered recovery is governance. Developer discretion is governance, where developers control the dominant implementation. Distributed control is still control. Fragmented governance is still governance. There is no point in the system at which “no one is in charge” in the sense the slogan requires.

The English courts have already stared directly at this question. In Tulip Trading Ltd v van der Laan (2023), a company that owned a large quantity of bitcoin, but had lost the private keys controlling it, sued the developers of several networks — including the developers associated with BTC and the Bitcoin Association for BSV — arguing that they owed it fiduciary and tortious duties to take steps that would restore its access. At first instance the claim was struck out. The Court of Appeal reversed, holding that there was a serious issue to be tried: it was arguable that developers who exercise real control over a network’s software, making discretionary decisions on behalf of all participants, occupy a role with fiduciary characteristics. Birss LJ framed the claimant’s case as the contention that “the decentralised governance of bitcoin really is a myth,” and held that whether decentralisation is genuine is a question of fact to be established at trial on evidence, not a legal shield to be assumed in the developers’ favour. His most quoted line is the one that matters most for the slogan: “the internet is not a place where the law does not apply.”

The state of the authority should be stated precisely. There is, as yet, no English decision holding that developers owe such duties; there is a Court of Appeal decision that the question is arguable enough to be tried — that the issue turns on facts about how control is actually exercised over a network, and that those facts cannot simply be assumed away by invoking the word “decentralisation.” That is itself a significant ruling for the thesis here. It means the question of who governs a blockchain is a justiciable legal question, to be answered by examining where control really sits, rather than a matter the protocol can place beyond the court’s reach by description alone. Whether or not a fiduciary duty is ultimately made out on any given set of facts, the court’s willingness to entertain the question is the point: the governance of a ledger is something the law will look into.

The argument for the proposition does not, in any event, depend on the outcome of any single piece of litigation. It rests on the distinction between two kinds of finality. Technical finality means the system treats a transaction as settled according to its own rules. Legal finality means the law declines to unwind, reallocate, compensate, seize, or otherwise disturb the value consequences. These are not the same thing, and conflating them is the whole error. A technically final transfer can still generate a proprietary claim, a restitutionary claim, criminal liability, a tax consequence, fiduciary liability, sanctions exposure, or a court-ordered remedy. Code can make a transfer happen. It cannot determine whether the signer was coerced, mistaken, hacked, defrauded, insolvent, mentally incapacitated, acting outside their authority, laundering proceeds, breaching a trust, or subject to a freezing order. Those are questions of right, and questions of right are decided by law. Code may settle the ledger; law settles the rights.

Where digital cash is recoverable: from possession to control points

If blockchains are governance-bound, then the recovery framework built for physical cash carries over — but with one structural change. For physical notes, recovery is possession-based: you chase the thing or its traceable substitute. For digital value, recovery shifts to control points — the institutions and parties with the practical ability to freeze, transfer, or surrender the asset.

Digital value can be reclaimed in several ways, none of them defeated merely by the assertion of technical finality. It can be recovered as property, where the legal system recognises the asset as ownable — which, as we have seen, English law now does. It can be recovered by tracing, where movement through addresses, accounts, exchanges, bridges, and substitutes can be established. It can be recovered through restitution, where a recipient has been unjustly enriched. It can be recovered through court orders against intermediaries — exchanges, custodians, wallet providers, stablecoin issuers — who hold the practical levers. And it can be recovered through criminal confiscation or civil recovery where the value represents proceeds of crime, the digital case being no different in principle from the cash case.

The most instructive recent decision shows both the promise and the limits, and it refuses to flatter either side. In D’Aloia v Persons Unknown (2024), the High Court held — for the first time at a contested trial rather than an interim hearing — that a stablecoin, Tether (USDT), is property under English law: neither a thing in possession nor a thing in action, but a distinct form of property capable of being traced and of forming the subject matter of a trust. That is a significant doctrinal win for victims. And yet the claimant lost against the exchange he was pursuing. He had been defrauded of around £2.5 million in stablecoins, which moved through fourteen “hops” before some portion reached a wallet on the exchange Bitkub. The court accepted that a constructive trust could in principle bind an exchange that received identifiable misappropriated crypto, and that the asset could in principle be followed through a mixture. But on the facts, the claimant’s expert evidence — a “black box” tracing methodology that could not be transparently audited — failed to prove that his coins, rather than someone else’s, had reached the wallet. The claim failed not because crypto is beyond recovery, but because recovery, like all recovery, depends on proof.

That is the honest shape of digital recoverability, and it points in both directions at once. The law is willing to treat digital value as property, to trace it, to impose trusts on those who receive it with notice, and to order intermediaries to give it up. But the practical difficulty of tracing fungible value through mixtures and across jurisdictions is real, and it is often decisive. Recoverability is genuine; it is also hard. And critically, it scales with control: a custodial stablecoin is far easier to freeze than a bearer banknote, a transparent ledger far easier to trace than a privacy coin, a programmable CBDC reversible by design if a state chooses to build it so. The cruder claim — that digital assets are simply unrecoverable — is false. The accurate claim is that recovery has migrated from the holder’s pocket to the system’s control points, and that whoever holds those points answers to the courts.

The law of controlled amnesia

Return, finally, to the negotiation we began with, because the whole field reduces to it.

Cash is best understood not as a form of money but as a legally tolerated form of bounded bearer finality. That tolerance is strongest where value is low, exchange is ordinary, and the social benefit of frictionless payment exceeds the risk of concealed wrongdoing. It weakens as value, opacity, abnormality, and systemic consequence rise. A five-hundred-dollar cash payment is cash in the full sense: ordinary, final, private, practically uninvestigated. A ten-million-dollar transfer in banknotes, tokens, stablecoins, or ledger entries cannot claim the same treatment merely by wearing a cash-like form. At that scale the transaction is a governance problem, and the law treats it as one.

Digital systems do not escape this logic; they re-enact it. Blockchains, smart contracts, stablecoins, and CBDCs are not outside the law. They are governance-bound arrangements whose rules can be altered, contested, enforced, ignored, or overridden by the institutions and courts that surround them. Code can execute a transfer. It cannot determine whether the transfer was valid, recoverable, criminal, taxable, voidable, or final. Those determinations belong to law, and they always have.

None of this means cash should be abolished, and the governance counterweight is worth stating because it is easy to forget amid the enforcement logic. Cash is in retreat but it is not obsolete, and the law has recognised as much. In the United Kingdom, the share of payments not made in cash rose from 46% to 86% between 2002 and 2022, thousands of free-to-use ATMs were removed, and bank branches closed at pace. Yet around three million adults still rely on cash, a substantial share of small businesses still take most of their takings in it, and a global IT outage in 2024 reminded everyone what dependence on a single digital rail can cost. The Financial Conduct Authority’s access-to-cash regime, in force from 18 September 2024 under powers granted by the Financial Services and Markets Act 2023, now requires designated banks and building societies to assess and fill significant gaps in cash provision. Inclusion, privacy, and resilience are not nostalgia. They are reasons the institution of cash still earns its keep — and reasons the right answer is calibration, not abolition.

Figure 4. Cash is in retreat, but not obsolete — the governance counterweight. The collapse of cash’s share of payments sits alongside the persistence of a population that depends on it, which is why the policy response has been to guarantee access rather than to let cash disappear. Figures drawn from the FCA’s access-to-cash policy materials (PS24/8) and related FCA statements, 2024.

So the field has a shape, and the shape is a single principle worn by many doctrines. Cash is recoverable when the law can identify a claimant, a wrong, a proprietary or restitutionary path, a statutory basis, or criminal taint — without defeating the ordinary function of cash as circulating money. It is not recoverable merely because some prior holder suffered a loss, if the value has passed in good faith, for value, without notice, and in ordinary circulation. The same logic governs digital value, with recovery migrating from possession to control points but never disappearing.

The law of cash, in the end, is a law of controlled amnesia. Cash must forget enough of its past to circulate, but not so much that theft, mistake, fraud, and criminal proceeds become legally purified by movement. The strong claim made for distributed ledgers — that technical finality can manufacture the kind of forgetting that places value beyond the reach of courts — asks the law to forget everything. It will not. It never has. And the cases now coming through the courts, on stablecoins and tracing and developer duties, are the law saying so, one judgment at a time.

Notes and references

The standard applied here is set out in the note that follows. Primary legal authorities are cited in the conventional form; the holdings, and the quoted passages, were taken from the judgments and the legislation themselves, not from secondary summaries.

Cases

1. Miller v Race (1758) 1 Burr 452; 97 ER 398 (KB) — bank note treated as currency; no property in currency once paid away in good faith for value; per Lord Mansfield.

2. Banque Belge pour l’Etranger v Hambrouck [1921] 1 KB 321 (CA) — tracing of misappropriated money through bank accounts.

3. Lipkin Gorman v Karpnale Ltd [1991] 2 AC 548 (HL) — action for money had and received; restitution for unjust enrichment against an innocent recipient of stolen money; recognition of the change-of-position defence.

4. AA v Persons Unknown [2019] EWHC 3556 (Comm) — cryptoassets are property, applying the property test in National Provincial Bank Ltd v Ainsworth [1965] 1 AC 1175 (HL) (per Lord Wilberforce); rivalrousness of bitcoin.

5. Tulip Trading Ltd v van der Laan [2023] EWCA Civ 83 (CA) — arguable (serious issue to be tried) that software developers owe fiduciary/tortious duties to coin owners; whether a network is genuinely decentralised is a question of fact to be established on evidence at trial; “the internet is not a place where the law does not apply” (Birss LJ). The Court did not decide that such a duty exists. The fiduciary test applied was that in Bristol and West Building Society v Mothew [1998] Ch 1 (CA).

6. D’Aloia v Persons Unknown Category A & Ors [2024] EWHC 2342 (Ch) — first trial-level holding that a stablecoin (USDT) is property under English law; constructive trust against a receiving exchange arguable in principle; tracing through mixtures available in principle; claim nonetheless failed on the facts for want of transparent, reliable tracing evidence.

7. R (World Uyghur Congress) v National Crime Agency [2024] EWCA Civ 715 (CA) — adequate consideration anywhere in a supply chain does not, by itself, cleanse property of criminal taint or break the chain; analysis of POCA s 308.

Legislation

7. Proceeds of Crime Act 2002, Part 5 (esp ss 240–241, 266, 278, 304, 305, 308) — recoverable property; following and tracing; the bona fide purchaser exception in s 308 (”good faith, for value and without notice”). Part 7 (esp ss 327–329, 340) — money-laundering offences and the meaning of criminal property. — Financial Services and Markets Act 2023; Financial Services and Markets Act 2000, Part 8B — statutory basis for the FCA access-to-cash regime.

8. Property (Digital Assets etc) Bill (introduced to Parliament 11 September 2024) — operative clause: a thing is not prevented from being the object of personal property rights merely because it is neither a thing in possession nor a thing in action.

Official and regulatory materials

10. Law Commission, Digital Assets: Final Report (Law Com No 412, 2023), and Digital Assets as Personal Property: Supplemental Report and Draft Bill (2024) — recommendation to confirm a distinct “third category” of personal property for certain digital assets.

11. Financial Conduct Authority, Access to Cash (Policy Statement PS24/8, July 2024) and associated FCA statements (2024) — rules in force 18 September 2024; fourteen designated firms; cited figures on the decline of cash payments (46% → 86%, 2002–2022), ATM removals, SME cash reliance, and the c. 3 million adults who rely on cash. — Bank of England, public guidance on legal tender and banknotes — legal tender concerns the discharge of debts and does not compel retail acceptance.

Scholarship (cited as relied upon, not independently vouched for)

— Angela Walch, ‘In Code(rs) We Trust: Software Developers as Fiduciaries in Public Blockchains’ in Philipp Hacker and others (eds), Regulating Blockchain: Techno-Social and Legal Challenges (OUP 2019) — cited because the Court of Appeal in Tulip Trading expressly referred to it; it is reported here as a source the court relied upon, not as a work whose full argument has been independently re-analysed for this essay.

Economics (canonical works, cited for their established central propositions)

2. R H Coase, ‘The Problem of Social Cost’ (1960) 3 Journal of Law and Economics 1; and ‘The Nature of the Firm’ (1937) 4 Economica 386 — transaction costs as the determinant of institutional arrangements.

3. Douglass C North, Institutions, Institutional Change and Economic Performance (CUP 1990) — institutions as the “rules of the game” that reduce uncertainty in exchange.

4. Oliver E Williamson, The Economic Institutions of Capitalism (Free Press 1985) — governance structures as responses to transaction hazards.

A note on sources and the standard applied

This essay is built, deliberately, on a backbone of primary legal sources — decided cases and statutes — together with official and regulatory publications. Those are the load-bearing citations, and their holdings and quoted words were taken from the judgments and legislation themselves. Where a normative or evaluative claim is made (for instance, that “calibrated finality” is the right governance posture, or that cashness is best understood as scalar), it is presented as the essay’s argument, not as a demonstrated fact; the descriptive legal propositions are what the cited authorities establish.

Two further points of candour. The economics references (Coase, North, Williamson) are cited for their established central propositions, which are not in dispute, rather than for fine-grained empirical findings; readers should treat them as canonical signposts, not as freshly re-derived results. And the single secondary academic source named — Walch — is cited only because the Court of Appeal relied on it; it is reported as the court’s reliance, not as a work whose full text has been independently analysed here.

Keywords: human flourishing, economics, scarcity, human action, subjective value, institutions, property, rule of law, calculation, dispersed knowledge, Mises, Hayek, Adam Smith, dogma, ordered liberty, markets, morality.

Abstract

Economics is routinely caricatured as a cold arithmetic of money, a justification of greed, or a quasi-religious worship of markets indifferent to human life — and on the strength of that caricature it is dismissed as dogma. This essay argues that the caricature is false and that the dismissal is exactly backwards: properly understood, economics is the least dogmatic of the human sciences, because it is the disciplined study of human action under scarcity, uncertainty, knowledge-limits, institutions, and moral constraint, and its defining habit is the refusal to look away from consequences. The argument proceeds in stages. It begins from the human person — a being with ends, limited means, time, and the unavoidable necessity of choice — and recovers a serious account of human flourishing, which is not a pile of consumption goods but a condition of ordered liberty: the capacity to plan, create, own, exchange, raise a family, build enterprises, preserve capital, and pass something better to those who follow. That capacity, the essay shows, rests on conditions no society can flourish without — secure property, sound money, the rule of law, enforceable contract, open markets, reliable institutions, rights that protect creators, and a culture that honours making rather than envying it. It then locates economics correctly as a moral science of consequences: a discipline that cannot supply ultimate ends, but that clarifies the means, constraints, and effects through which whatever ends we choose must be pursued, and that therefore cannot be separated from morality without becoming either cold technocracy or sentimental ruin. From there it draws the line that the title turns on. Dogma is not the possession of principles; principles are indispensable. Dogma is the refusal of reality when reality contradicts the slogan — and by that test the dogmatists are the systems that suppress their own feedback: the socialist who promises abundance and blames sabotage when calculation fails, the anarchist who calls institutions illegitimate while living on their benefits, the anti-intellectual-property advocate who calls copying freedom while ignoring the creator who bore the cost, the technocrat who imagines he can optimise a society whose knowledge is dispersed beyond any planner’s reach, and the market-absolutist who forgets that markets rest on moral and legal foundations they cannot themselves supply. Real economics is anti-dogmatic precisely because it builds in correction — loss, bankruptcy, exit, substitution, the adjustment of prices — and asks of every rule not whether it flatters an ideology but how it actually works in human life. The essay closes where it began, with the human person: a person flourishes not in a void but inside an order of law, property, trust, sound money, open markets, and moral restraint, and economics, far from being the enemy of that flourishing, is the map of the constraints and consequences within which it is possible. Human flourishing is not produced by dogma. It is produced by free persons acting within a moral and institutional order that protects creation, rewards responsibility, disciplines error, and lets civilisation compound across generations.

I. Economics is about human life, not sterile equations

Begin by refusing the caricature, because almost everything false about the popular view of economics follows from it. The caricature holds that economics is about money — about profit, markets, the accumulation of goods — and that the economist is a person who knows the price of everything and the value of nothing, reciting slogans about growth while the actual texture of human life goes on somewhere else. On that view economics is indeed a kind of dogma: a narrow creed of acquisition, indifferent to the things that make life worth living.

But that is not what economics is, and it never was. Economics begins with the human person acting in the world. A person has ends he cares about and means too limited to pursue them all at once. He has time, which runs in one direction and cannot be recovered. He has knowledge, which is partial, and faces a future, which is uncertain. And because of all this he must choose — must weigh the present against the future, must decide whether to consume or save, to work or rest, to build or wait, to risk or preserve, to teach his children or pursue his own projects, to start the enterprise or keep the secure post. Economics is the study of human beings making those choices under those constraints, and of what happens — to them and to everyone around them — as a result.

This makes economics a discipline of reality before it is a discipline of money. It is the study of consequences: of what actually follows when people act under scarcity, and when the institutions around them reward or punish particular kinds of conduct. The price of a thing enters the story not because money is the point but because a price is a piece of information about scarcity and desire that a person needs in order to choose well. Profit and loss enter not as moral verdicts but as signals that an effort has served others or failed to. The market enters not as an idol but as one of the arrangements through which dispersed people coordinate their plans without any of them commanding the others. To mistake the instruments for the subject — to think economics is about money the way a caricature of medicine might think it is about scalpels — is to miss the whole of it.

And here is the first reason the dismissal of economics as dogma is exactly backwards. Economics is not dogma because reality is not dogma. Scarcity is not dogma. Time is not dogma. Risk is not dogma. The need to choose is not dogma. These are not ideological commitments that an economist has elected to hold; they are features of the human condition that no act of will can repeal. A discipline organised around them is not a creed demanding assent. It is an attempt to take seriously the situation every human being is actually in.

II. What human flourishing actually requires

If economics is finally about human life, then the question it serves is the question of human flourishing — and here too a caricature has to be cleared away, because flourishing is routinely confused with mere consumption, as though to live well were simply to have more.

It is not. Human flourishing is the capacity to live as a responsible, creative, reasoning person within a stable moral and institutional order. It includes work that means something, a family one can provide for and protect, the education of one’s children, the exercise of creativity, the earning of a good name, the building of an enterprise, security against violence and ruin, the freedom to give, and a culture worth belonging to. A pile of goods is, at most, an input to some of these, and beyond a point it is not even that. The miser surrounded by wealth he cannot enjoy is not flourishing; the harried consumer drowning in possessions and starved of meaning is not flourishing; and a society that measures itself only by the height of its consumption has mistaken the smoke for the fire.

But — and this is where economics rejoins the moral picture rather than retreating from it — flourishing so understood does not happen in a vacuum. It has conditions, and the conditions are exactly the things a serious economics studies. Flourishing requires property secure enough that a person can plan across time, confident that the field he cultivates, the house he builds, the capital he saves, and the work he creates will still be his tomorrow. It requires law stable enough to support trust, so that strangers can deal with one another without each transaction being a gamble. It requires markets open enough that effort can find its reward through exchange rather than through favour. It requires institutions reliable enough to lower the uncertainty that would otherwise paralyse action. It requires money sound enough to preserve the possibility of calculation, so that saving is not silently confiscated and plans reaching into the future are not dissolved by a debased unit of account. It requires contracts enforceable enough to permit cooperation among people who will never meet. It requires rights strong enough to protect those who create value from those who would simply take it. And it requires a culture serious enough to honour production and responsibility rather than to envy and resent them.

Figure 1. What human flourishing requires. Not a pile of goods, but a condition of ordered liberty resting on institutions that allow a person to act across time.

This is the answer to crude materialism, and it cuts in both directions at once. Against those who reduce economics to acquisition, it insists that the point of all the property and exchange and calculation is not goods for their own sake but the human capacity to plan, create, and live responsibly across time. And against those who imagine that flourishing can be conjured by good intentions alone — by declaring rights to things without securing the conditions that produce them — it insists that the conditions are real and that they cannot be wished into being. Flourishing is not a pile of goods. It is ordered liberty: a condition in which free persons, secured by institutions and restrained by law and morality, can pursue excellence, take responsibility, and build something that lasts.

III. Economics as a moral science of consequences

To see why economics cannot be dogma, and why it also cannot stand apart from morality, it helps to be precise about what economics can and cannot do — because its limits are as important as its powers, and dogma is precisely what fills the space when those limits are forgotten.

Economics cannot tell a man his ultimate ends. It cannot tell him whether to love his children, honour his promises, write the book that is in him, start the company, devote himself to a cause, or spend his life in quiet service. Those are questions of morality, of meaning, of the kind of person one chooses to be, and economics has nothing to say about them in the imperative. What economics can do — and it is no small thing — is illuminate the means, the constraints, and the consequences through which whatever ends a person or a society chooses must actually be pursued. It can explain why some institutional arrangements make honourable action easier and others make it nearly impossible. It can show how incentives shape conduct over time, how secure property encourages people to maintain and improve what they hold, how price signals coordinate the knowledge of millions who could never coordinate by design, how the debasement of money quietly destroys the capacity to plan, how the threat of confiscation drives investment into hiding or out of existence, and how pervasive uncertainty suppresses the very production a society depends on.

This is exactly where the phrase “not dogma” earns its meaning, and the contrast is sharp. Dogma says: repeat the slogan regardless of consequences, and judge by the purity of the intention. Economics says: examine the structure of action and the actual effects of the rules, and judge by what happens to the human beings who must live under them. Dogma treats good intentions as sufficient — it announces a goal and assumes the goal will be reached because it was sincerely meant. Economics insists that outcomes matter, because human beings live with outcomes and not with intentions, and that a policy meant to help can harm, and frequently does, when it ignores how people actually respond to the incentives it creates. The economist is not the enemy of the moralist here; he is the moralist’s indispensable ally, the one who asks, of any noble end, “and by what means, and with what effects, and at whose cost?” — the questions without which good intentions become, with depressing regularity, the architecture of avoidable suffering.

So economics is a moral science not because it dictates morals but because it cannot be separated from them without distortion. It takes human ends as given and asks how they can be served; it refuses to pretend that the asking is neutral, because the answers bear directly on whether real people will flourish or suffer. To do economics seriously is already to care about consequences, and to care about consequences is already a moral act.

IV. Scarcity, choice, and why morality cannot ignore it

At the root of all this lies scarcity, and it is worth dwelling on, because scarcity is both the reason economics exists and the reason a morality that ignores economics turns destructive.

Scarcity means that not every end can be pursued at once. It is not merely that goods are limited; almost everything that matters is scarce. Time is scarce — the hour spent on one thing is gone from every other. Attention is scarce. Labour is scarce, and skill scarcer still. Capital is scarce, and the institutional competence to deploy it scarcer than that. Trust is scarce. Even genius — the rare capacity to see what others miss and to make what did not exist — is desperately scarce. Because all of this is true, choices are not optional and they are not free of cost: to do one thing is to forgo another, and the forgone alternative is the real price of every act.

And because choices have consequences, the rules a society sets up — the things it rewards and the things it punishes — shape what it gets, reliably and without regard to anyone’s intentions. If a society punishes production and rewards appropriation, it will get less production and more appropriation. If it attacks property, it will get less of the stewardship that property encourages, because no one improves what may be taken. If it weakens contracts, it will get less cooperation, because strangers will not rely on promises that cannot be enforced. If it socialises the fruits of invention, it will get less invention, or it will drive invention into secrecy. If it rewards political access over service to customers, it will get courtiers and corruption rather than producers. None of these is a moral judgment smuggled in; each is a statement about how human beings respond to incentives over time, and each has been confirmed by experience often enough to count as something close to law.

This is why morality and economics cannot be divorced, and why the divorce is dangerous rather than merely incomplete. A moral system that ignores scarcity becomes what one might call sentimental tyranny: it promises everything to everyone — security, abundance, dignity, care — as though these were free, and then, when reality refuses to supply what was promised, it reaches for coercion to close the gap between the promise and the possible. The promise was generous; the result is the boot. A morality serious about human beings must therefore be serious about scarcity, because only then can it ask the questions that turn good intention into good outcome: not merely “what do people deserve?” but “what arrangements actually produce the things people need, and what arrangements destroy them?” Compassion that refuses this question does not thereby become more compassionate. It becomes, in its effects, the opposite of what it intends.

V. Property as the foundation of human agency

Of all the conditions of flourishing, property is the one most often mistaken for mere greed, and getting it right is central to the whole argument — because property is not, at bottom, about having. It is about agency across time.

Property is the institutional recognition that a person may control, use, exclude others from, invest in, defend, transfer, and pass on a valuable interest. What that recognition buys is continuity of action. It allows a person to say: this field will be cultivated and not merely stripped, because its harvests will be mine; this machine will be maintained, because I keep what it produces; this business will be built over years, because it will still be mine when the years have passed; this invention will be developed through its long unprofitable infancy, because I will share in its maturity; this book will be written, because the work will be recognised as mine; this capital will be saved rather than consumed, because the future it secures is my own. Strip property away and every one of these sentences becomes unsayable. Planning collapses into immediate consumption, because there is no point in deferring what others may seize; stewardship collapses into use-it-or-lose-it; and the long, patient, future-directed activity on which civilisation depends gives way to force, favour, or the grey paralysis of bureaucracy. Hayek put the deep point with characteristic precision: the system of private property, he observed, is the most important guarantee of freedom not only for those who own but scarcely less for those who do not, because it is only when control of the means of production is divided among many independent hands that no single power can dominate us all.

And the property that flourishing requires is not only the tangible kind. It includes land and tools and machines and homes, certainly; but it includes equally the company share and the debt, the contract and the goodwill, the patent and the copyright and the trade secret, the registered design and the commercial identity. These intangible forms are not lesser or suspect for being intangible; they are the instruments through which created value — the value a person brings into being by thought, effort, risk, and disclosure — is secured to its creator long enough for him to recover what its making cost and to be moved to make again. A doctrine that protects the physical object but treats created value as common the moment it proves useful has not defended property; it has gutted it, leaving the husk of possession while surrendering the thing that made property worth having. Flourishing requires the protection of created value, not merely of physical objects, because it is in the creation of value — not the holding of matter — that human agency most fully expresses itself.

VI. Institutions are not optional decorations

If property is the foundation of agency, institutions are the structure that makes property — and everything built on it — possible, and the fantasy that they can be dispensed with is among the most damaging in circulation.

No advanced society has ever flourished without institutions. Property itself, money, contract, inheritance, courts, registries, standards, accounting, reputation, professional norms, the whole apparatus of commercial law — these are not ornaments hung on an economy that would function just as well without them. They are the machinery through which strangers cooperate peacefully at scale, the means by which a person in one city can rely on the promise of a person he will never meet in another. Take them away and you do not get a freer version of the same prosperity; you get the world before them, in which exchange shrinks to the circle of personal acquaintance and trust, enforced by private force, local dominance, clan loyalty, and the threat of retaliation.

Here the essential distinction — the one the anti-institutional imagination consistently blurs — is between limited government and no government. Limited government means law-bound institutions confined to a definite and defensible task: protecting property, enforcing contracts, adjudicating disputes, preventing violence, and restraining coercion, while staying out of the business of planning the economy or directing the lives of citizens. No government means the dissolution of those general rules into private power — the collapse of the courthouse into the strong arm, of title into possession, of contract into whatever can be privately enforced. These are not two points on a single scale such that less government is always more freedom; they are different kinds of order, and past a certain point less institutional structure means not more liberty but less, because liberty itself is an achievement of institutions and not a residue left when they are removed. Small government is not no government. It is government confined to the defence of the institutional order that makes liberty possible.

This is precisely the position of the economists most often conscripted for the opposite view. Mises and Hayek were relentless critics of socialism, of central planning, of arbitrary and discretionary intervention — and they were nothing of the kind toward law, title, courts, and the institutional order. Mises held the state to be the indispensable apparatus that secures peace, observing that the anarchist’s dream founders on the plain fact that some people will not, of their own accord, respect the rules on which social cooperation depends, so that an institution capable of restraining them is not a regrettable concession but a requirement of civilised life. Hayek devoted his maturity to showing that the rule of law — general, abstract, prospective rules binding everyone, the state included — is not the enemy of the market but its precondition, the framework without which the spontaneous order of exchange cannot arise at all. To invoke their critique of planning as if it were a brief against institutions is to mistake the doctor’s warning against poison for a campaign against food. They opposed the planning state. They built their whole defence of liberty on the institutional order that the anti-institutional fantasy would tear down.

VII. Markets as discovery, not dogma

The market, too, is constantly mistaken for an object of dogmatic worship, as though those who defend it did so out of mystical faith in an invisible hand. The truth is the reverse: the case for the market is a case from humility, and it is one of the least dogmatic arguments in the whole of social thought.

The argument runs like this. The knowledge a society needs in order to coordinate its activity — who wants what, how urgently, at what cost, with what alternatives, under what local conditions — is not concentrated anywhere. It is dispersed across millions of minds, much of it tacit, none of it available in its entirety to any single person or board or ministry. Prices are the mechanism by which this scattered knowledge is condensed into signals that each person can act on without needing to know the whole: a rising price says “this has become scarcer or more wanted, economise or supply more,” and it says so to everyone at once, carrying information that no planner could gather. Profit and loss then discipline error after the fact — rewarding the judgments that served others, penalising those that did not — and competition allows many different judgments to be tried at once, so that better ways of doing things can be discovered rather than decreed. The market, in short, is a discovery procedure: a way of finding out what works in a world where no one can know in advance.

This is not faith; it is the acknowledgment of ignorance, and it is the exact opposite of the planner’s posture. The planner imagines that he can know enough — about preferences, technologies, conditions, and their endless interactions — to direct society toward a better outcome than the uncoordinated choices of its members would reach. The defender of the market makes the humbler and far better-supported claim that no one can know that much, and that a system which harnesses dispersed knowledge through prices and corrects its own mistakes through profit and loss will outperform any attempt to substitute a single mind’s plan for the knowledge of the many. Hayek named the lesson with a phrase that ought to hang over every ministry: “The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.” The point is not that markets are perfect — they are not, and no serious economist claims they are. The point is that markets contain correction mechanisms — loss, bankruptcy, the loss of reputation, the substitution of alternatives, the free entry of new competitors and the exit of failed ones, the constant readjustment of prices — that dogmatic systems lack and actively suppress. To prefer the system that learns from its errors over the system that forbids the admission of error is not dogma. It is the beginning of wisdom about the limits of human knowledge.

VIII. Dogma is the refusal of correction

We can now say sharply what dogma actually is, because the popular usage — which treats any firm conviction as dogma — gets it exactly wrong, and the error matters.

Dogma is not having principles. Principles are necessary; a mind without them is not open but empty, and a society without them cannot sustain the trust on which everything else depends. Dogma is something more specific and more corrosive: it is the refusal of reality when reality contradicts the slogan. It is the determination to hold the conclusion fixed and to explain away, suppress, or simply ignore whatever evidence threatens it. Adam Smith — who, it is worth remembering, founded modern economics and also wrote a treatise on the moral sentiments, and was therefore the last person to imagine economics could be amoral — drew the portrait two and a half centuries ago in his description of the doctrinaire reformer, the “man of system,” who “is apt to be very wise in his own conceit, and is often so enamoured with the supposed beauty of his own ideal plan of government, that he cannot suffer the smallest deviation from any part of it.” That is dogma exactly: the plan held sacred, and reality forbidden to interfere.

By this test, the dogmatists are not those who hold that institutions matter or that incentives are real. They are the systems that keep a slogan and look away from the one fact that would refute it. Socialist dogma promises that collective ownership will produce abundance, and when calculation fails — as Mises demonstrated it must, since the abolition of private ownership of the means of production destroys the market prices for capital goods without which rational allocation is impossible — it blames saboteurs, wreckers, and hostile encirclement rather than the slogan. Anarchist dogma declares all institutions illegitimate while quietly relying on the title, the contract, and the court whose legitimacy it denies. Anti-intellectual-property dogma announces that copying is freedom and looks away from the creator who bore the entire cost of producing the thing before anyone could copy it. Technocratic dogma insists that experts can optimise society and ignores the dispersed knowledge, the incentives, and the unintended consequences that defeat every such attempt. And market-absolutist dogma, where it genuinely exists, forgets that markets rest on moral and legal foundations they cannot themselves supply, and treats as self-sustaining an order that in fact depends on institutions and norms outside the market’s own logic.

Figure 2. Dogma is the refusal of correction. Each dogma keeps a slogan and looks away from a single decisive fact; real economics looks at the fact and builds in the feedback that dogma suppresses.

What unites these, and what makes them dogma in the strict sense, is the suppression of feedback. Each holds its conclusion immune from the evidence that would test it, and each must therefore silence or explain away the correction that reality keeps trying to supply. Real economics is anti-dogmatic by constitution, because its central question is not “does this confirm the creed?” but “how does this rule actually work in human life, and what happens to people when it operates?” — a question that keeps the door open to being wrong, which is the one thing dogma cannot abide.

IX. Human flourishing requires creators, not parasites

A society that means to flourish must, above all, protect those who create, and the failure to do so is at once a moral failure, an economic failure, an institutional failure, and a cultural one. This is where the abstract argument touches something urgent.

The structure of creation is always the same: the inventor, the author, the engineer, the scientist, the farmer, the builder, the entrepreneur, the teacher, the designer, the organiser — each brings value into being before others can consume it, bearing the cost and the risk and the uncertainty in advance, and disclosing the result so that others may benefit. A society that protects this activity teaches its members that creation is worth attempting; a society that treats the creator’s work as common material the instant it becomes useful teaches them the opposite — that the wise course is to wait, to copy, to let someone else bear the cost of discovery and then to take the result for nothing. Over time a society gets what it rewards, and a society that rewards the copyist over the creator will find itself with more of the former and less of the latter, which is to say with less of everything that creation produces.

The doctrine that would strip the creator of his claim is therefore wrong on every axis at once. It is morally wrong, because it denies a person the value he brought into being and would not exist but for him. It is economically wrong, because it externalises the benefit of creation to all while leaving the cost with the originator alone, which is precisely the arrangement that produces too little of whatever is so treated. It is institutionally wrong, because it severs the link between effort, risk, disclosure, and reward on which the incentive to create depends. And it is culturally wrong, because it glorifies the latecomer who takes and mocks the producer who makes, inverting the honour a healthy society pays to those who add to the common stock. The vocabulary in which this doctrine is usually dressed compounds the offence: it calls the person who reproduces a finished work a “competitor,” as though he had produced an alternative rather than copied an original, and it calls the person who made the thing a “monopolist,” as though defending one’s own creation were a kind of oppression. A society cannot flourish by calling the parasite a competitor and the creator a monopolist. The names are not innocent; they are the means by which appropriation is laundered into liberty and the producer is taught to feel ashamed of producing.

X. Why economics rejects envy dressed as policy

A flourishing society must care for the poor, the vulnerable, the sick, and the young, and must take seriously the questions of opportunity and stability on which a decent common life depends. Economics does not deny this; it insists on something the sentimentalist forgets, which is that the manner of the caring determines whether it helps or harms. And here a distinction must be drawn that is easy to blur and disastrous to lose: the distinction between compassion and envy.

Compassion looks at suffering and asks how to relieve it. Envy looks at success and asks how to pull it down. The two can wear the same political clothing and speak the same vocabulary of fairness, but they are opposites in their effect, because compassion seeks to raise the floor while envy seeks to lower the ceiling, and lowering the ceiling does nothing for those on the floor except to remove the thing whose existence might have lifted them. To take from producers because they produced more is not to create abundance; it is to punish the behaviour that made abundance possible, and a society that does so reliably gets less of it. The error is to imagine that wealth is a fixed pile to be redistributed rather than a flow to be produced, and that the flow will continue unchanged no matter how its sources are treated.

Disentangle, then, the things that envy-as-policy runs together. Charity is voluntary and moral — the free choice to give, which is among the highest expressions of a flourishing life and is corrupted, not perfected, when it is replaced by compulsion. Justice protects title and supplies remedy — it secures what is rightfully held and repairs what is wrongfully taken. Social responsibility operates within production and stewardship — the recognition that those who have built and hold owe something to the community whose institutions made the building possible. Confiscation, by contrast, attacks the source of production itself, and dressed though it may be in the language of any of the first three, it is none of them. The point is emphatically not that the weak should be abandoned; it is the reverse. A society that destroys its own productive capacity in the name of equality will have less with which to help anyone, and the poor it claimed to serve will be the first to discover that a smaller pie, however evenly divided, leaves everyone with less. Real compassion is therefore inseparable from a sober understanding of what produces the wealth that compassion would deploy — which is to say, compassion needs economics, lest it become, with the best intentions, a machine for manufacturing the poverty it set out to relieve.

XI. The role of law in moral markets

A persistent confusion sets law against liberty, as though every legal rule were a subtraction from freedom and the freest society were the one with the least law. The confusion dissolves once one sees that law is not one thing, and that the question is never “law or no law” but “what sort of law.”

Law can certainly destroy markets and crush flourishing. When it becomes arbitrary, so that no one can know in advance what is permitted; when it becomes retrospective, punishing yesterday’s lawful act by today’s decree; when it becomes politicised, bending to favour whoever holds power; when it becomes confiscatory, treating private holdings as a fund for redistribution — then law is indeed the enemy of liberty, and the economist is among its sharpest critics. But the same instrument, differently used, is the precondition of everything markets achieve. Law that protects property, enforces contracts, defines liability, governs evidence and title, supplies the corporate form and the rules of banking and insolvency and succession and remedy — this law does not constrain the market from outside; it constitutes the market from within, supplying the framework of enforceable expectation without which exchange among strangers cannot occur at all. There is no serious market without a legal order, just as there is no serious game without rules; the rules do not oppose the play, they make the play possible.

So the relevant distinction is between good law and bad. Good law is general, applying to everyone rather than singling out favourites; stable, so that people can rely on it; knowable, so that they can conform to it; prospective, governing the future rather than rewriting the past; and protective of property and contract, securing the conditions of action. Bad law is arbitrary, retrospective, politicised, discretionary, redistributive, or designed to reward a favoured group at the expense of the rest. This is Hayek’s distinction, and it is the heart of his answer to the false opposition of law and liberty: the rule of law is not opposed to freedom but is one of freedom’s conditions, because it is only under general, prospective, knowable rules — rules that bind the powerful as much as the weak — that a person can plan his own life rather than live at the discretion of those who hold power. Liberty, on this view, is not the absence of law; it is life under the right kind of law, and the project of a free society is not to abolish the courthouse but to confine it to the defence of the rules that make liberty possible.

XII. Economics and morality: neither can replace the other

We arrive at the synthesis the whole argument has been building toward, which is also the surest guard against dogma in either direction: economics and morality are distinct, each is indispensable, and neither can replace the other without producing a characteristic ruin.

Economics without morality becomes cold calculation — a technical mastery of means detached from any serious account of ends, efficient at producing whatever it is pointed at and silent on whether that thing is worth producing. It can optimise a process and have nothing to say about whether the process should exist; it can maximise a quantity and be indifferent to what the quantity is for. Pursued alone, it shrinks the human person to a node in an optimisation and forgets that the point of all the optimising was supposed to be a life worth living. Morality without economics becomes the opposite disaster — sentimental destruction, a generosity of intention untethered from any grasp of consequence, promising what cannot be delivered and reaching for coercion when reality declines to cooperate. It can will the end and ignore the means, declare the right and destroy the conditions that produce the thing the right is to, and leave behind, with a clear conscience, more suffering than indifference would have caused.

Figure 3. Human flourishing sits where moral ends meet economic understanding. Each discipline alone produces a characteristic ruin; only together do they sustain ordered liberty.

A serious civilisation needs both, held together. It needs moral ends — justice, compassion, the honouring of responsibility and creation — and it needs the economic understanding without which those ends cannot be pursued: the knowledge of incentives, the respect for scarcity, the discipline of consequence. It needs law, and it needs limits on law. It needs markets, and it needs the institutions and norms that keep dealing honest. Human flourishing sits exactly at this intersection, and Adam Smith embodied it before the disciplines were even fully separate: the man who wrote The Wealth of Nations also wrote The Theory of Moral Sentiments, which opens by observing that “how selfish soever man may be supposed, there are evidently some principles in his nature which interest him in the fortune of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it.” The founder of economics began from sympathy, not greed — and the lesson is that the two disciplines were never meant to be torn apart. Economics tells us that good intentions do not repeal scarcity. Morality tells us that efficiency is not the whole of life. Together they support a society in which people can create, trade, own, cooperate, and rise.

XIII. Austrian economics, properly understood

It is worth anchoring all this in the tradition that has thought hardest about it, without turning the essay into sectarian exposition — because the Austrian school is itself routinely mistaken for a cult of slogans, and recovering what it actually holds is part of recovering what economics is.

Austrian economics is not a creed of catchphrases. It is a tradition concerned with human action, subjective value, entrepreneurship, economic calculation, the role of time and the structure of capital, the dispersion of knowledge, and the institutional order within which all of these operate. Its two central results are not slogans but discoveries about how the social world works. Mises showed that without private property in the means of production there can be no market prices for capital goods, and without such prices rational economic calculation breaks down — so that a fully socialised economy is not merely inefficient but, in the strict sense, blind, unable to know whether it is creating wealth or destroying it. Hayek showed that the knowledge a society needs is dispersed beyond the reach of any central mind, and that evolved rules and institutions coordinate human action in ways no deliberate design could match — so that the conceit of comprehensive planning founders not on bad intentions but on an insurmountable epistemic limit.

Notice what these results do and do not imply, because the anti-institutional reading gets it precisely wrong. Neither position requires anarchism; both Mises and Hayek defended the limited but real state and the legal order it secures. Neither denies institutions; both make institutions central, Mises in his account of the calculating market and Hayek in his account of the rule of law and spontaneous order. Neither holds that intangible value is unreal or unimportant; Mises treats money, credit, expectations, and entrepreneurial judgment as central economic facts, and both regard the framework of property and contract — much of it intangible — as the precondition of prosperity. And neither supports the socialisation of created value; the whole thrust of the tradition is that secure property and the price signals it generates are what make calculation, coordination, and creation possible. The Austrian insight, stated without slogans, is this: a flourishing society needs not the absence of rules but rules of the right kind — rules that protect the conditions under which human beings can act, calculate, discover, and cooperate. That is the reverse of the doctrine that borrows the Austrian name to attack the institutional order. It is a defence of that order, properly limited, as the ground of everything else.

XIV. Why economics is not dogma

Gather the thread of the argument into the claim the title makes. Economics is not dogma, and the demonstration is now in hand.

Economics is not dogma because it does not ask for obedience to a slogan. It asks something harder and humbler: that human action be taken seriously, that scarcity be acknowledged rather than wished away, that institutions be examined by their actual effects rather than their advertised intentions, that incentives not be ignored because we would prefer they did not operate, that property be understood as a condition of agency rather than treated as an embarrassment, that production be recognised as prior to distribution because there is nothing to distribute that has not first been produced, and that law be assessed by whether it enables or destroys the conditions of flourishing. None of this is a creed demanding assent. All of it is an insistence on attending to reality, which is the precise opposite of dogma.

And so the indictment that brands economics a dogma can be returned to sender with the names corrected. Economics is not the enemy of human flourishing; bad economics is, and ideological economics is. Economics reduced to envy, to slogans, to the conceit of central planning, to anti-institutional fantasy, or to anti-property resentment — these are the enemies of flourishing, and they are enemies precisely because they are dogmatic, because they hold a conclusion fixed and refuse the correction reality keeps offering. Real economics is a discipline of humility before reality: it expects to be surprised, it builds in the feedback that tells it when it is wrong, and it asks of every rule and every policy not whether it flatters a faith but how it works in the lives of the human beings who must live under it. The systems that cannot survive that question are the dogmas. The discipline that insists on asking it is not.

XV. Conclusion: flourishing requires ordered liberty

Return, at the end, to the human person, because that is where the argument began and where it belongs.

A person flourishes when he can think and work and build, when he can own what he has made and exchange it freely, when he can teach and create, protect his family and enter into binding agreements, preserve the fruits of his labour and develop his skill, and pass something better to those who come after him. None of this happens in a void. It happens inside an order — an order of law that is general and stable rather than arbitrary, of property secure enough to support planning, of trust underwritten by enforceable promise, of money sound enough to preserve calculation, of markets open enough to reward service, and of moral restraint serious enough to honour creation and discipline greed. Take away that order and the person does not become freer; he becomes the prey of whoever is stronger, and the long, patient, future-directed work that builds civilisation gives way to the short horizon of force and favour.

Economics is not dogma because it is not a substitute for life, and was never meant to be one. It is a map of the constraints and consequences within which human life is actually lived — a map that shows why slogans fail, why institutions matter, why property matters, why creators matter, and why no civilisation has ever been built, or can be, on appropriation dressed as freedom. It does not tell us what to live for. It tells us, with a rigour that humbles every planner and every ideologue, what it takes to make a world in which living for something is possible. Human flourishing is not produced by dogma. It is produced by free persons acting within a moral and institutional order that protects creation, rewards responsibility, disciplines error, and allows civilisation to compound across generations.

References

Austrian primary sources

• Mises, Ludwig von. Human Action: A Treatise on Economics. 3rd rev. ed. Chicago: Henry Regnery, 1966. (Human action and subjective value: “economics is not about things and tangible material objects… it is about men, their meanings and actions.” The state as indispensable and the rejection of anarchism. Economic calculation: without private ownership of the means of production there are no market prices for capital goods, and rational allocation breaks down.)

• Mises, Ludwig von. “Economic Calculation in the Socialist Commonwealth.” 1920. Reprinted, Auburn, AL: Ludwig von Mises Institute, 1990. (The original statement of the calculation problem.)

• Hayek, F. A. The Fatal Conceit: The Errors of Socialism. Ed. W. W. Bartley III. Chicago: University of Chicago Press, 1988, p. 76. (”The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.”)

• Hayek, F. A. “The Use of Knowledge in Society.” American Economic Review 35, no. 4 (1945): 519–530. (Dispersed knowledge; the price system as a discovery and coordination mechanism.)

• Hayek, F. A. The Constitution of Liberty. Chicago: University of Chicago Press, 1960. (The rule of law as a condition of liberty; coercion; “it is not the source but the limitation of power which prevents it from being arbitrary.”)

• Hayek, F. A. Law, Legislation and Liberty. 3 vols. Chicago: University of Chicago Press, 1973–1979. (Spontaneous order; nomos and thesis; the legal order as the institutional precondition of the market; the delimitation of protected domains.)

• Hayek, F. A. The Road to Serfdom. Chicago: University of Chicago Press, 1944, ch. 7. (”The system of private property is the most important guarantee of freedom, not only for those who own property, but scarcely less for those who do not.”)

• Hayek, F. A. “The Pretence of Knowledge.” Nobel Memorial Lecture, 1974. (The lesson of humility before the insuperable limits of knowledge.)

Classical foundations

• Smith, Adam. The Theory of Moral Sentiments. 1759. (Book I, ch. 1, opening: “How selfish soever man may be supposed, there are evidently some principles in his nature, which interest him in the fortune of others…” Part VI, the “man of system” who “cannot suffer the smallest deviation” from his ideal plan.)

• Smith, Adam. An Inquiry into the Nature and Causes of the Wealth of Nations. 1776. (Markets channelling private interest to public benefit; the division of labour.)

• Aristotle. Nicomachean Ethics. (Eudaimonia — flourishing as activity in accordance with virtue — as the proper end of human life; the background to the account of flourishing developed here.)

Note on method and scope. Every quotation attributed to Hayek, Mises, or Smith reflects the verified wording of the cited work, checked at the passage or page level (Hayek, Fatal Conceit, p. 76; Smith, Theory of Moral Sentiments, Book I ch. 1 and Part VI), not drawn from an abstract or secondary gloss; Aristotle’s eudaimonia is invoked as the classical background and paraphrased rather than quoted. This essay argues a constructive thesis — that economics is a non-dogmatic, consequence-attentive science of human action, that human flourishing is a condition of ordered liberty resting on specific institutions, and that economics and morality are jointly necessary and mutually irreplaceable. It defends limited government and the institutional order, criticises socialist, anarchist, anti-intellectual-property, technocratic, and market-absolutist dogma alike, and treats “dogma” in the strict sense of the refusal of correction rather than as a synonym for principle. Where it characterises a position — the calculation problem, the dispersed-knowledge argument, the structure of the anti-IP case — the characterisation reflects the verified content of the sources cited, and any contested empirical claim about the effects of particular policies is presented as a tendency grounded in incentives, not as a settled quantitative result.

Keywords: intellectual property, Friedrich Hayek, rule of law, spontaneous order, nomos, institutions, limited government, Mises, socialisation, monopoly, competition, appropriation, positive theory of property, delimitation, copyright, patents.

Abstract

This series has removed a costume and supplied an economics. The first essay showed that the anti-intellectual-property argument is an anarchist politics wearing the name of Austrian economics, and that Mises and Hayek were not anarchists. The second showed that Mises diagnosed an external-economy problem and called the legal response a question of the delimitation of property rights, not a verdict of illegitimacy, and that all developed property is institutional, so the charge “artificial because it requires law” indicts the whole architecture of ownership. This third and final essay takes up the move the abolitionist has left in reserve — the appeal to statelessness, the claim that because these rights depend on the state specifically they are illegitimate, and that a stateless order could secure complex property and exchange without them — and it shows that Hayek’s life work is the standing refutation of exactly that move. Hayek’s enemy was never order, law, or the state; it was planned command substituted for evolved order — central direction in place of the coordination achieved by general rules and market prices. He defended the rule of law as the institutional precondition of the market, holding that a free order is one in which all are “restrained only by rules of just conduct of universal application,” and that power is held non-arbitrary not by its democratic source but by “the limitation of power.” His distinction between grown law (nomos) and made legislation (thesis) is a distinction within legality, not a rejection of it; and his account of the judge’s task — the protection of expectations through “the delimitation of protected domains” — uses the very word Mises used, and locates the legal task exactly where this series has located it. The anarchist collapses Hayek’s central distinction, treating all legal recognition as if it were central command, and concludes that the courthouse itself is the enemy. It is not. Small government is not no government: Mises held the state to be the indispensable apparatus that secures peace, and a market order without enforceable title, contract, and adjudication is not a market but possession defended by private force. The essay then exposes the collectivist core the abolitionist’s vocabulary conceals. The argument has a structure — deny the creator’s exclusive claim, declare the thing too useful to be controlled, rename appropriation as freedom, recast enforcement as oppression, and elevate collective access above the producer’s title — and it is the structure of every socialisation of property, regardless of the advocate’s intent. Once “usefulness defeats ownership” is admitted, nothing stops the same logic from reaching the house, the factory, the medicine, and the field; it is socialism at the point of value, private effort to create and collective entitlement once created. The two favourite slogans fall with it: “state-granted monopoly” proves too much, because all property excludes and the word “monopoly” merely makes ordinary exclusion sound sinister; and “copying is competition” is false, because the competitor produces an alternative while the copyist reproduces a finished form, arriving after the uncertainty has been conquered by someone else to announce that imitation is liberty. The positive theory follows naturally: intellectual property is institutionalised control over created value, one delimited form among many, differing in its incidents from land and shares and debt exactly as they differ from one another, and no more a counterfeit for having its own incidents than a lease is a counterfeit for being temporal. The conclusion returns to the series’ thesis. Mises did not reject intangible value because it was intangible; Hayek did not reject institutions because they were legal. Austrian economics is a theory of markets, property, calculation, and the institutional order that makes them possible. The anti-IP anarchism that borrows its name is not a purer Austrianism but a corruption of it — a doctrine that takes the critique of socialism and turns it into an argument for socialising the work of creators, then calls the theft freedom.

I. The last move: the appeal to statelessness

The first two essays disposed of the economic substance of the anti-intellectual-property argument. The category-claim — that intangible, non-rival, legally constituted value cannot be property — was shown to be foreign to Austrian subjectivism, to invert Mises’s actual diagnosis, and to indict all developed property if applied consistently. But the abolitionist has one move left, and it is the move that turned the debate political in the first place. It is the appeal to statelessness: the claim that because patents and copyright depend on the state for their recognition and enforcement, they are creatures of political power rather than genuine rights, and that a stateless order could secure complex property and exchange without them.

This move is what the anarchism in “anti-IP anarchism” actually consists of, and it is worth meeting directly, because it is the only part of the argument that is genuinely about the state rather than about intellectual property in particular. And the answer to it is not some novel rejoinder; it is Hayek, whose entire mature project is a sustained demonstration that the rule-bound institutional order the anarchist treats as the enemy is in fact the precondition of everything the anarchist claims to want. To appeal to statelessness against intellectual property, while invoking the authority of Hayek, is to enlist against the courthouse the one thinker who spent a career explaining why the courthouse must stand.

II. Hayek: the rule of law is the precondition, not the enemy

The decisive thing to understand about Hayek — and the thing the anti-IP writer must obscure to claim him — is what Hayek opposed. He did not oppose order. He did not oppose law. He did not oppose the state as such. What he opposed was a specific thing: planned command substituted for evolved order — the discretionary direction of resources by a central authority in place of the impersonal coordination achieved by general rules and market prices. The whole of his political philosophy turns on the difference between a general rule, which applies to everyone and enables each person to use his own knowledge for his own purposes, and a specific command, which directs particular people to particular ends chosen by the authority. The first is the substance of the rule of law; the second is the substance of central planning. Hayek was the great enemy of the second and the great defender of the first.

This is not a reconstruction; it is his explicit thesis. The condition of liberty he defends, he wrote, is one in which all are allowed to use their knowledge for their purposes, restrained only by rules of just conduct of universal application — and such a system, he added, can be achieved and maintained only if all authority, including that of a majority, is limited in the exercise of coercive power by general rules. Liberty, for Hayek, is not the absence of rules; it is life under general, abstract, universal rules rather than under particular commands. And the limitation of power, not its source, is what keeps it from being arbitrary: there is no justification, he insisted, for the belief that so long as power is conferred by democratic procedure it cannot be arbitrary, because it is not the source but the limitation of power that prevents arbitrariness. A legal order that subjects everyone, the state included, to general rules known in advance is, for Hayek, the very architecture of freedom.

Figure 1. Hayek’s enemy was command, not order. He defended the rule of law against both lawlessness and central planning; the anarchist collapses the two and indicts the law itself.

Hayek’s most developed jurisprudence makes the point sharper still, and it ties this whole series together. In Law, Legislation and Liberty he distinguishes grown law (which he calls nomos) from made legislation (thesis), and spontaneous order (cosmos) from constructed order (taxis). The distinction the anti-IP writer needs Hayek to be making — between legitimate stateless custom and illegitimate state law — is not the distinction Hayek draws at all. Hayek’s distinction is within legality: between law that evolves through the resolution of disputes and the protection of expectations, and legislation that is deliberately constructed. Both are law; Hayek’s worry is the colonisation of the first by the second, not the abolition of either. And crucially, the legal order in Hayek is the institutional precondition of the market order: the framework of rules and the practices of property and exchange are what allow the spontaneous order of the market to get off the ground in the first place. The market does not precede law and then suffer it; the market is constituted by a framework of legal rules, without which it cannot exist. Adam Ferguson’s observation, which Hayek made his own, was that institutions are “the result of human action, but not the execution of any human design” — they are grown, not made. But grown is not the same as absent. A grown legal order is still a legal order, and the rules of property are precisely the grown institutions that Ferguson and Hayek had in mind.

And here is the detail that closes the circle with the second essay. Hayek describes the function of the judge, and of law itself, as the protection of expectations — and he says that the maximal coincidence of expectations is achieved by “the delimitation of protected domains.” Delimitation. It is the exact word Mises used when he called intellectual property “a problem of the delimitation of property rights.” Mises, analysing the economics of a created good, said the legal task was to delimit the rights over it. Hayek, analysing the function of law as such, said the legal task was to delimit protected domains so that expectations could coincide and people could plan. The two giants of the school the anti-IP writer invokes both located the legal task in exactly the same place — the delimitation of protected domains over valuable things — and intellectual property is one instance of that task, not an exception to it. The anarchist who cites Hayek against the courthouse has not read the chapter in which Hayek explains that the courthouse exists to do the very thing the anarchist says cannot legitimately be done.

III. Small government is not no government

The appeal to statelessness trades on a conflation that a single distinction dissolves: the conflation of limited government with no government, on the ground that both oppose socialism. They do both oppose socialism. They are not the same thing, and treating them as the same is the central error of the anarchist’s political case.

Mises and Hayek defended limited government: a state bound by law, restrained in its powers, forbidden to plan and command, but real, and charged with the indispensable functions of recognising title, enforcing contracts, adjudicating disputes, protecting property, and restraining violence. Mises was explicit that this is not a grudging concession but a requirement of social cooperation. The state, he held, is the social apparatus of compulsion and coercion, it holds the monopoly of violent action, and it is indispensable — the anarchist, he wrote, overlooks the undeniable fact that some people are too narrow-minded or too weak to adjust themselves spontaneously to the conditions of social life, and even setting those aside, there remain the infants, the aged, and the insane; a society that wishes to preserve peace must be prepared to restrain those who would break it, and the institution that does so is the state. This is limited government — a courthouse that knows its limits — and it is the opposite of both socialism and anarchism.

The distinction matters for intellectual property because the abolitionist treats every appeal to a legal institution as a step toward the planning state. It is no such thing. A court that adjudicates whether one party has copied another’s work is not a central planner directing the economy; it is performing exactly the dispute-resolving, expectation-protecting function that Hayek identified as the heart of the rule of law and that Mises identified as the indispensable work of the limited state. A classical-liberal order is not an empty field on which the strong take what they can hold; it is a framework of enforceable rules within which people may use their own knowledge for their own purposes. Small government is not an empty courthouse. It is a courthouse that knows its limits. A market order without enforceable property, contract, and adjudication is not a freer market; it is not a market at all. It is possession, retaliation, private force, and clan enforcement — the very condition that the institutions of property and law exist to escape. The abolitionist’s stateless order does not liberate exchange. It returns it to the world before law, in which there is no created form to own because there is no settled way to own anything.

IV. The collectivist core: socialism at the point of value

Now expose what the abolitionist’s vocabulary conceals, because beneath the language of freedom the anti-intellectual-property argument has a structure, and the structure is collectivist regardless of what its advocates intend or call themselves.

Lay the structure out in sequence. First, deny the creator’s exclusive claim to the thing he made. Second, justify the denial by appeal to usefulness: the thing is too useful, too important, too needed by others to be left under one person’s control. Third, rename the appropriation that follows as “freedom” — the freedom of others to use what they wish. Fourth, recast the creator’s attempt to enforce his claim as “oppression,” “monopoly,” “aggression.” And fifth, elevate the public’s desire for access above the producer’s title, so that the collective appetite for the thing becomes the ground for overriding the right to it. That is the argument, stripped of its vocabulary. And it is, step for step, the structure of every socialisation of property that has ever been proposed.

Figure 2. The anti-IP argument’s structure is the structure of every socialisation of property. Once usefulness defeats ownership, the same logic reaches the house, the factory, and the field.

The point is not that the anti-intellectual-property advocate is a secret socialist; many are sincere libertarians who would be appalled at the comparison. The point is structural, and it is more damning for being structural: the form of the argument is collectivist whatever the intent behind it. Each of its five moves can be applied, without modification, to any valuable thing that others want. Deny the landlord’s exclusive claim because housing is too important to be left under private control; rename the expropriation as the tenants’ “freedom”; recast the landlord’s enforcement of his title as “oppression”; elevate the collective need for shelter above his ownership. Deny the manufacturer’s claim to the factory because production is too important; deny the farmer’s claim to the field because food is too important; deny the inventor of a medicine his claim because health is too important. The logic that abolishes intellectual property because created forms are useful and non-rival is the same logic that abolishes every form of property the moment the public’s appetite for the underlying good is asserted as superior to the owner’s title. Once usefulness defeats ownership, all property is conditional on collective appetite. That is the principle the anti-IP argument introduces, and a principle with that shape does not stay confined to patents and copyright. It is socialism at the point of value: private effort to create, and collective entitlement the moment the creation proves useful to others.

V. The two slogans, dismantled

Two slogans carry most of the anti-intellectual-property argument’s rhetorical weight, and both fail under inspection.

The first is “state-granted monopoly.” The word “monopoly” is doing illegitimate work, because all property excludes, and exclusion is not monopoly. The owner of a house exclusively controls that house; the owner of a parcel exclusively controls that parcel; the owner of a share has exclusive rights attached to that share. None of these is called a monopoly, because we understand that owning a defined thing and excluding others from that thing is simply what property is. A patent confers exclusive rights over a defined, disclosed invention for a defined term; it does not confer control of an entire market, and it does not prevent others from inventing differently. A copyright protects a particular expression; it does not stop anyone from writing his own book on the same subject. A trade mark prevents confusion as to source; it does not prevent trade. The slogan collapses ordinary, defined exclusion — the essence of all property — into “monopoly,” a word chosen to make the normal sound sinister. And the collapse proves too much in the now-familiar way: if exclusion is monopoly, then property is monopoly, and the anti-IP argument has become an argument against property as such. The owner of land is then a “monopolist” of his field, the owner of a house a “monopolist” of his home. An argument that can only condemn intellectual property by condemning all property has not isolated its target.

The second slogan is “copying is competition,” and it is false because it conflates two different acts. A competitor produces an alternative: a rival author writes a better book, a rival inventor develops a different machine, a rival firm builds a competing system. That is competition, and it is the engine of the market — the discovery, through rivalry, of better ways of serving human wants. A copier does something else entirely. He takes the completed form another produced and reproduces it. He does not offer an alternative; he offers the same thing, made by someone else, while the originator who bore the cost of producing it is stripped of the ability to recover that cost. The asymmetry is the one this series has drawn throughout: the creator acts before the outcome is known, under uncertainty, bearing the risk and the cost of production; the copier acts after, once the valuable form has been revealed and its market demonstrated, bearing only the small cost of reproduction and no risk at all. The copyist is not the heroic entrepreneur of Austrian theory. He is the man who waits until entrepreneurial uncertainty has been conquered by someone else, and then announces that imitation is liberty. To call this competition is to flatter the copier with a name that belongs to the rival who actually produces something new.

VI. The positive theory: institutionalised control over created value

Having cleared the slogans, state the positive account plainly, because the anti-IP writer’s deepest rhetorical advantage is the suggestion that the defender of intellectual property has no theory of it beyond “the state says so.” He does. Intellectual property is institutionalised control over created intangible value: a delimited legal right, constituted in the same way every developed property right is constituted, fitted to an asset — a created form — whose character differs from land and from chattels, and whose incidents therefore differ too.

That its incidents differ is not an objection; it is the ordinary condition of property. No two property forms share the same incidents. Land can be perpetual and spatial; a lease is temporal; a mortgage is conditional and exists to secure a debt; a company share is corporate and confers a residual claim; a debt is relational and is an enforceable claim against a person; a trust splits legal from beneficial ownership. Each of these is unquestionably property, and each has incidents the others lack. A patent is limited in term, requires disclosure, and is defined by a technical claim; a copyright is expressive and bounded by the idea/expression distinction; a trade mark turns on commercial identity and the prevention of confusion; a trade secret rests on confidentiality and the wrong of its breach; goodwill is reputational. These differ from land — and from one another — exactly as a lease differs from a freehold and a share differs from a debt.

Figure 3. Different incidents, one institution of property. Intellectual property differs from land in its incidents exactly as a lease, a share, and a debt do — and no more counts as counterfeit for it.

So the demand that intellectual property prove itself genuine by sharing the incidents of land — by being perpetual, tangible, rivalrous — is a demand that no property form can meet, because no property form shares all the incidents of any other. Different incidents do not destroy property. They reveal legal architecture fitted to different forms of value. A proper Austrian treatment of intellectual property does not ask whether it resembles land; it asks the questions that fit an institution of delimited rights over created value: does the right support private ordering and contracting; does it internalise to the creator the value he would otherwise produce as an external economy; does it operate through general rules rather than discretionary privilege; does it avoid the overbroad claim, the excessive term, the rent-seeking abuse that a delimited right can degenerate into? Those are real questions, and they admit real criticism — of bad patents, of long terms, of abusive litigation. They are questions of delimitation, of where to draw the boundary, which is exactly where Mises and Hayek both located the legal task. They are not the anarchist’s question, which is not “where should the boundary be drawn” but “there is no boundary to draw, because the thing is intangible and the state recognises it.” That question, this series has shown, is not Austrian economics. It is an anarchism that has borrowed the name.

VII. Conclusion: the costume, the courthouse, and the back door

Stand back and see the whole shape of it. The anti-intellectual-property argument presents itself as the consistent terminus of Austrian economics, and it is the opposite — a corruption of Austrian themes, assembled by grafting an anti-institutional anarchism onto a tradition that never contained it.

Mises did not reject intangible value because it was intangible. He saw, more precisely than the abolitionist, that the services of a disclosed creation are inexhaustible, and he saw the other half too — that the creation must be produced at a cost — and he named the result the extreme case of external economies and called the legal response a question of the delimitation of property rights. He took no side on the answer, as the abolitionist’s own footnote admits. Hayek did not reject institutions because they were legal. He spent a career showing that the rule of law, the grown order of general rules, the delimitation of protected domains, is not the enemy of the market but the precondition without which there is no market at all. Between them they located the legal task in the same place and gave it the same name. Neither was an anarchist. Neither held that a right is illegitimate because a state recognises it. Both held that legal order is what makes property and exchange possible in the first place.

The anti-IP anarchist position is therefore not a purer Austrianism. It is a hollowing-out of Austrian economics and a refilling of the shell with something foreign to it. It takes the Austrian critique of socialism — the demonstration that abolishing private ownership of the means of production destroys the calculation on which civilisation depends — and it turns that critique into an argument for socialising the work of creators: for denying the producer’s exclusive claim, declaring his creation too useful to control, renaming the appropriation freedom, recasting his enforcement as oppression, and elevating the collective appetite above his title. It says the creator may labour privately but must surrender control the moment his product becomes useful to others. That is not the economics of Mises. It is not the jurisprudence of Hayek. It is not Austrian economics at all.

It is collectivism in a black-and-gold costume, selling socialism through the back door while calling the theft freedom. And the answer to it is the one the two greatest economists of the tradition it impersonates would have given: that markets require property, that property requires institutions, that institutions require law, and that the sentence on which all of it rests — this is mine; you may not take it without my consent — is not abolished but destroyed when the created form is declared free for the taking the instant it proves worth taking. The courthouse the anarchist would tear down is the building in which that sentence is kept. Mises and Hayek knew what it was for. The copyist’s socialism does not — or pretends not to, which in the end is the same thing.

References

Hayek, primary

• Hayek, F. A. The Constitution of Liberty. Chicago: University of Chicago Press, 1960. (Liberty as life under general rules; coercion as one person’s actions made to serve another’s will; “it is not the source but the limitation of power which prevents it from being arbitrary”; civilisation made possible by the principles of liberty; dispersed knowledge.)

• Hayek, F. A. Law, Legislation and Liberty. 3 vols. Chicago: University of Chicago Press, 1973–1979. (Nomos/thesis and cosmos/taxis; the thesis that a free order is one in which all use their knowledge for their purposes “restrained only by rules of just conduct of universal application,” with authority “limited in the exercise of coercive power by general rules”; the judge’s task and “the delimitation of protected domains”; the legal order as the institutional precondition of the market.)

• Hayek, F. A. The Road to Serfdom. Chicago: University of Chicago Press, 1944. (The rule of law versus central direction.)

• Hayek, F. A. “The Use of Knowledge in Society.” American Economic Review 35, no. 4 (1945): 519–530. (The price system as coordinator of dispersed knowledge.)

• Ferguson, Adam. An Essay on the History of Civil Society. 1767. (Institutions as “the result of human action, but not the execution of any human design” — the observation Hayek adopted.)

Mises, primary

• Mises, Ludwig von. Human Action: A Treatise on Economics. 3rd rev. ed. Chicago: Henry Regnery, 1966. (The state as the indispensable apparatus of compulsion; the rejection of anarchism, “too narrow-minded or too weak to adjust themselves spontaneously to the conditions of social life”; intellectual property as “a problem of the delimitation of property rights” and the external economies of creators, p. 661.)

The anti-IP position under examination

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (2001): 1–53. (n. 38: “Mises expressed no opinion on the issue.”)

• Hoppe, Hans-Hermann. A Theory of Socialism and Capitalism. Boston: Kluwer, 1989. (The anarcho-capitalist, physical-integrity-only framework — distinct from Austrian economics.)

The institutional and economic background

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” NBER, 1962. (Inappropriability.)

• Landes, William M., and Richard A. Posner. “An Economic Analysis of Copyright Law.” Journal of Legal Studies 18, no. 2 (1989): 325–363.

• De Soto, Hernando. The Mystery of Capital. New York: Basic Books, 2000. (Property as a legal-institutional construction; dead capital.)

Note on method and scope. Every quotation attributed to Hayek or Mises reflects the verified wording of the cited work, checked at the passage level, not drawn from an abstract or secondary gloss. Ferguson’s phrase is attributed to Ferguson, as the observation Hayek adopted, not to Hayek. This is the third and final essay of the series. It argues that Hayek’s rule-of-law jurisprudence refutes the appeal to statelessness, that small government is not no government, that the anti-IP argument’s structure is collectivist regardless of intent, that the “monopoly” and “competition” slogans fail, and that intellectual property is one delimited form of property among many. The series criticises the abolitionist category-claim and the anarchist political case, not any particular statute, term, or scope; it concedes throughout the legitimacy of Austrian criticism of overbroad or abusive intellectual-property regimes, which is a question of delimitation rather than of existence. Where the essay reports what a source argues — Mises’s diagnosis, Hayek’s distinctions, Kinsella’s concession — the report reflects the verified content of that source, and Hoppe is identified as an anarcho-capitalist, because it is the anarchism and not Austrian economics that generates the abolitionist conclusion.

Keywords: intellectual property, Ludwig von Mises, external economies, public goods, non-rivalry, inexhaustibility, delimitation of property rights, appropriability, Stephan Kinsella, institutional property, De Soto, dead capital, copyright, patents.

Abstract

The first essay in this series removed the costume: it showed that the anti-intellectual-property argument is an anarchist politics wearing the name of Austrian economics, and that Mises and Hayek were not anarchists. This second essay takes up the economic substance, and it begins where the abolitionist most wants to claim Mises and least deserves to. Mises did discuss patents and copyright, in Human Action, and the abolitionist treats this discussion as an endorsement. It is nothing of the kind. What Mises did was identify a problem with a precise structure. The services of an intellectual creation — a formula, a recipe, a design, a text once disclosed — are inexhaustible: a known formula, in his words, renders unlimited services, and does not lose anything of its capacity to produce however often it is used, so that in this respect it is not an economic good at all but a free good. But the production of that creation is costly, and the cost is borne by one person, before disclosure, under uncertainty. Put the two together and you have what Mises called the extreme case of external economies: the creator produces a benefit that others enjoy without paying for it, because once the creation is disclosed its non-rival services flow freely to all. Mises’s conclusion from this was not that intellectual property is illegitimate. It was that the matter is “a problem of the delimitation of property rights,” and that with the abolition of patents and copyright authors and inventors would, for the most part, become producers of external economies. He drew the economic implication and, as Kinsella’s own footnote concedes, expressed no opinion on the legal answer. The abolitionist converts this neutral diagnosis into a dogma — non-rival, therefore never property — that Mises never stated and that does not follow from anything he said. The essay then develops the moral and economic consequence Mises’s framing exposes: abolishing intellectual property does not remove cost or coercion from the world; it relocates them, forcing the originator to internalise the cost of creation while the imitator externalises the benefit, so that the creator becomes an unpaid producer for others under the banner of “freedom.” Finally, the essay turns the abolitionist’s central slogan — that intellectual property is “artificial” because it depends on the state — against the whole of property. All developed property is institutional. A field exists in nature; ownership of the field does not. Land title, mineral rights, water rights, mortgages, company shares, debts, negotiable instruments, and trusts are every one of them institutional layers laid over a natural substrate, none of them found lying in the world, all of them constituted by law. Intellectual property slots into precisely the same architecture. The objection “it requires law, therefore it is artificial” does not isolate patents and copyright; it indicts land title and the share certificate and the mortgage with exactly equal force, and an argument that abolishes all property to reach intellectual property has refuted itself, not its target.

I. The passage the abolitionist wants, and what it actually says

The anti-intellectual-property writer reaches for Mises early, because Mises is the most authoritative name available to him and because Mises did, unlike many economists, address the question directly. The reach is understandable. It is also a misreading, and the misreading is worth correcting in detail, because once Mises’s actual structure is on the table the abolitionist’s use of him collapses.

Here is what Mises actually did. In the part of Human Action dealing with the limits of property rights and the problems of external costs and external economies, he observed that certain things have a peculiar economic character: their services are inexhaustible. A formula, a recipe, a piece of technological knowledge, once it is known, can be used without limit and without depletion. Mises’s own example is the recipe for preparing coffee: a thing rendering such unlimited services, he wrote, is the knowledge of the causal relation it embodies; the recipe, provided it is known, renders unlimited services, does not lose anything of its capacity to produce however often it is used, has an inexhaustible productive power, and is therefore not an economic good. Such recipes, he said, are as a rule free goods, because their ability to produce definite effects is unlimited. This is the non-rivalry point that the abolitionist treats as decisive — and Mises stated it first and more precisely than the abolitionist does.

But Mises did not stop where the abolitionist stops, and the continuation is everything. The reason intellectual creation poses a problem rather than simply a happy abundance of free goods is that the inexhaustible services have to be produced, and producing them is costly. The formula is a free good once it exists and is known; bringing it into existence is not free at all. Mises identified this directly as the extreme case of external economies: the characteristic mark of the formula, he wrote — the mental device directing a technological procedure — is the inexhaustibility of the services it renders, and the extreme case of external economies is shown in the production of the intellectual groundwork of every kind of processing and constructing. The creator bears the cost of producing something whose services, once disclosed, flow without charge to everyone. That is the external economy: a benefit conferred on others that the producer cannot capture.

Figure 1. Mises’s diagnosis, not a verdict. Inexhaustible services plus costly production yield an external economy; Mises called the legal response a question of delimitation and took no side on it.

And now the decisive sentence, the one the abolitionist never quite quotes in full. Having laid out the external-economy structure, Mises drew his conclusion, and his conclusion was a careful refusal to draw the abolitionist’s. The matter, he said, is “a problem of the delimitation of property rights,” and with the abolition of patents and copyright, authors and inventors would for the most part be producers of external economies. That is the whole of it. He did not say patents and copyright are illegitimate monopolies. He did not say non-rival goods can never be property. He said the question is one of delimitation — of where a legal order should draw the boundary of property rights around a good whose services are inexhaustible — and he noted the economic consequence of drawing it one way rather than another. Kinsella himself, in a footnote to Against Intellectual Property, concedes exactly this: Mises, he writes, expressed no opinion on the issue, merely drawing the economic implications from the presence or absence of such laws. The abolitionist’s own text admits that Mises did not hold the abolitionist position.

So the structure of the abolitionist’s appeal to Mises is precisely backwards. He takes a thinker who identified a problem and declined to prescribe a solution, and he presents him as having prescribed the abolitionist’s solution. He takes “this is a question of where to draw the boundary” and reports it as “there is no boundary to draw.” This is not interpretation; it is inversion. Mises diagnosed; the abolitionist forges a verdict and signs Mises’s name to it.

II. Why “non-rival, therefore not property” does not follow

It is worth pausing on the inference the abolitionist actually needs, because stated plainly it is a non-sequitur, and Mises’s framing is what exposes it.

The abolitionist argues: the services of a disclosed creation are non-rival; non-rival goods cannot sensibly be owned, because ownership is a device for allocating things that cannot be used by everyone at once; therefore intellectual property is not property. The hidden premise is the middle one — that non-rivalry is disqualifying, that a good whose services are inexhaustible is for that reason ineligible to be the object of a property right. But nothing establishes that premise, and Mises’s analysis quietly denies it. Mises says the services of the formula are inexhaustible and that in respect of those services the formula is a free good. He does not say that the creation cannot be the object of a delimited legal right. On the contrary, by framing the matter as one of delimitation, he treats it as an open question whether and how a legal order should define such a right — which presupposes that defining one is at least possible and possibly desirable, not that it is conceptually ruled out.

The reason non-rivalry does not disqualify is that property rights do not exist only to ration scarce services among competing users. They also exist to secure to a producer the return on a costly act of production, so that the act will be undertaken. Rationing is one function of property; appropriability is another. A good can be non-rival in its services and still pose an appropriability problem in its production, and it is the second problem, not the first, that intellectual property addresses. The abolitionist’s inference works only if rationing is the sole legitimate function of property — only if a right that secures appropriability rather than rationing scarce use is somehow not a real property right. But that is exactly the question at issue, and it cannot be settled by definition. Mises did not settle it by definition. He named both sides of the problem — the inexhaustible services and the costly production — and left the delimitation to be worked out. The abolitionist settles it by definition, in his own favour, and calls the result Austrian.

III. The creator as unpaid producer for others

Develop now the consequence that Mises’s external-economy framing makes unavoidable, because it disposes of the most seductive line the abolitionist has: that abolishing intellectual property simply restores market freedom, removing a coercive state-granted privilege and letting exchange flow unobstructed. Mises’s analysis shows that this is false. Abolition does not remove cost or coercion from the world. It relocates them.

Consider what the producer of an intellectual creation actually does, and when. An inventor invests time, capital, experiment, repeated failure, and finally disclosure — and the disclosure is the very act that makes the knowledge available to others. An author invests labour, thought, research, structure, and the long uncompensated period of composition. A designer invests creative judgment and technical refinement; a developer invests architecture, implementation, testing, and documentation. Every one of them bears the cost of production before the outcome is known, before the market has revealed whether the work will succeed, under the full weight of uncertainty. And every one of them acts first. The copier, by contrast, arrives afterwards. He acts only once the uncertainty has been resolved by someone else, once the valuable form has been revealed and its market demonstrated. He bears the cost of reproduction, which is small, not the cost of production, which was large; and he bears no risk, because the risk was discharged by the creator before he arrived.

Figure 2. The creator as unpaid producer for others. Abolition relocates cost rather than removing it: the creator internalises the cost of production while the imitator externalises the benefit.

This is the point at which the abolitionist’s vocabulary does its concealing work, and Mises’s vocabulary undoes it. The abolitionist calls the copier’s freedom to reproduce “the free market” and the creator’s wish to control reproduction a “state-granted monopoly.” But what is actually happening, in Mises’s terms, is that abolition forces the originator to internalise the cost of creation while permitting the imitator to externalise the benefit. The creator becomes a producer of external economies — a person who bears a private cost and confers a public benefit he cannot capture. There is no neutral, coercion-free baseline here that abolition simply restores. There is a choice between two rules, each of which determines who may take from whom. Under a regime of delimited rights, the creator may exclude the copier from the created form for a defined period and thereby capture some of the return. Under abolition, the copier may take the created form from the creator the moment it is disclosed. Abolishing intellectual property does not abolish coercion from the world. It changes who may take: the copier from the creator, under the moral perfume of “freedom.” Mises’s external-economy framing is what lets us see this clearly, because it identifies the loss — the uncompensated benefit conferred on others — that the abolitionist’s vocabulary is designed to hide.

None of this, it must be said carefully, settles the policy question by itself. That abolition makes creators producers of external economies does not prove that any particular patent term is optimal, or that copyright should last as long as it does, or that the existing statutes draw the boundary in the right place. Mises’s point, and this essay’s, is narrower and more durable: it destroys the glib claim that abolition is a simple restoration of market freedom. It is not. It is a rule choice that reallocates the value of created forms away from those who create them and toward those who copy them, and it must be defended as such, on its consequences, not smuggled in as the mere absence of an artificial interference.

IV. The institutional foundation of all property

Now turn the abolitionist’s central slogan against the whole of property, because the slogan — that intellectual property is “artificial” because it depends on the state for its existence — is the move that, followed consistently, destroys far more than its author intends. This is the deepest lesson of Mises’s word “delimitation,” because delimitation is an institutional act, and all property requires it.

Begin with the thing the abolitionist treats as the paradigm of natural, pre-political property: land. A field is physical; you can stand on it, plough it, fence it. But ownership of the field is not physical and is not found in nature. The field exists in the world; the title does not. Ownership of land requires boundaries that someone has surveyed and recorded, a system of registration, rules of priority between competing claimants, mechanisms of conveyance and inheritance, remedies for trespass and dispossession, and an apparatus of enforcement. Strip all of that away — strip away the survey, the registry, the court, the sheriff — and you do not have natural property in land; you have a field that the strongest or the most numerous can occupy until someone stronger or more numerous arrives. What converts the physical field into owned property is an institutional layer laid over the natural substrate, and that layer is constituted by law.

And what is true of land is true of every developed form of property, without exception. Mineral rights require a legal separation of the subsurface estate from the surface. Water rights require allocation rules. A building’s ownership runs through deeds, mortgage priorities, remedies, and enforcement. A company’s assets are owned through corporate law, the share, and the doctrine of limited liability — none of which is a physical thing. A debt is an enforceable obligation, given effect by the law of contract and, where it is embodied in a negotiable instrument, by commercial law. A bank balance is a structured claim recognised by accounting convention, contract, and statute. A warehouse receipt, a bill of lading, a trust interest — each is an institutional construction, an artifact of law laid over some underlying fact, and each is unquestioningly treated as property.

Figure 3. All developed property is institutional. Intellectual property slots into the same architecture as land, shares, and debt; the “requires law” objection cannot isolate it.

The abolitionist’s slogan, set against this, proves far too much. He says patents and copyright are artificial because they require law to exist and to be enforced. But land title requires law to exist and be enforced. The company share requires law. The mortgage, the debt, the negotiable instrument, the trust — every one of them requires law, and none of them is found in nature. If “requires legal recognition” entails “artificial privilege rather than genuine property,” then the entailment does not stop at intellectual property; it sweeps through the entire architecture of ownership, dissolving land title and corporate shares and secured debt along with copyright and patent. The title deed is not a natural object. The cadastral map is not a fruit of the soil. The share certificate is not a cow. The mortgage is not a rock. Yet each belongs to the architecture of property, and each is constituted by exactly the kind of legal-institutional act the abolitionist treats as disqualifying when, and only when, the asset in question is an intellectual creation.

The Peruvian economist Hernando de Soto built an entire account of development around this insight, though from the other direction: his thesis is that the assets of the poor in the developing world remain “dead capital” — incapable of being mortgaged, traded, or leveraged — precisely because they lack the formal institutional layer of title and registration that converts a physical holding into legally cognisable property. The physical house exists; without the institutional layer it cannot function as property. De Soto’s point and Mises’s word “delimitation” meet here. Property is not the brute fact of a thing; it is the institutional delimitation of rights over a thing. And once that is seen, the abolitionist’s demand that intellectual property justify itself by being natural, tangible, and pre-political is revealed as a demand that no property can meet — not land, not shares, not debt, not money. The demand is not a principle that intellectual property happens to fail. It is a solvent that destroys property as such, applied selectively to the one form the abolitionist has decided to dislike.

V. What Mises leaves us, and what comes next

Gather the thread. Mises did not hand the abolitionist a weapon; he handed him a problem, and the problem cuts the other way. He saw, more precisely than the abolitionist sees, that the services of a disclosed creation are inexhaustible and in that respect free. But he saw the other half too — that the creation must be produced at a cost borne by one person before any return — and he named the result the extreme case of external economies. From this he drew not a verdict but a question: how a legal order should delimit property rights around such a good. He took no side on the answer, as the abolitionist’s own footnote admits. To present Mises as an abolitionist is to invert him.

And the word he chose, delimitation, opens onto the deeper point. Delimitation is an institutional act, and all developed property rests on institutional acts of exactly this kind. The field is not property until it is surveyed, titled, and defended; the company is not owned until corporate law constructs the share; the debt is not collectible until contract and commercial law give it force. Intellectual property is one more delimitation of rights over one more kind of valuable thing — a created form rather than a parcel of land or a bundle of corporate claims — and it is constituted by law in precisely the way they are. The abolitionist’s charge that it is “artificial because it requires law” is therefore not an argument against intellectual property in particular. It is an argument against property in general, and an argument that proves too much refutes itself.

What remains is the political half of the abolitionist’s case — the claim that because these rights depend on the state, specifically, they are illegitimate, and that a stateless order could secure complex property and exchange without them. That claim is Hayek’s territory, and it is the subject of the third and final essay, “Hayek’s Courthouse,” which will show that the rule-bound institutional order Hayek spent his life describing is not the enemy of the market but its precondition, that small government is not no government, and that the anti-intellectual-property argument, taken to its root, is not a defence of the market at all but a path to socialising created value — and, behind it, all value — through the back door. Mises gave us the economics of the created good. Hayek gives us the institutions without which no good, created or natural, is property at all.

References

Mises, primary

• Mises, Ludwig von. Human Action: A Treatise on Economics. 3rd rev. ed. Chicago: Henry Regnery, 1966. Page-level anchors: p. 128 (the coffee recipe; a known formula “renders unlimited services,” “its productive power is inexhaustible; it is therefore not an economic good”); p. 364 (”such recipes are, as a rule, free goods as their ability to produce definite effects is unlimited”); p. 661 (the external-economy structure; “the extreme case of external economies”; “this is a problem of the delimitation of property rights”; with abolition, authors and inventors “would for the most part be producers of external economies”).

The anti-IP position under examination

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (2001): 1–53. See n. 38: “Mises expressed no opinion on the issue, merely drawing the economic implications from the presence or absence of such laws” — the concession that Mises did not hold the abolitionist view.

The economics of appropriability (the structure Mises anticipated)

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity, 609–626. Princeton: Princeton University Press (NBER), 1962. (Inappropriability and the disclosure paradox.)

• Landes, William M., and Richard A. Posner. “An Economic Analysis of Copyright Law.” Journal of Legal Studies 18, no. 2 (1989): 325–363. (Fixed cost of expression vs near-zero cost of copying.)

The institutional character of property

• De Soto, Hernando. The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else. New York: Basic Books, 2000. (Untitled assets as “dead capital”; property as a legal-institutional construction over a physical substrate.)

• Searle, John R. The Construction of Social Reality. New York: Free Press, 1995. (Institutional facts; status functions; how collective recognition constitutes ownership.)

• Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). (Originality, not effort, as the basis of copyright — relevant to delimitation, against a labour rationale.)

Note on method and scope. Every quotation attributed to Mises reflects the verified wording of Human Action (3rd rev. ed., 1966) at the cited pages, checked against the text, not drawn from an abstract or secondary paraphrase. The claim sometimes attributed to Mises that intellectual creations stand outside property because they are “immaterial, intangible, and impalpable” is not Mises’s wording and is not used here; his actual distinction is the inexhaustibility of the services a disclosed formula renders, and it is represented as such. This is the second of three essays. It argues that Mises diagnosed an external-economy problem and treated the legal response as a matter of delimitation rather than illegitimacy, and that the institutional character of all property defeats the “artificial because legally constituted” objection. It does not claim Mises endorsed intellectual property — he expressly did not — and it does not defend any particular statute, term, or scope; the legitimacy of Austrian criticism of overbroad or abusive regimes is a question of delimitation, conceded throughout. The political half of the abolitionist case — the appeal to statelessness — is addressed in the third essay.

Keywords: intellectual property, scarcity, rivalry, appropriability, comparative institutional analysis, Kinsella, Plant, Machlup, Arrow, anticommons, cumulative innovation, open source, copyleft, Benkler, patent reform, human flourishing.

Abstract

This essay argues that the serious question about intellectual property is not whether one is “for” it or “against” it — a tribal framing that substitutes a badge for an argument — but which institutional order, judged against an explicit and operational standard of human flourishing, best supports invention, disclosure, the recovery of investment, competition, access, and the diffusion of knowledge; and it argues, after engaging the strongest versions of the case on both sides, that the answer is neither blanket defence of the existing system nor abolition, but reform. It goes deeper than the surface debate by engaging the foundation the serious anti-IP case actually rests on: not the slogan that “you cannot own ideas,” nor even the claim that intellectual property is a non-consensual negative servitude on another’s physical property, but the prior premise from which both descend — the scarcity-and-rivalry argument running from Hoppe through Kinsella and grounded in Arnold Plant’s 1934 demonstration that property rights in patents and copyright “do not arise out of the scarcity of the objects which become appropriated” but “make possible the creation of a scarcity… which could not otherwise be maintained.” The essay shows that this argument proves too much: the same rivalry test that would deny protection to ideas would equally dissolve trade secrets, goodwill, reputation, choses in action, spectrum, and other recognised property interests that govern non-rivalrous or rival-only-in-use goods; that the relevant scarcity is not in the idea, which is non-rival in use, but in the appropriability of the returns to its costly production, which is the genuine public-good problem Kenneth Arrow identified in 1962; and that “binding non-consenting others” cannot be the disqualifier, because every property right, beginning with first acquisition of land, binds the whole world without its consent. Having met the rights-based foundation on its own ground rather than deflecting it, the essay turns to the economics and refuses to overclaim in the other direction: the effect of intellectual property on innovation is, as Fritz Machlup concluded for the United States Senate in 1958, genuinely uncertain — “if we did not have a patent system, it would be irresponsible… to recommend instituting one; but since we have had a patent system for a long time, it would be irresponsible… to recommend abolishing it” — and the modern literature confirms a two-sided picture, with patents plausibly aiding recovery in high-fixed-cost, easily-copied, disclosure-dependent fields such as pharmaceuticals (the contested Tufts estimate notwithstanding) while the anticommons literature of Heller and Eisenberg, the cumulative-innovation analyses of Scotchmer and of Bessen and Maskin, and the historical evidence of Moser show that strong rights can also retard the sequential innovation that is the real engine of progress, and even the anticommons thesis is itself empirically disputed. The honest conclusion is that intellectual property imposes two opposing error costs — under-protection that leaves costly disclosure and production unrewarded, and over-protection that manufactures scarcity and blocks cumulative innovation — and that the optimum is interior and sector-specific, which is precisely what reform calibrates and what abolition and maximalism, the two corners, both refuse. The essay sets out a concrete, category-by-category reform agenda, much of it already enacted by the courts (eBay, KSR, Alice, Octane Fitness, Google v. Oracle), and engages the apparent counterexamples — open source, copyleft, Creative Commons, Wikipedia — to show that they do not demonstrate the viability of abolition but confirm the thesis, because copyleft and Creative Commons are licences built upon copyright that use the right to compel openness and would collapse without it; they are the most successful reform of intellectual property ever deployed, not its absence. Decisively, abolishing intellectual property does not abolish control; it migrates control out of a time-limited, disclosed, expiring public right and into perpetual, undisclosed private enclosure — trade secrecy, contracts of adhesion, and the technical lock — that systematically favours large capital over the individual creator, as the right-to-repair conflict shows in the open. The philosophical literature confirms that intellectual-property rights require justification rather than being self-evident, and the justifications — Lockean, personality-based, and utilitarian — each have force and limits; but none of the serious objections refutes the category, and measured against the contractual and secretive enclosure that abolition actually yields, a properly limited and reformed system of creator rights has the stronger comparative-institutional case. The standard throughout is institutional performance against human flourishing, operationalised as a definite checklist, and the conclusion the foundations, the evidence, and the comparison jointly support is reform, not abolition.

I. Beyond the badge: what the serious argument is actually about

The usual way of arguing about intellectual property guarantees that nothing is learned, because it is not an argument but an act of membership. One declares oneself “pro-IP” or “anti-IP,” and the declaration sorts the speaker into a tribe, summons the tribe’s slogans, and licenses the dismissal of the other side before a single premise has been examined. So I will state plainly what this essay is. It is not a defence of the existing system, which is in many respects indefensible and which I criticise concretely below. It is not a brief for any school. It is an attempt to answer the one question that can be argued rather than asserted — which institutional order best lets human beings invent, disclose, create, invest, recover their costs, compete, learn, and pass an improving civilisation forward — and to answer it by engaging the strongest versions of the case on both sides, including the deep one that most popular debate never reaches.

That deep version matters because the surface arguments are weak and the serious argument is strong, and a treatment that only refutes the weak ones is worthless. “You cannot own ideas” is too vague to be useful and, stated baldly, is a straw man: no defensible account of intellectual property claims ownership of abstract thoughts. The “negative servitude” argument — that a patent is a non-consensual encumbrance on others’ physical property — is stronger, but as I showed in an earlier version of this argument, it presupposes the very thing in dispute, namely that the owner’s liberty already included the use the right restricts. But neither of these is the foundation. The foundation, in the most rigorous libertarian and economic statements of the anti-IP case, is a claim about scarcity and rivalry — and that is the claim this essay engages first, because everything else descends from it.

The standard against which I will judge the competing institutions is human flourishing, and because that phrase is often invoked and never defined, let me operationalise it at the outset. By human flourishing I mean not a mood but a definite checklist of what a good institutional order must deliver: it must support invention, encourage disclosure rather than secrecy, allow the recovery of investment, preserve competition, secure access, promote the diffusion of knowledge, protect small creators against fast copying, sustain institutional trust, and enable long-run cumulative progress. An intellectual-property regime, or its abolition, is to be judged by how it performs against those criteria, item by item, sector by sector. That is the test, and it is neither doctrinal purity nor bare efficiency; it is the social function that property rules exist to serve.

II. The scarcity argument — and why it proves too much

The most rigorous case against intellectual property does not begin with ideas or servitudes. It begins with a theory of why property exists at all. On this theory — developed by Hans-Hermann Hoppe and made the centrepiece of Stephan Kinsella’s Against Intellectual Property — property rights exist to resolve conflict over scarce, in the sense of rivalrous, goods. Two people cannot both eat the same apple or stand on the same square foot of land; their uses are mutually exclusive; property assigns the right to decide and so prevents violence. From this premise the conclusion is meant to be inexorable. Ideas, the argument runs, are not scarce in this sense: they are non-rivalrous, because my use of a technique does not exclude yours — we can both use it — just as, in Jefferson’s image, one taper may be lit from another without darkening the first. If ideas are non-rivalrous, there is no possibility of conflict over them, and therefore no need for property in them. And here the argument reaches its sharpest point, drawing on Arnold Plant’s classic 1934 papers: ordinary property rights respond to a scarcity that already exists, but property rights in patents and copyright, Plant wrote, “do not arise out of the scarcity of the objects which become appropriated… they make possible the creation of a scarcity of the products appropriated which could not otherwise be maintained.” Intellectual property, on this account, does not track a pre-existing scarcity; it manufactures one, and in doing so hands the holder a measure of control over everyone else’s tangible property — their presses, their computers, their factories. The negative-servitude conclusion is the output of this argument; the rivalry premise is its engine.

This is a serious argument and it deserves a serious answer on its own ground, not a deflection. Here is the answer, in three parts.

Figure 1. The anti-IP case’s true foundation is the rivalry premise, not the negative-servitude conclusion. Engaged on its own ground, the premise proves too much — it would dissolve much recognised property — and points to the real question: the appropriability of returns to production.

First, the rivalry test proves too much, because it would dissolve a great deal of property that no one proposes to abolish. Confidential information and trade secrets are protected by law, yet information is the paradigm non-rivalrous good — my knowing your secret does not stop you knowing it. Goodwill and business reputation are protected through the law of passing off and trade marks, yet they are not rivalrous in the apple-and-land sense. A chose in action — a debt, a share, a contractual right — is a protected interest in value, not a rivalrous physical object. Radio spectrum and airspace are allocated as property though they are not consumed by use in the way an apple is. If the rivalry premise were a sound test of what may be property, it would strike all of these; since no one accepts that conclusion, the premise cannot be doing the work the argument needs. One cannot wield a principle against patents that, applied consistently, would also abolish trade secrets — which is awkward in the extreme for an argument that elsewhere relies on secrecy and contract as the substitutes for intellectual property.

Second, and more deeply, the rivalry framing misidentifies where the scarcity is. It is true that an idea, once it exists, is non-rival in use: any number of people can apply it at once without depleting it. But the economically relevant scarcity was never in the finished idea. It is in the production of the idea and the appropriability of the returns to that production. The investment that produces a new drug, a new design, a new body of expression is enormous, rivalrous, and sunk; the resources poured into it cannot simultaneously be poured into anything else; and the returns that would justify the investment may be wholly unappropriable if the result, once disclosed, can be copied for nothing. This is precisely the problem Kenneth Arrow identified in 1962: information goods have a public-good character that creates a genuine allocation problem — not because anyone is harmed by another’s use of the idea, but because the production of socially valuable information may not occur at all if its returns cannot be captured. The anti-IP argument scores a real point against a bad justification (the idea that one “owns” an idea the way one owns an apple) and then treats it as though it had answered the real question, which is institutional and economic: how, if at all, should a society secure the appropriability of returns to costly production of non-rival goods? That question does not answer itself, and “ideas are non-rivalrous” does not answer it either.

Third, the “without consent” strand of the argument — that intellectual property illegitimately binds non-consenting third parties — cannot be the disqualifier, because every property right does exactly that. When I homestead a plot of land, I impose on the entire rest of humanity a duty not to enter it, and not one of them consented. First acquisition is, in this precise sense, a non-consensual imposition on everyone else; it is the way property works. The owner of a trade secret binds non-consenting others not to exploit a wrongfully obtained formula; the holder of a contract binds assignees; the beneficiary of an easement binds successors who never agreed. “It binds people who did not consent” is true of property as such and therefore cannot single out intellectual property as illegitimate. What can be disputed is which impositions are justified — and that returns us, once again, to the real question of whether produced, disclosed contribution can ground a protected interest, which the rivalry premise was supposed to settle and does not.

None of this establishes any particular intellectual-property right. It establishes that the deepest and most rigorous version of the anti-IP case does not refute the category either, because its central premise proves too much, mislocates the scarcity, and rests on a feature common to all property. What survives the argument is the genuine question the rivalry premise was meant to foreclose: the appropriability of returns to costly production of non-rival goods is a real problem, and how a society should address it is institutional and empirical. To that we now turn.

III. The rights question, answered positively

Before the economics, the rights question deserves a positive answer rather than a second deflection, because a fair criticism of the comparative-institutional approach is that it can seem to duck “what is just?” while answering “what works?” The honest position answers both, and uses each to discipline the other.

The positive account is this. The law already grants protected interests in produced value across many domains, on a principle that has nothing to do with creating matter from nothing. The farmer who does not create the soil acquires a protected interest in the crop he produces, through accession, specification, and improvement; the person who generates confidential information acquires a protected interest through the law of confidence; the builder, the manufacturer, and the author of a fixed work all acquire protected interests in what they have made, though every atom and every word pre-existed them. The relevant act was never metaphysical creation; it is the ordering of existing materials, symbols, and information into a valuable arrangement that did not exist before in that form. If the law can and does ground protected interests in produced value generally, then a specific, fixed, disclosed intellectual contribution is not disqualified from grounding one merely because it is intellectual. Whether it should, and within what limits, is the further question — but the categorical denial is a bare premise, false to the structure of the law it claims to describe.

The philosophical literature has worked this ground thoroughly, and it is worth being precise about what it does and does not deliver, because the honest answer engages it rather than gesturing at it. William Fisher’s canonical survey identifies four families of justification: the utilitarian (rights as incentives to produce socially valuable works), the Lockean labour-desert (one acquires a property interest by mixing one’s labour with the unowned), the personality theory descending from Hegel (expressive works as extensions of the author’s self), and the social-planning theory (rights shaped to foster a just and attractive culture). Each has genuine force and genuine limits. The Lockean argument, as Justin Hughes and Lawrence Becker developed and Wendy Gordon refined it, grounds a real intuition — that one has a claim to what one has produced — but, as critics from Robert Nozick onward noted, the “mixing labour” metaphor strains when the thing produced is non-rival and the “commons” of ideas is not depleted by its appropriation. The personality theory explains our strong intuitions about authorship and attribution but fits expressive works far better than functional inventions. The utilitarian theory is the workhorse of actual law but, as Edwin Hettinger sharply observed, contains a paradox: copyright and patent “establish a right to restrict the current availability and use of intellectual products for the purpose of increasing the production and thus future availability and use of new intellectual products” — restricting access now to expand it later — which is coherent only if the future gain exceeds the present restriction, and that is an empirical matter that no a priori argument settles. Robert Merges, in the most sustained recent attempt to ground intellectual property philosophically, concedes the essential point that frames this whole essay: these rights “require clear philosophical justification precisely because they are exceptions to the default norm of open access and the public domain.” That concession is correct, and it cuts against the maximalist as much as the abolitionist. Intellectual-property rights are not self-evident; they bear a burden of justification; and the burden is discharged, if at all, partly by showing that they produce better consequences than the alternative — which is the consequences question, to which the rights answer hands off.

The relationship between the two questions is therefore not a tension to be hidden but the structure of the answer. The rights account sets the outer bound: it establishes that a creator right is not inherently a theft from others, that produced and disclosed contribution can ground a protected interest, and so that the category is not ruled out at the threshold. Within that bound, the consequences account does the calibrating work — fixing scope and term by what actually serves flourishing. A natural-rights theorist who sets a patent’s term at “forever” because creation grounds ownership has proved too much; a consequentialist who would seize a disclosed invention the instant it proves useful has ignored the wrong done to its maker. The content of property justice — the precise delineation of what is owned and for how long — is worked out partly by reference to what property is for. That is not consequentialism dissolving rights; it is the recognition, old and respectable, that rights and their function are not finally separable.

IV. The economics, without overclaiming: two error costs

If the rights question cannot rule the category out, and cannot by itself fix its scope, then the scope is an economic question — and here intellectual honesty requires resisting the temptation to overclaim in the pro-IP direction, exactly as the asymmetry fallacy must be resisted in the anti-IP direction. The most authoritative review of the evidence reached a verdict of genuine uncertainty, and it has not been overturned.

In an exhaustive 1958 study for the United States Senate, Fritz Machlup concluded: “No economist, on the basis of present knowledge, could possibly state with certainty that the patent system, as it now operates, confers a net benefit or a net loss upon society… If we did not have a patent system, it would be irresponsible, on the basis of our present knowledge of its economic consequences, to recommend instituting one. But since we have had a patent system for a long time, it would be irresponsible, on the basis of our present knowledge, to recommend abolishing it.” That last sentence is, almost word for word, the thesis of this essay — and it comes not from an advocate but from the most careful empirical reviewer the question has had. Arnold Plant, a generation earlier, had gone further, concluding that “the science of economics as it stands to-day furnishes no basis of justification for this enormous experiment.” The empirical case for strong patents was weak in 1934, weak in 1958, and — despite enormous subsequent work — remains genuinely contested today. Anyone who tells you the evidence clearly favours either strong intellectual property or its abolition is overclaiming.

The reason the evidence is two-sided is that intellectual property imposes two opposing error costs, and the literature populates both.

Figure 2. Intellectual property imposes two opposing error costs, and the serious literature sits on both sides. The optimum is interior and sector-specific — which is what reform calibrates and what both abolition and maximalism refuse.

On the side of under-protection: where the fixed cost of creation is high, the result is easily copied, and the law requires disclosure in order to sell, exclusivity plausibly does real work. Arrow’s appropriability problem is genuine; remove all protection and some socially valuable production that depends on capturable returns will not be financed, and disclosure will be replaced by secrecy. Pharmaceuticals are the paradigm: development is marked by enormous risk-adjusted fixed costs, and although the most-cited estimate — the Tufts Center’s figure of roughly $2.6 billion per approved drug, which explicitly includes the cost of failures and of capital — is contested (its critics, including Knowledge Ecology International, argue it is inflated and partly built on assumptions rather than project-level data), even a figure a fraction of that size establishes that exclusivity-period pricing is substantially recovery rather than pure rent. The figure cuts both ways: it refutes “patent pricing is pure extraction,” and its contestation refutes “patent pricing is pure recovery.” The recovery-versus-rent split is empirical and sector-specific, and pretending it is settled in either direction is dishonest.

On the side of over-protection: where innovation is cumulative — where each advance builds on the last — strong rights can reduce total innovation by blocking the follow-on work that is the real engine of progress. Suzanne Scotchmer’s analysis of “standing on the shoulders of giants” and the formal results of James Bessen and Eric Maskin on sequential innovation both show that in such fields patents can be net-negative. Michael Heller and Rebecca Eisenberg’s 1998 work identified the anticommons: a proliferation of overlapping upstream patents can deter downstream innovation by forcing would-be innovators to assemble too many licences, so that more property yields less useful output. Plant’s manufactured-scarcity point bites here, as do the everyday pathologies of thickets, trolls, and evergreening. And here honesty requires a turn that cuts against my own side: the anticommons thesis is itself empirically contested — subsequent surveys found the predicted anticommons “has not materialised significantly,” and serious scholars such as Epstein and Kuhlik argued Heller and Eisenberg overstated it. I am not entitled to deploy the anticommons as a decisive stick while ignoring that its magnitude is disputed, any more than the abolitionist is entitled to an idealised commons. Mark Lemley’s demonstration that simultaneous and independent invention is the historical norm rather than the exception — the “myth of the sole inventor” — cuts the same way, undermining the heroic-inventor story that the strongest pro-patent rhetoric relies on, and supporting a genuine independent-invention defence as a reform.

The conclusion that the two error costs force is not a verdict for either corner. It is that the right scope and term minimise the sum of the two costs, and that the sum differs by sector: it is high where fixed costs are large, copying is easy, and disclosure is forced, and low where innovation is cumulative or where reputation and complementary revenue already finance production. Calibrating that interior optimum, sector by sector, is exactly what reform is — and exactly what abolition and maximalism, the two corners, both refuse.

V. The pathologies are real — and so reform is concrete

Nothing in the foundational or economic argument defends the existing system, and the most common failure of pro-IP writing is to slide from “creator rights can be legitimate” to “the current statutes are fine.” They are not, and honesty requires naming the pathologies concretely, at the scale and persistence they actually have.

In patents: the non-practising entity or troll that produces nothing and acquires patents only to extract settlements under threat of injunction — a problem so visible that Justice Kennedy named it from the bench in eBay. The patent thicket and anticommons, in which overlapping rights make it impossible to build without licensing a dozen owners. Evergreening, by which a pharmaceutical monopoly is extended through trivial reformulations long after the original invention should have entered the public domain. The granting of obvious or trivial patents that should never have issued, and of abstract software and business-method patents that claim an idea by reciting “do it on a computer” — a class that James Bessen and Michael Meurer, in Patent Failure, showed function less like property with clear boundaries than like a minefield of unclear claims that defeats the notice function property is supposed to serve. In copyright: term extensions driven by corporate lobbying that have stretched protection far beyond anything an incentive rationale could justify, locking up twentieth-century culture for generations; and the use of anti-circumvention law to block lawful repair and interoperability. These are not reasons to abolish intellectual property. They are reasons to reform it — and the move from “this institution performs badly” to “this institution should not exist” requires the extra step, which abolition rarely takes, of showing that what replaces it performs better. A corrupt court is an argument for a better court, not for no courts.

Reform is not a gesture; it is a concrete agenda, and it differs by category. For patents: raise the obviousness bar, as KSR v. Teleflex (2007) did; discipline abstract and functional claims, as Alice v. CLS Bank (2014) did; create a genuine independent-invention defence, which Lemley’s evidence on simultaneous invention supports; shift fees against abusive suits, as Octane Fitness (2014) eased; deny automatic injunctions to non-practising entities, as eBay v. MercExchange (2006) did; curb evergreening; and shorten effective terms in fast-moving fields where the cumulative-innovation cost is highest. For copyright: roll back the lobby-driven term extensions; restore registration and formalities; keep fair use robust; protect reverse-engineering and interoperability as fair use, as Google v. Oracle (2021) did when it held that reimplementing an API’s declaring code was fair use; and narrow anti-circumvention so it cannot block lawful repair. For trade secret: keep it bounded to genuine confidentiality and breach, protect employee mobility by limiting non-competes, and prevent it from becoming a perpetual end-run around the patent disclosure bargain. For trade mark: keep it to preventing consumer confusion and protecting goodwill, and resist its expansion into the control of language and expression. Much of this agenda is not utopian; it is already underway through the courts, which is itself the strongest evidence that reform — not abolition — is the realistic path. A system capable of eBay, KSR, Alice, Octane, and Oracle is a system that can be reformed, and the task is to continue and complete that work, not to burn the institution down.

VI. The counterexamples, properly understood: open source confirms the thesis

The strongest objection to the comparative case is empirical and concrete, and it deserves to be met at depth rather than waved away: open-source software, copyleft, Creative Commons, Wikipedia, and the open-hardware communities plainly sustain enormous creative output on models of disclosure, reputation, and complementary revenue, with little reliance on the exclusionary use of intellectual property. Yochai Benkler gave this its canonical theorisation as commons-based peer production — a genuine third mode of organising production alongside the firm and the market, in which large numbers of diversely motivated contributors aggregate small contributions outside both managerial hierarchy and price signals. Josh Lerner and Jean Tirole explained the individual economics: contributors are moved by intrinsic motivation and by signalling — the reputational and career returns to visible, high-quality work — and firms profit through complementary goods and services rather than through exclusivity. If all of this works, and it manifestly does, does it not show that abolition is viable and that the claim that abolition “favours capital” is overblown?

Figure 3. The open-source counterexample, properly understood, confirms the thesis. Peer production is real and powerful — but copyleft and Creative Commons run on copyright, and would collapse without it.

It would — except that these models are not the absence of intellectual property. They are built upon it, and the point is not rhetorical but mechanical. The General Public Licence that makes copyleft work is a copyright licence: it uses the author’s copyright to impose a binding, running condition — that derivative works remain open and carry the same licence forward. Benkler’s own definition of peer production names “copyleft or permissive licensing” as one of its constitutive criteria; the openness is engineered through the right, not in its absence. Remove the underlying copyright and the share-alike obligation becomes unenforceable: there is no longer any legal lever to stop a well-capitalised firm from taking the open code, improving it behind a proprietary wall, and releasing nothing back. Creative Commons is the same device — a structured deployment of copyright to compel attribution and sharing, not an alternative to copyright — and Wikipedia’s licensing rests on the same foundation. These communities are therefore not evidence that creator rights are unnecessary. They are evidence that creator rights, reformed and redirected, can be made to serve openness — which is an argument for reform and against abolition, not the reverse. The honest reading is that the counterexamples are among the best illustrations of the thesis: control built on the public right of copyright produces openness; remove the right, and openness loses its enforcement mechanism while secrecy and scale lose nothing. Copyleft is, quite literally, the most successful reform of intellectual property ever deployed — a redirection of the right toward openness — and it would collapse without the right it deploys.

This also lets me state the “favours capital” claim with the precision the evidence demands, neither overclaiming nor retreating. It is not universal. It is strongest precisely where fixed costs are high, copying is easy, and disclosure is unavoidable, and weakest where complementary revenue, reputation, and low coordination costs already finance production — which is why peer production thrives in software, where the marginal contributor’s cost is low and the complementary-revenue and signalling returns are high, and is far rarer where bringing a product to market requires hundreds of millions in sunk, unrecoverable investment. The claim is a statement about a tendency under specified conditions, not a law of nature; and the open-source case, properly understood, does not refute it but maps its boundary.

VII. The decisive point: control migrates, it does not vanish

We come, finally, to the move that the corrected comparison exists to set up, and which the foundational, economic, and counterexample analyses have now equipped. Abolishing intellectual property does not abolish control over created value. It relocates control — out of a public, time-limited, disclosed, expiring legal right, and into private mechanisms that are perpetual, undisclosed, and unaccountable. The control does not disappear; it changes form, and the new form is worse on the very dimensions of the flourishing checklist that the abolitionist invokes.

The private mechanisms are already in use and would expand to fill the space abolition created: trade secrecy, non-disclosure agreements, employee confidentiality and non-compete restrictions, compartmentalised research, encrypted firmware, server-side software that never leaves the vendor’s control, diagnostic and repair locks, restricted APIs, closed standards, contractual access terms, reverse-engineering prohibitions, licensing in place of sale, binding arbitration, and class-action waivers. Each controls created value without disclosure, without a time limit, and without the public-domain endpoint that a patent or copyright eventually reaches. Trade secrecy is the natural substitute for the patent: where the patent discloses in exchange for bounded protection, the trade secret protects precisely by not disclosing, potentially forever. This is why Moser’s historical finding cuts the way it does — remove patents and innovation migrates to where secrecy works, which is less disclosure, not more. The abolitionist who imagines that removing patents yields open knowledge has the direction of the effect backwards.

And the migration is not neutral between large and small, which is the distributive heart of the matter and the point at which the earlier sections converge. Large firms are far better placed to exploit the private substitutes than small creators are. A large firm can maintain secrecy at scale behind compartmentalised research and a wall of NDAs; a lone inventor usually cannot, because he must disclose his invention to attract the investment that would let him build it — the very appropriability problem of Section IV, now with a distributive edge. A large firm can draft, impose, and enforce contracts of adhesion across millions of transactions; a small creator has weak bargaining power and cannot easily litigate a breach. A large firm can build the technical locks — encrypted firmware, server-side execution, authentication gates — that enclose a product against repair and copying; a small creator cannot. Stripped of a recoverable period of exclusivity, the small creator is exposed to exactly the fast copying that exclusivity existed to prevent, very often by the large firm with the scale to bring his idea to market faster and cheaper. Abolition may transfer power from individual creators to capital-rich corporations. It is not the friend of the little inventor its rhetoric supposes.

The right-to-repair conflict shows the stack of private control operating in the open, and shows that the “control” at issue is rarely a single patent that abolition would dissolve. Modern agricultural equipment is locked against owner and independent repair not by one right but by a layered combination: copyright in the embedded software, backed by the anti-circumvention provisions of Section 1201 of the Digital Millennium Copyright Act, which make it unlawful to bypass the digital lock even to repair a machine one owns; trade secrecy in the diagnostic information; the technical lock itself, the engine control unit that refuses to function after an unauthorised repair; and, underneath it all, the manufacturer’s contractual claim — John Deere has pressed this since 2015 — that because the tractor runs on code, the farmer does not really own it but holds only “an implied license for the life of the vehicle to operate the vehicle.” Strip out the patent layer, or even the copyright layer, and the rest of the stack stands: the contract, the trade secret, and the lock remain. This is why the contest has had to be fought on every front at once — a Federal Trade Commission and state-attorneys-general antitrust suit that survived a motion to dismiss, renewable repair exemptions to Section 1201 from the Copyright Office, state right-to-repair statutes beginning with Colorado’s, and a proposed federal FARM Act — and it is telling that even these reforms generally preserve the manufacturer’s trade secrets. The lesson for the abolition debate is exact: control over created value lives in a stack of contract, secrecy, technical measures, and law, and removing any single layer leaves the others standing. Abolishing “intellectual property” in the narrow sense would not free the farmer; it would leave him facing the contract, the secret, and the lock.

VIII. Conclusion: reform, honestly

The honest conclusion is the one the slogans on both sides are built to prevent, and it is a conclusion for reform — reached, I have tried to ensure, without idealising the existing system and without ducking the hardest foundational objection.

The category of intellectual property survives the serious arguments against it. It survives the intangibility objection, because debts, shares, companies, trusts, and confidential information are all intangible institutional interests in value and are property nonetheless. It survives the negative-servitude argument, which presupposes the very thing in dispute. And it survives the deepest objection, the scarcity-and-rivalry argument, which proves too much — it would dissolve trade secrets and goodwill along with patents — misidentifies the scarcity, which lies in the appropriability of returns to costly production rather than in the non-rival idea, and rests on a non-consensual binding of third parties that is a feature of all property, beginning with the first claim to land. What survives every one of these is a genuine and difficult question: which produced intellectual contributions should generate protected interests, of what scope, for how long. That question is answered not by slogans but by engaging the rights account and the consequences account together, against an operational standard of human flourishing.

When that engagement is honest, several things hold at once, and they fit together rather than competing. The rights account establishes that creator rights can be legitimate but cannot fix their scope. The economic evidence is genuinely two-sided — Machlup’s verdict of “irresponsible to institute, irresponsible to abolish” remains the honest summary — because intellectual property imposes two opposing error costs whose sum is minimised at an interior, sector-specific optimum. The existing system has real and serious pathologies that demand concrete reform, much of it already begun by the courts. The open-source counterexamples confirm rather than refute the comparative thesis, because copyleft and Creative Commons are reforms of copyright that would collapse without it. And, decisively, abolishing intellectual property does not abolish control but migrates it into perpetual, undisclosed, private enclosure that favours large capital over the individual creator — as the right-to-repair stack shows in the open.

So the burden the abolitionist bears is the one he most often refuses: to specify the replacement system and show that it produces a better society than a reformed system would. Measured against the contractual and secretive enclosure that abolition actually yields, a properly limited system of creator rights — bounded, disclosed, expiring, distinct across its categories, calibrated to the two error costs sector by sector, and pruned of its real abuses — has the stronger comparative-institutional case. The test, from first to last, is not which tribe is purer. It is which order better lets human beings invent, disclose, invest, compete, learn, and flourish. The answer that survives the foundations, the evidence, and the comparison is not the status quo, and it is not abolition. It is reform — and the case for it is stronger, not weaker, for engaging the deepest objection on its own ground and admitting exactly where the evidence runs against it.

References

Property theory and the institutional character of ownership

• Honoré, A. M. “Ownership.” In Oxford Essays in Jurisprudence, ed. A. G. Guest. Oxford: Oxford University Press, 1961.

• Penner, J. E. The Idea of Property in Law. Oxford: Oxford University Press, 1997.

• Merrill, Thomas W., and Henry E. Smith. “What Happened to Property in Law and Economics?” Yale Law Journal 111 (2001): 357–398.

• Demsetz, Harold. “Toward a Theory of Property Rights.” American Economic Review 57, no. 2 (1967): 347–359. (Property rights emerge to internalise externalities.)

Philosophy of intellectual property (both sides)

• Fisher, William. “Theories of Intellectual Property.” In New Essays in the Legal and Political Theory of Property, ed. Stephen Munzer. Cambridge: Cambridge University Press, 2001. (The four-theory survey: utilitarian, Lockean-labour, personality, social-planning.)

• Hughes, Justin. “The Philosophy of Intellectual Property.” Georgetown Law Journal 77 (1988): 287–366.

• Becker, Lawrence C. “Deserving to Own Intellectual Property.” Chicago-Kent Law Review 68 (1993): 609–629.

• Gordon, Wendy J. “A Property Right in Self-Expression: Equality and Individualism in the Natural Law of Intellectual Property.” Yale Law Journal 102 (1993): 1533–1609.

• Hettinger, Edwin C. “Justifying Intellectual Property.” Philosophy and Public Affairs 18, no. 1 (1989): 31–52. (The paradox of restricting present use to expand future use; leans toward alternative reward systems.)

• Merges, Robert P. Justifying Intellectual Property. Cambridge, MA: Harvard University Press, 2011, esp. pp. 3–5. (IP rights as exceptions to the default of open access, requiring justification.)

• Palmer, Tom G. “Are Patents and Copyrights Morally Justified? The Philosophy of Property Rights and Ideal Objects.” Harvard Journal of Law and Public Policy 13, no. 3 (1990): 817–865. (The libertarian critique.)

The scarcity/rivalry foundation of the anti-IP case

• Plant, Arnold. “The Economic Theory Concerning Patents for Inventions.” Economica (New Series) 1, no. 1 (1934): 30–51. (”Property rights in patents and copyright make possible the creation of a scarcity… which could not otherwise be maintained.”)

• Plant, Arnold. “The Economic Aspects of Copyright in Books.” Economica (New Series) 1, no. 2 (1934): 167–195.

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (2001): 1–53. (The rivalry/scarcity argument, building on Hoppe; the negative-servitude conclusion.)

• Hoppe, Hans-Hermann. A Theory of Socialism and Capitalism. Boston: Kluwer, 1989. (Scarcity/rivalry as the precondition of property — the premise Kinsella builds on. Note: Hoppe is an anarcho-capitalist, a tradition distinct from the limited-government Austrianism of Mises and Hayek.)

The economics of innovation, disclosure, and the contested evidence

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity. Princeton: Princeton University Press (NBER), 1962. (The appropriability/public-good problem — the case for some protection.)

• Machlup, Fritz. An Economic Review of the Patent System. Study No. 15, Subcomm. on Patents, Trademarks, and Copyrights, U.S. Senate Committee on the Judiciary, 85th Cong., 2d Sess. (1958), esp. pp. 79–80. (The “irresponsible to institute / irresponsible to abolish” verdict; net effect genuinely uncertain.)

• Heller, Michael A., and Rebecca S. Eisenberg. “Can Patents Deter Innovation? The Anticommons in Biomedical Research.” Science 280, no. 5364 (1998): 698–701. (The anticommons — whose empirical magnitude is itself contested; see the “Whither the Research Anticommons?” review literature and Epstein & Kuhlik.)

• Scotchmer, Suzanne. “Standing on the Shoulders of Giants: Cumulative Research and the Patent Law.” Journal of Economic Perspectives 5, no. 1 (1991): 29–41; and Innovation and Incentives. Cambridge, MA: MIT Press, 2004.

• Bessen, James, and Eric Maskin. “Sequential Innovation, Patents, and Imitation.” RAND Journal of Economics 40, no. 4 (2009): 611–635.

• Bessen, James, and Michael J. Meurer. Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovators at Risk. Princeton: Princeton University Press, 2008. (Patents’ failure of the notice function, especially in software.)

• Moser, Petra. “How Do Patent Laws Influence Innovation? Evidence from Nineteenth-Century World’s Fairs.” American Economic Review 95, no. 4 (2005): 1214–1236.

• Lemley, Mark A. “The Myth of the Sole Inventor.” Michigan Law Review 110 (2012): 709–760. (Simultaneous and independent invention as the norm — supports an independent-invention defence.)

• Boldrin, Michele, and David K. Levine. Against Intellectual Monopoly. Cambridge: Cambridge University Press, 2008; and “The Case Against Patents.” Journal of Economic Perspectives 27, no. 1 (2013): 3–22.

• DiMasi, Joseph A., Henry G. Grabowski, and Ronald W. Hansen. “Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs.” Journal of Health Economics 47 (2016): 20–33. (The ~$2.6 billion figure — cited with its criticisms; includes failures and cost of capital.)

• Mazzucato, Mariana. The Entrepreneurial State. London: Anthem Press, 2013. (Public funding of basic research as a partial upstream alternative.)

Open source and commons-based peer production

• Benkler, Yochai. The Wealth of Networks: How Social Production Transforms Markets and Freedom. New Haven: Yale University Press, 2006; and “Commons-Based Strategies and the Problems of Patents.” Science 305 (2004): 1110–1111. (Peer production — defined to include “copyleft or permissive licensing,” i.e. built on copyright.)

• Lerner, Josh, and Jean Tirole. “Some Simple Economics of Open Source.” Journal of Industrial Economics 50, no. 2 (2002): 197–234. (Signalling and complementary-goods incentives.)

• von Hippel, Eric. Democratizing Innovation. Cambridge, MA: MIT Press, 2005.

• The GNU General Public License and the Creative Commons licence suite. (Copyleft and CC as deployments of copyright to compel openness — reform of IP’s terms, not its absence.)

Patent and copyright reform in the courts

• eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006); KSR International Co. v. Teleflex Inc., 550 U.S. 398 (2007); Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014); Octane Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545 (2014); Google LLC v. Oracle America, Inc., 593 U.S. 1 (2021).

The migration of control

• Digital Millennium Copyright Act, 17 U.S.C. § 1201 (1998).

• Federal Trade Commission and state attorneys general v. Deere & Co. (filed January 2025; motion to dismiss denied, 2025); In re Deere & Co. Repair Services Antitrust Litigation (N.D. Ill., MDL); Freedom for Agricultural Repair and Maintenance (FARM) Act (introduced 2025); Colorado Consumer Right to Repair Agricultural Equipment Act (2023, eff. 2024).

• Radin, Margaret Jane. Boilerplate: The Fine Print, Vending Machines, and the Rule of Law. Princeton: Princeton University Press, 2013.

The Austrian and classical-liberal background (distinguished, not conflated)

• Mises, Ludwig von. Human Action. 3rd rev. ed. Chicago: Henry Regnery, 1966. (Produced intellectual goods and the “delimitation of property rights”; limited government distinguished from anarchism.)

• Hayek, F. A. Law, Legislation and Liberty. 3 vols. Chicago: University of Chicago Press, 1973–1979.

Note on method and scope. This is a further expanded version, built to add depth at the points a reviewer identified as thin: the foundations and the literature. It now engages the scarcity-and-rivalry argument (Hoppe → Kinsella → Plant) that is the true engine of the serious anti-IP case, rather than only its negative-servitude conclusion, and answers it on its own ground (it proves too much; the scarcity is in appropriability, per Arrow; non-consensual binding is universal to property); it gives a positive rights-based account and engages the philosophical literature on both sides (Fisher’s four theories, Hughes, Becker, Gordon, Hettinger’s paradox, Merges, Palmer); it frames the economics as two opposing error costs and engages the literature on each, anchored by Machlup’s authoritative 1958 verdict and including the cumulative-innovation work of Scotchmer and of Bessen and Maskin, Bessen and Meurer on notice failure, Lemley on simultaneous invention, and Boldrin and Levine, while flagging that the anticommons thesis is itself contested; it sets out a concrete category-by-category reform agenda, much of it already enacted by the named courts; and it engages the open-source counterexamples at depth, showing via Benkler’s own definition that copyleft and Creative Commons are built on copyright and so confirm the migration thesis rather than refute it. Every quotation (Plant, Machlup, Hettinger, the John Deere “implied license” position) reflects the verified wording of the cited source, checked at passage or page level; no claim is made that any specific technology “depended on patents”; the Tufts/DiMasi pharmaceutical figure is presented with its criticisms and treated as cutting both ways; and the reform cases reflect their actual holdings. The representation of the anti-IP position is stated at full strength before it is answered, and Austrian limited-government economics (Mises, Hayek) is distinguished from anarcho-capitalism (Hoppe, Rothbard) rather than conflated with it. The essay defends reform — not the status quo and not abolition — and the standard of judgment throughout is institutional performance measured against human flourishing. The author-reputation point raised in earlier review is deliberately not addressed in the body: it is an ad hominem with respect to the argument, which stands or falls on its merits regardless of who advances it, and inserting reputational material into a general treatment of intellectual property would be a category error.

Keywords: intellectual property, comparative institutional analysis, property theory, contract, trade secrets, negative servitude, Stephan Kinsella, right to repair, DMCA, patent disclosure, drug development cost, human flourishing, Mises, Hayek, Rothbard, corporate enclosure.

Abstract

Debates over intellectual property are usually conducted as contests of allegiance — one declares oneself “pro-IP” or “anti-IP” and the slogan does the work that argument should — and this essay rejects that framing in favour of a comparative-institutional one: the proper question is not whether the existing intellectual-property system is good, which it manifestly is not in every particular, but what institutional order best supports invention, disclosure, authorship, investment, recovery, competition, access, and the diffusion of knowledge. The argument proceeds from a methodological correction and then makes its substantive case. The correction is that most anti-IP arguments commit an asymmetry fallacy: they measure the real intellectual-property system, with its trolls, its overlong terms, its litigation costs, and its bad patents, against an imagined frictionless world of free and open knowledge, when the only valid comparison is between a limited or reformed system of creator rights and the actual replacement system that abolition would leave in place — contract, secrecy, trade secrets, technical locks, access controls, server-side software, licensing-not-sale, private arbitration, and corporate scale. Against that corrected baseline the essay argues four things. First, property is already institutional and largely non-corporeal: debts, shares, bank balances, choses in action, security interests, trusts, goodwill, confidential information, and securities entitlements are all legally constructed interests in value rather than tangible objects, so “intellectual property is intangible and legally constructed” cannot by itself be an objection without taking down much of commercial law with it. Second, the strongest anti-IP argument — that an intellectual-property right is a non-consensual negative servitude imposed on another’s material property — is not decisive but question-begging, because it presupposes that the owner’s liberty already included the use the right restricts, which is precisely the point in dispute; every enforceable right restricts some uses of material property, and the real question is whether the restricted use was ever part of the owner’s legitimate liberty in the first place. Third, the live dispute is therefore not the vague claim that “you cannot own ideas” but the precise one of whether a specific, produced, objective intellectual contribution — fixed, disclosed, reduced to practice, embodied, or commercially used under defined conditions — can generate a protected legal interest; and the observation that creation is never creation from nothing does not settle that question, since all production is the rearrangement of existing matter and symbol into valuable form, and the law already grants protected interests to produced value across many domains. Fourth, and decisively for the comparison, abolishing intellectual property does not abolish control; it migrates control out of a time-limited, disclosed, expiring public right and into perpetual, undisclosed, private enclosure — trade secrecy, contractual lock-in, and the technical lock — and that migration systematically favours large capital, which can keep secrets at scale, draft and enforce contracts of adhesion, and build technical locks, over the individual creator, who needs disclosure to raise capital and cannot maintain secrecy at scale. The essay illustrates the migration with the right-to-repair conflict, where control over agricultural equipment rests not on any single patent but on a stack of copyright anti-circumvention law, trade secrecy, technical locks, and the manufacturer’s claim that the buyer holds only an implied licence; and it treats the pharmaceutical case with deliberate care, noting that the headline figure for the cost of developing an approved drug is both large and contested, which is exactly why the recovery-versus-rent question is empirical and sector-specific rather than resolvable by slogan. The conclusion is not that every existing statute is justified — bad intellectual property supports reform, not abolition — but that the burden the abolitionist must discharge is to specify the replacement system and show it produces a better society, and that, measured against the contractual and secretive enclosure that abolition actually yields, a properly limited system of creator rights has the stronger comparative-institutional case. The standard throughout is human flourishing, not doctrinal purity.

I. Against the tribal frame

The first thing to say about the intellectual-property debate is that the usual way of conducting it guarantees that nothing will be learned. One announces oneself “pro-IP” or “anti-IP,” and the announcement functions as a tribal badge: it sorts the speaker into a camp, summons the camp’s stock arguments, and licenses the dismissal of the other camp’s. The label arrives before the analysis and replaces it. This is not philosophy; it is identity maintenance conducted in the vocabulary of philosophy.

So let me set the badges aside. This essay is not a defence of every existing intellectual-property statute — many of them are bad, and I will say where. It is not a utilitarian brief, nor an Objectivist one, nor an anarchist one. It is an attempt to ask the only question that can actually be argued rather than merely asserted: what institutional order best enables human beings to live, create, invent, disclose, invest, recover their costs, compete, learn, and pass an improving civilisation forward? That question does not presuppose an answer. It might turn out that a world with much weaker intellectual property, or none, serves those ends better; it might turn out that a limited and reformed system serves them better. The point of starting from the question rather than the slogan is that the answer is allowed to be discovered rather than declared.

Framed this way, the debate becomes a comparative-institutional one, and that reframing immediately exposes a fallacy so common that most anti-IP arguments are built on it. The fallacy is asymmetric comparison: setting the real intellectual-property system, with all its genuine pathologies — patent trolls, copyright terms that now stretch absurdly long, ruinous litigation costs, patents granted on obvious or trivial “inventions” — against an imagined alternative world of frictionless free knowledge and costless competition. That is not a comparison between two institutions. It is a comparison between one institution’s real flaws and another institution’s imagined perfection, and it is rigged from the outset, because every real institution loses to an imaginary one.

Figure 1. The methodological correction. The valid comparison is between a reformed system of creator rights and the actual no-IP substitute, not between flawed real IP and an imagined frictionless ideal.

The valid comparison is between institutions that could actually exist. On one side, a limited and reformed system of creator rights — bounded in scope, limited in time, defeated by independent creation and prior art, and pruned of the abuses that disfigure the current system. On the other, the system that abolition would actually leave standing, which is emphatically not a commons of open knowledge. It is the system of private ordering that firms already use and would use far more in the absence of public creator rights: contract, trade secrecy, non-disclosure agreements, no-reverse-engineering clauses, technical protection measures, access controls, server-side software, licensing in place of sale, binding arbitration, and the simple advantages of corporate scale. The serious question is which of these two produces the better outcomes. An argument that skips this comparison — that treats “the current law is flawed” as if it were a proof that abolition is correct — has not done the work. Flaws in a real institution are an argument for reforming it unless and until someone shows that the available alternative is better, and showing that requires describing the alternative honestly.

II. Property is already institutional

A great deal of anti-IP argument rests on an unstated picture of what property really is: tangible, corporeal, a scarce physical object you can stand on or hold or kick. On that picture intellectual property looks like an impostor — a legal fiction, an intangible, a thing conjured by the state rather than found in the world. But the picture is false to the actual structure of modern property, and once that is seen, the “intangible, therefore not real property” move loses its force.

Consider the assets that make up the overwhelming bulk of value in a modern economy. A debt is not a physical object; it is an enforceable claim against a person, a chose in action. A company share is not a thing; it is a bundle of rights against an abstraction called a corporation, itself a legal person that exists nowhere in physical space. A bank balance is a structured claim recognised by contract, accounting, and statute. Security interests, mortgages, leases, easements, trusts, options, goodwill, licences, insurance claims, insolvency entitlements, securities held through intermediaries — every one of these is a legally constructed interest in value, not a tangible object, and every one is unquestioningly treated as property, bought, sold, pledged, inherited, and defended in court. The balance sheets of the world’s largest firms are dominated by intangible and institutional assets. To say that intellectual property cannot be genuine property because it is intangible and legally constructed is to say something that, applied consistently, would dissolve debts, shares, companies, and most of commercial law along with it.

The careful anti-IP theorist has a reply to this, and it must be met rather than dodged. The reply is that these intangible commercial interests are, at bottom, still claims over scarce physical resources or over persons — a debt is ultimately a claim on the debtor’s assets, a share on the company’s, and so on — whereas intellectual property purports to be a claim over something non-rivalrous, an idea or pattern that can be used by everyone at once without depletion. That is a real distinction, and it deserves a real answer. But notice what the answer must be. It cannot be the crude one, “these are intangible too, so anything goes.” It must be this: if institutional legal relations can legitimately structure claims to value, priority, transfer, exclusion, and enforcement in debt, shares, companies, security interests, and — tellingly — confidential information and trade secrets, then intellectual property cannot be rejected merely because it is institutional or non-corporeal. Some further argument is required to distinguish it. The intangibility objection, by itself, proves far too much; it is not the end of the argument but a demand for the real one. And the real one, as we will see, turns out to be about negative servitudes and non-rivalry, not about tangibility at all.

III. The negative-servitude argument and its hidden premise

The strongest version of the anti-IP case — the one associated above all with Stephan Kinsella — does not rest on the crude intangibility claim. It rests on a more precise and more powerful idea: that an intellectual-property right is a non-consensual negative servitude imposed on other people’s material property. A patent, on this view, does not give the patentee a thing; it gives him a veto over what everyone else may do with their things — their factories, their machines, their paper, their computers — and a veto over another’s property that the other never consented to is, the argument runs, an illegitimate encumbrance, a negative easement conjured without agreement. This is a serious argument, and it is worth setting out in its full logical form, because its structure is where it both gets its force and conceals its weakness.

The argument runs like this. A owns a factory. B holds a patent that prevents A from using the factory in a certain way — say, to manufacture a device embodying B’s claimed invention. Therefore B has imposed a non-consensual negative servitude on A’s property. Non-consensual negative servitudes are illegitimate. Therefore B’s patent is illegitimate. Each step appears to follow, and the conclusion appears forced.

Figure 2. The negative-servitude argument carries a hidden premise. Calling the right a servitude presupposes that the owner’s liberty already included the restricted use — which is precisely what is in dispute.

The trouble is in the move from step two to step three, and it is a buried assumption rather than an inference. To describe B’s patent as imposing a servitude on A’s property is already to assume that the use in question — manufacturing B’s invention — was part of A’s legitimate liberty as owner of the factory to begin with. If it was, then yes, the patent takes something from A. But if it was not — if A’s ownership of the factory never included the liberty to appropriate B’s specific protected contribution — then the patent takes nothing from A that was his, and the “servitude” framing simply mislabels the situation. The whole question is whether the creator B has a protected interest capable of delimiting what A may do, and the negative-servitude argument does not answer that question; it presupposes a particular answer to it. It assumes that the creator has no such interest, and then derives the illegitimacy of the right from that assumption. The label “negative servitude” is doing the work of an argument it has not made.

This is clearer once one notices that every enforceable right restricts some uses of material property, and we do not regard that restriction as itself a usurpation. A landowner may not use his land to create a nuisance that poisons his neighbour. A printer may not use his own press and his own paper to run off forged bonds. A factory owner may not use his own machines to exploit a stolen trade secret. A trustee may not treat trust assets in his possession as his own. A person under contract may not use his own voice and his own time in breach of an agreed restriction. In every case a right held by someone else restricts what the material owner may do with his own physical property — and in no case do we think this shows the other’s right to be an illegitimate servitude. Why not? Because in each case the restricted use was never part of the owner’s legitimate liberty. The question, then, is never “does this right restrict some material action?” — all rights do — but “was the restricted action ever part of the owner’s rightful liberty in the first place?” The negative-servitude argument cannot answer that question by restating it as a label. It has to engage the real issue, which is whether intellectual creation can generate a protected interest at all.

IV. The real question: can created value generate a protected interest?

So the debate is not, despite the slogans, about whether “you can own ideas.” That phrase is too vague to be useful, and stated baldly it is a straw man, because no defensible version of intellectual property claims ownership of abstract thoughts floating free in the mind. The serious question is precise: can a specific, produced, objective intellectual contribution — one that is fixed in a tangible medium, or disclosed, or reduced to practice, or embodied in a device, or expressed in a determinate form, or commercially deployed under defined conditions — generate a legally protected interest? That is a far narrower claim than “owning ideas,” and answering it requires distinctions the slogan erases: between an abstract idea and a fixed expression of it, between a vague concept and a specific patent claim, between knowing a method and being the one who reduced it to a working, disclosed practice, between a brand identifier and the goodwill it carries, between confidential information and the public domain.

A common anti-IP move at this point is to say that creation cannot be a source of any right because no one creates anything from nothing — all human production merely rearranges pre-existing matter, energy, symbol, and information. The farmer does not create the soil; the builder does not create atoms; the engineer does not create physics; the author does not create language; the programmer does not create mathematics; the inventor does not create natural law. This is true, and it is also beside the point, because the relevant act was never metaphysical creation from nothing. It is economically and institutionally meaningful production: the ordering of existing materials into a valuable arrangement, system, expression, process, or embodiment that did not exist before in that form. The farmer who does not create the soil nonetheless produces a crop; the law has no difficulty granting him a protected interest in it, through doctrines of accession, specification, improvement, and produced assets, even though every atom of the crop pre-existed his labour. The observation that production is rearrangement does not distinguish intellectual creation from any other kind of production, all of which is rearrangement, and none of which is thereby denied the capacity to generate protected interests. The denial that produced intellectual value can generate a protected interest is therefore not a conclusion that follows from “nobody creates matter.” It is a bare premise, and the honest course is to bring it into the open and argue it, rather than smuggle it in under a truism about atoms.

None of this establishes that any particular intellectual-property right is justified. It establishes something narrower and prior: that the question is a real one, that it cannot be settled by the intangibility slogan or the negative-servitude label or the creation-from-nothing truism, and that defending some creator rights does not commit one to the absurd position of owning vague thoughts. It clears the ground. What is built on the ground depends on the comparison, and the comparison depends on being honest about what abolition would actually produce.

V. Disclosure, expiry, and the different categories

Two features of the existing system are systematically omitted by the abolitionist account, and both bear on the comparison. The first is that a patent is not only an exclusionary right; it is also a disclosure mechanism. The patent bargain, in its design, trades a limited period of exclusivity for the public teaching of how the invention works — a teaching that enters the published record, becomes available for others to learn from and design around, and falls into the public domain when the patent expires. This means that an expired patent is not evidence that the patent was irrelevant; it may be evidence that the disclosure-and-expiry structure worked exactly as intended, drawing the invention into the open in return for a bounded period of protection. Whether the system in fact delivers this disclosure benefit is an empirical question, and a fair one — but the abolitionist who describes patents only as monopoly grants has omitted half of the instrument’s structure.

The second omission is that “intellectual property” is not one thing, and treating it as a single bloc is a category error that corrupts the analysis. Patent, copyright, trade mark, trade secret, and design rights have different justifications, different structures, different durations, and different failure modes. A patent confers time-limited exclusivity over a disclosed, claimed invention. Copyright protects a particular expression, not the idea behind it, and is defeated by independent creation. A trade mark prevents consumer confusion as to source and protects goodwill; it does not restrain competition in the underlying product. A trade secret rests on confidentiality and the wrong of its breach, and — crucially for what follows — it is the mechanism that expands when patents contract. Design rights protect defined visual features. An argument that lumps these together and condemns or defends them as a unit cannot be right, because what is true of one is frequently false of another. A serious treatment must take them apart, and in particular must notice that the categories trade off against one another: weaken patents, and you do not get openness; you get more reliance on the category specifically designed to keep knowledge secret.

VI. The decisive point: control migrates, it does not vanish

Here is the heart of the comparative case, and the point the asymmetric comparison is designed to obscure. Abolishing intellectual property does not abolish control over created value. It relocates control — out of a public, time-limited, disclosed, expiring legal right, and into private mechanisms that are perpetual, undisclosed, and unaccountable. The control does not disappear. It changes form, and the new form is worse on precisely the dimensions the abolitionist claims to care about.

What are the private mechanisms into which control migrates? They are already in use, and they would expand to fill the space that abolition created: trade secrecy, non-disclosure agreements, employee non-compete and confidentiality restrictions, compartmentalised research, black-box devices, encrypted firmware, server-side software that never leaves the vendor’s control, diagnostic and repair locks, cloud dependencies, restricted APIs, closed standards, contractual access terms, reverse-engineering prohibitions, licensing in place of sale, binding arbitration clauses, and class-action waivers. Each of these is a method of controlling created value without disclosure, without a time limit, and without the public-domain endpoint that a patent or copyright eventually reaches. Trade secrecy, in particular, is the natural substitute for patents: where a patent discloses in exchange for bounded protection, a trade secret protects precisely by not disclosing, potentially forever. The abolitionist who imagines that removing patents yields open knowledge has the direction of the effect backwards. It yields less disclosure, not more.

Figure 3. Abolishing IP migrates control rather than removing it. The private substitutes are perpetual, undisclosed, and favour large capital over the individual creator.

And this is where the distributive consequence becomes decisive, because the migration is not neutral between large and small. The anti-IP argument often imagines that abolition hurts entrenched monopolists and helps everyone else; the opposite is at least as likely. Large firms are far better positioned to exploit the private mechanisms than small creators are. A large firm can maintain secrecy at scale, behind compartmentalised research and a wall of NDAs; a lone inventor usually cannot, because he needs to disclose his invention to attract the investment that would let him build it. A large firm can draft, impose, and enforce contracts of adhesion — clickwrap, shrinkwrap, platform terms, dealer agreements, licence-not-sale — across millions of transactions; a small creator has weak bargaining power and cannot easily litigate a breach. A large firm can build the technical locks — encrypted firmware, server-side software, authentication gates — that enclose a product against repair and reverse-engineering; a small creator cannot. A large firm has capital, speed to market, distribution, brand, data, manufacturing scale, and the ability to absorb failure; it can survive in a world without intellectual property by deploying all the other mechanisms of control it commands. The small creator, stripped of a recoverable period of exclusivity, is left exposed to exactly the fast copying that exclusivity existed to prevent — and copied, very often, by the large firm with the scale to bring the product to market faster and cheaper. Abolition, in other words, may transfer power from individual creators to capital-rich corporations. It is not obviously the friend of the little inventor that its rhetoric supposes.

The right-to-repair conflict is a precise illustration of how the stack of private control actually works, and it shows that the “control” at issue is rarely a single patent that abolition would dissolve. Modern agricultural equipment is locked against owner and independent repair not by one right but by a layered combination: copyright in the embedded software, backed by the anti-circumvention provisions of Section 1201 of the Digital Millennium Copyright Act, which make it unlawful to bypass the digital lock even to repair a machine one owns; trade secrecy in the diagnostic information; the technical lock itself, the engine control unit that refuses to function after an unauthorised repair; and, underpinning all of it, the manufacturer’s contractual claim — John Deere has argued since 2015 — that because the tractor runs on code, the farmer does not really own it but holds only “an implied license for the life of the vehicle to operate the vehicle.” Strip out the patent layer, or even the copyright layer, and the rest of the stack stands: the contract, the trade secret, and the technical lock remain. This is why the contest has had to be fought on multiple fronts at once — a Federal Trade Commission and state-attorneys-general antitrust suit against the manufacturer that survived a motion to dismiss, renewable repair exemptions to Section 1201 granted by the Copyright Office, state right-to-repair statutes beginning with Colorado’s, and a proposed federal FARM Act — and it is telling that even these reform efforts generally preserve the manufacturer’s trade secrets. The lesson for the abolition debate is exact: control over created value lives in a stack of contract, secrecy, technical measures, and law, and removing any single layer leaves the others in place. Abolishing “intellectual property” in the narrow sense would not free the farmer; it would leave him facing the contract, the secret, and the lock.

VII. The pharmaceutical case, handled with care

No discussion of this kind can avoid pharmaceuticals, and none should pretend the question is easy, because it is the case where the stakes of the recovery-versus-rent distinction are highest and the empirical uncertainty greatest. The framework I am working from flags this area as one requiring special care, and it is right to.

Here is the disciplined version of the issue. Drug development is characterised by very high fixed costs, very high failure rates, long timelines, and heavy regulatory burdens; a great many candidates fail for every one that reaches the market, and the cost of the failures is borne by the successes. The most-cited estimate of the cost of bringing a single approved drug to market — from the Tufts Center for the Study of Drug Development — puts the figure at roughly $2.6 billion, a number that explicitly includes both the out-of-pocket cost of the failures and the cost of capital over a decade-plus development period. If that figure is even approximately right, then prices charged during the period of patent exclusivity are not, or not only, monopoly extraction; they are substantially the recovery of an enormous risk-adjusted fixed cost, without the prospect of which the investment would not be made. That cuts hard against the crude claim that patent pricing is pure rent.

But intellectual honesty requires the other half, and the framework demands it. The $2.6 billion figure is contested. Critics point out that it rests in part on assumptions rather than project-level data, that the cost-of-capital adjustment roughly doubles the headline number, that the sample is self-selected from industry-supplied data, and that the figure has been used to justify high prices in public debate. Some portion of exclusivity-period pricing plainly is rent extraction rather than recovery; the question is the proportion, and it varies by drug, by company, and by therapeutic area. The honest conclusion is therefore not “patents are recovery, so prices are justified,” nor “patents are monopoly, so prices are extraction,” but that the recovery-versus-rent split is an empirical, sector-specific question that neither slogan answers — which is precisely why the comparative-institutional method matters more here than anywhere else. And the comparison still has to be run against the real alternative: in a world without pharmaceutical patents, what recovers the cost of the failures, and what prevents a developer’s disclosed compound from being manufactured immediately by a competitor who bore none of the development risk? The honest answer is that the alternative mechanisms — secrecy (largely unavailable for a molecule that must be disclosed to be approved and prescribed), first-mover advantage, regulatory exclusivity, or public funding of development — each have serious limits, and an abolitionist case has to specify which of them does the work and show that it does it better. That is a hard question, and pretending it is easy in either direction is the one thing the evidence forbids.

VIII. Burden of proof, reform, and the standard of judgment

Two asymmetries in how this debate is usually conducted deserve to be named and corrected, because correcting them is most of what a fair treatment requires.

The first is an asymmetry of burden. The abolitionist frequently demands that the defender of intellectual property prove every benefit of every right, while offering, as the substitute, only the airy assurance that “the market will discover alternatives.” That is not an argument; it is burden-shifting. A call for abolition is not mere criticism of existing arrangements; it is itself a positive policy claim, and it carries its own burden of proof. The abolitionist must specify what replaces the abolished rights, how disclosure is preserved or improved, how small inventors and authors recover their costs, how corporate secrecy is restrained, how contractual lock-in is prevented from expanding, how innovation incentives change, how consumers are affected, and how long-term knowledge accumulation fares. “The market will work it out” is not a replacement model; it is the refusal to provide one. Both sides bear a burden here, and the abolitionist’s is at least as heavy as the defender’s, because abolition is the proposal for change.

The second is an asymmetry of comparison, already named, but worth restating as a rule of judgment: the fact that the current law is flawed is not a proof that abolition is correct. It is a proof that the current law should be reformed — its terms shortened where they are too long, its trolls disarmed, its obvious patents denied, its litigation made less ruinous — unless someone independently shows that the available alternative is better than a reformed system. Bad intellectual property is an argument for good intellectual property, not for none, in exactly the way that a corrupt court is an argument for a better court rather than for the abolition of courts. The move from “this institution is performing badly” to “this institution should not exist” requires the extra step of showing that what replaces it performs better, and that step is the one the comparative-institutional method exists to force.

And the standard against which both the reformed system and its abolitionist alternative must finally be judged is not doctrinal purity — not whether a rule follows from homesteading, or satisfies a definition of property, or coheres with an ideology. It is institutional performance measured against human flourishing: whether the order in question enables invention, disclosure, authorship, capital formation, exchange, competition, the recovery of investment, the protection of small creators, consumer benefit, the diffusion of knowledge, institutional trust, and long-term progress. Property rules themselves are institutional norms that exist to support human action, cooperation, and the reduction of conflict; the question of what rights should exist cannot be severed from the social function of rights. The anti-IP theorist who says that justice is prior to consequences is not answered by denying justice, but by observing that the content of property justice is itself worked out partly by reference to what property is for. That is not consequentialism swallowing rights; it is the recognition that rights and their function are not finally separable.

IX. Conclusion

The honest position on intellectual property is not a tribal one, and the honest essay cannot end with a slogan, because the case it has made is comparative rather than absolute. Let me state the conclusion at the level of confidence the argument actually supports.

The category of intellectual property is not refuted by the observations usually marshalled against it. It is not refuted by intangibility, because debts, shares, companies, trusts, and confidential information are all intangible institutional interests in value and are property nonetheless. It is not refuted by the negative-servitude argument, because that argument presupposes the very thing in dispute — that the creator has no protected interest — and dresses the presupposition as an inference. It is not refuted by the truth that creation is never creation from nothing, because all production is rearrangement and the law protects produced value across every domain. What remains, once these are cleared away, is a genuine and difficult question — whether specific, produced, objective intellectual contributions should generate protected interests, and under what limits — and that question is to be answered by comparing real institutions, not by measuring a flawed system against an imagined utopia.

When the comparison is run honestly, the decisive fact is that abolishing intellectual property does not abolish control over created value; it migrates that control out of a public, disclosed, time-limited, expiring right and into private, undisclosed, perpetual enclosure — trade secrecy, the contract of adhesion, and the technical lock — and that migration systematically favours large capital over the individual creator, because capital can keep secrets at scale, draft and enforce the contracts, and build the locks, while the lone creator needs disclosure to raise capital and cannot do any of these things. The right-to-repair conflict shows the stack of private control operating in the open; the pharmaceutical case shows why the recovery-versus-rent question is empirical and sector-specific rather than resolvable by slogan; and both show that the abolitionist’s imagined world of open knowledge is not what abolition would actually produce.

None of this is a defence of the existing system as it stands. The existing system has real and serious pathologies, and they call for real and serious reform — shorter terms, better examination, the disarming of trolls, the curtailment of abusive litigation, and a clear-eyed willingness to prune rights that do not earn their keep. But reform is the conclusion that the flaws support, not abolition, because the burden on the abolitionist is to specify the replacement system and show that it produces a better society, and measured against the contractual and secretive enclosure that abolition actually yields, a properly limited system of creator rights — bounded, disclosed, expiring, and distinct across its categories — has the stronger comparative-institutional case. The test, from first to last, is not which position is purer. It is which institutional order better enables human beings to create, disclose, invest, compete, learn, and flourish. That is the debate worth having, and it is the only one that the slogans, on either side, are designed to prevent.

References

Property theory and the institutional character of ownership

• Honoré, A. M. “Ownership.” In Oxford Essays in Jurisprudence, ed. A. G. Guest. Oxford: Oxford University Press, 1961. (The incidents of ownership; property as a bundle of legal relations rather than a thing.)

• Penner, J. E. The Idea of Property in Law. Oxford: Oxford University Press, 1997. (Property as a relation among persons with respect to things; the exclusion thesis.)

• De Soto, Hernando. The Mystery of Capital. New York: Basic Books, 2000. (Property as a legal-institutional construction; assets as “dead capital” without formal title.)

• Smith, Henry E., and Thomas W. Merrill. “What Happened to Property in Law and Economics?” Yale Law Journal 111 (2001): 357–398. (Property as in rem rights, against the pure contractarian reduction.)

The anti-IP position under examination

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (2001): 1–53. (IP as a non-consensual negative servitude/easement over others’ material property; the central argument addressed here.)

• Boldrin, Michele, and David K. Levine. Against Intellectual Monopoly. Cambridge: Cambridge University Press, 2008. (The economic case against patents and copyright; engaged as the empirical-economic variant.)

The economics of innovation, disclosure, and appropriability

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity. Princeton: Princeton University Press (NBER), 1962. (Inappropriability and the disclosure paradox.)

• Landes, William M., and Richard A. Posner. “An Economic Analysis of Copyright Law.” Journal of Legal Studies 18, no. 2 (1989): 325–363. (Fixed cost of creation vs near-zero cost of copying.)

• DiMasi, Joseph A., Henry G. Grabowski, and Ronald W. Hansen. “Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs.” Journal of Health Economics 47 (2016): 20–33. (The ~$2.6 billion per-approved-drug estimate, including failures and cost of capital — cited here as the contested headline figure, with its criticisms noted.)

The migration of control: contract, secrecy, and technical locks

• Digital Millennium Copyright Act, 17 U.S.C. § 1201 (1998). (Anti-circumvention of technological protection measures; the renewable repair exemptions granted by the Copyright Office.)

• In re Deere & Co. Repair Services Antitrust Litigation, N.D. Ill. (consolidated MDL); and Federal Trade Commission and state attorneys general v. Deere & Co. (filed January 2025; motion to dismiss denied, 2025). (The right-to-repair stack: copyright anti-circumvention + trade secrecy + technical lock + the “implied license” claim.)

• Freedom for Agricultural Repair and Maintenance (FARM) Act (introduced 2025); Colorado Consumer Right to Repair Agricultural Equipment Act (2023, effective 2024). (Statutory reform efforts that generally preserve trade-secret protection.)

• Radin, Margaret Jane. Boilerplate: The Fine Print, Vending Machines, and the Rule of Law. Princeton: Princeton University Press, 2013. (Contracts of adhesion as private regulation by the strongest drafter.)

The Austrian and classical-liberal background (distinguished, not conflated)

• Mises, Ludwig von. Human Action. 3rd rev. ed. Chicago: Henry Regnery, 1966. (Intellectual products and the “delimitation of property rights”; the calculation argument; the state as indispensable — distinguishing limited-government Austrianism from anarchism.)

• Hayek, F. A. Law, Legislation and Liberty. 3 vols. Chicago: University of Chicago Press, 1973–1979. (The legal order as the institutional precondition of the market; the delimitation of protected domains.)

Note on method and scope. This essay is built from an analytical framework whose explicit standard is comparative-institutional rather than tribal, and it follows that framework’s discipline: empirical claims are sourced and flagged where contested, and claims requiring validation are not asserted as settled. In particular, no claim is made that any specific technology “depended on patents”; the pharmaceutical cost figure (Tufts/DiMasi et al., ~$2.6 billion per approved drug) is presented as the cited headline estimate together with the documented criticism that it is inflated and partly assumption-based, and the recovery-versus-rent question is treated as empirical and sector-specific rather than resolved. The right-to-repair facts — DMCA § 1201, the John Deere “implied license” position, the FTC and state-AG antitrust suit and its survival of a motion to dismiss, the Copyright Office repair exemptions, the Colorado statute, and the FARM Act — reflect current reporting verified at the source level. The representation of the anti-IP position, especially the negative-servitude argument associated with Kinsella, is stated at full strength before it is criticised, and Austrian limited-government economics (Mises, Hayek) is distinguished from anarcho-capitalism (Rothbard, Hoppe) rather than conflated with it. The essay defends a comparative thesis — that a limited and reformed system of creator rights outperforms the contractual-and-secretive enclosure that abolition actually yields — and explicitly concedes the serious pathologies of the current system, which it treats as grounds for reform rather than abolition. The standard of judgment throughout is institutional performance measured against human flourishing.

Keywords: intellectual property, copyright, AI training, generative models, fair use, text and data mining, TDM, appropriability, public goods, fixed costs, marginal cost, idea–expression dichotomy, market harm, Bartz v Anthropic, Kadrey v Meta, Thomson Reuters v Ross, EU Copyright Directive, opt-out, licensing.

Abstract

The argument that abolishing intellectual property would licence “appropriation by scale” — the creator bearing the fixed cost of producing a work while a better-capitalised copier reproduces it at near-zero marginal cost — has a new and literal defendant: the generative model trained on the whole corpus. The structure is the same as the single copyist, only aggregated. Thousands of authors, journalists, researchers, illustrators, and coders each bear a fixed cost to produce one work; a training run ingests them all and yields a system that can reproduce the capability of the corpus — summaries, style, substitutes — at marginal cost approaching zero. This is, on its face, the appropriability problem of Arrow and of Landes and Posner, scaled to the level of an entire creative economy and concentrated in the hands of whoever owns the compute and the distribution. But honesty forbids the polemic that the abolitionist’s critics would most enjoy, because the case against AI training is genuinely harder than the case against the photocopier, and for two reasons the slogans on the creator side ignore. First, a trained model is not a verbatim copy: it ingests expression but, when working as designed, outputs new strings, and the idea/expression dichotomy together with Feist‘s rule that facts and learning are free cut both ways — what a model extracts is often exactly the unprotected layer. Second, copyright protects expression, not the act of reading or learning, and whether ingestion is “copying” or “use” is the live, unsettled legal question, not a settled wrong — a question the 2025 American decisions answered in conflicting ways (training held transformative in Bartz v. Anthropic and Kadrey v. Meta; market-substitution fatal to fair use in Thomson Reuters v. Ross; a $1.5 billion settlement turning not on training but on how the books were obtained), and which the EU has answered with a text-and-data-mining exception subject to rightsholder opt-out. The disciplined position, carried over from the companion essays, is the bounded-form test: training that reproduces a work’s protected expression, or produces a market substitute for it, is appropriation of the bounded form and is the copyist’s problem at scale; training that takes only the unprotected layer — facts, ideas, style, the statistical shape of language — and outputs new expression is the reader learning, which the law has always permitted. The hard cases sit between: ingestion makes intermediate copies even when the output infringes nothing, and a model that copies no single work may still erode the market for an entire class of works — appropriability harm without verbatim copying. The “you still have your file” defence fails here exactly as it failed for the photocopier; but so does “all training is theft,” and so does “all training is fair use.” The remedy is neither abolishing intellectual property nor banning the models — both are slogans — but pricing the appropriation of the form while keeping the learning free: licensing, opt-out and text-and-data-mining regimes, collective bargaining, and carve-outs. The principle is the one the companion essays defended. Only the machine is new.

I. The oldest problem, the newest defendant

The companion essays made a structural argument. Creative and inventive goods carry a high fixed cost of creation and a near-zero marginal cost of reproduction; in the absence of any right to exclude, the creator bears the cost of resolving the uncertainty — writing the book, discovering the invention, compiling the data — while a better-capitalised copier waits, reproduces the finished product at marginal cost, undersells the originator who alone paid the fixed cost, and captures the return. That is the appropriability problem, identified by Kenneth Arrow in 1962 and translated into the law of copyright by Landes and Posner in 1989, and the companion essays argued that abolishing intellectual property does not free the market but hands it to whoever owns the means of reproduction — appropriation by scale.

That argument was written as if its defendant were a publisher with a printing press. It now has a defendant that is far larger and far stranger: the generative model trained on the corpus of nearly everything that has been written, drawn, photographed, or coded. And the structure is identical, only aggregated. Where the single copyist took one finished book, the training run takes the finished output of a whole creative economy. Where the copyist reproduced one work at marginal cost, the model reproduces the capability of the corpus — the ability to summarise it, to write in its styles, to produce substitutes for the kinds of thing it contains — at a marginal cost that, once the model exists, approaches zero. The fixed cost is borne, as before, by the creators: thousands of them, each paying in time and risk to produce one work before any return. The capability is captured, as before, by the party with the complementary assets — here the compute, the data pipeline, and the distribution. The copyist’s problem did not go away. It went to industrial scale and acquired a new owner.

Figure 1. The copyist’s problem at industrial scale. The asymmetry is the same as the single copyist’s; the two disanalogies in amber are why the case is harder than the photocopier’s, and why the polemic must not skip them.

If the analysis stopped there, the conclusion would write itself: the model is the parasite the abolitionist warned about, only enormous, and the creators are the producers being stripped at the point of value. That conclusion is half-true, and a polemic that stopped there would be lying by omission — because the case against the machine is genuinely harder than the case against the press, and the next two sections are about why.

II. Why this is harder than the photocopier

The photocopier makes a copy. That is the whole of what it does, and it is why the copyist case is easy: the output is the input, reproduced. The trained model does something the photocopier does not, and the difference is not a technicality — it goes to the core of what copyright protects.

The first disanalogy is that a trained model is not a verbatim copy, and when it works as designed it does not output one. It ingests an enormous quantity of expression and, from it, learns a statistical structure — which words tend to follow which, how arguments are shaped, what the styles and registers of human writing are — and then it generates new strings that did not appear in any input. Copyright has a precise and ancient way of describing what is happening here, and it cuts against the creators as often as for them. The idea/expression dichotomy protects the specific expression and withholds protection from the idea, the method, the system, and the fact; and Feist holds that facts and the products of effort are free, that copyright rewards originality of expression and not labour or learning. What a model extracts from a text is, very often, exactly the unprotected layer — the facts it states, the ideas it advances, the statistical shape of the language it is written in — none of which copyright has ever given anyone the right to control. A human who reads a thousand novels and learns how to write one infringes nothing, because what he took was the unprotectable lesson, not the protectable form. The model’s defenders say it is doing the same thing at scale, and on the cases where the output is genuinely new expression, they are not obviously wrong. The dichotomy that the companion essays used against the abolitionist — IP protects the bounded form, not the idea — here cuts partly for the machine, because learning the idea is exactly what the law leaves free.

The second disanalogy is that copyright protects expression, not the act of reading or learning, and whether machine ingestion counts as “copying” or as “use” is a genuinely contested legal question rather than a settled wrong. When a person reads a book to learn from it, no copyright event occurs, because reading is not one of the exclusive rights. When a machine “reads” a book, it makes copies in the technical sense — it loads, tokenises, and processes the text, producing intermediate reproductions — and whether those copies are the actionable act, or whether they are an unactionable incident of a non-expressive use, is precisely the question the courts and legislatures are now fighting over. This is not a question the creator side gets to assume away by calling ingestion “theft,” and it is not a question the developer side gets to assume away by calling it “just reading.” It is open, and the honest essay says so.

These two disanalogies are why “you still have your file” — the slogan the companion essays demolished when a publisher used it to excuse reproducing a book — does not, by itself, resolve the AI case. It still fails as a defence: the fact that the author retains the manuscript is as irrelevant to a model that has appropriated the bounded form as it was to the publisher who reprinted it. But it no longer decides the case, because the prior question — did the model appropriate the bounded form at all, or only the unprotected layer the law leaves free? — is now live, and the answer is not the same for every model or every use.

III. The bounded-form test, applied to the machine

The companion essays’ organising idea was the bounded form: the wrong of copying is the appropriation of the determinate expressive or inventive form a producer made, not the taking of its value and not the failure to compensate effort. That test does real work here, because it sorts the AI cases that the slogans mash together.

Apply it as a question about the output, not a slogan about the technology. Does the model reproduce the work’s protected expression — verbatim or near-verbatim — or produce a market substitute for the specific work? If so, that is appropriation of the bounded form. The model is doing what the copyist did: displacing the work’s market, capturing the value of its form, occupying the demand the author created. The fact that it arrived at the reproduction through a training process rather than a photocopier changes nothing, because the bounded-form test looks at what was taken, not at the mechanism of taking. This is the copyist’s problem, at scale, and the appropriability argument applies with full force.

Or does the model take only the unprotected layer — the facts, the ideas, the style, the statistical shape of language — and output genuinely new expression? If so, that is, in principle, the reader learning. What was taken is what the law has always left free; the idea/expression dichotomy and Feist are not obstacles to this use but descriptions of why it is permitted. A model that has read widely and writes newly is doing what every author who has ever read widely and written newly has done, and the bounded-form test acquits it for the same reason it acquits the well-read novelist: the learning, not the form, is what was taken, and reading is not copying.

Figure 2. The bounded-form test applied to training. The two clear branches are easy; the value of the test is that it identifies the genuinely hard cases instead of pretending they are easy.

The point of the test is not that it makes every case easy. It is that it makes the easy cases easy and isolates the hard ones honestly, instead of the two slogans — “all training is theft” and “all training is fair use” — each of which simply asserts that one branch swallows the other. It does not. Some training reproduces the form; some takes only the lesson; and the interesting law, and the genuine moral difficulty, live in the cases that sit between.

IV. The hard case in the middle, named honestly

Two features of model training make the middle of that tree genuinely hard, and the disciplined position has to name them rather than wish them away.

The first is that ingestion makes intermediate copies even when the output infringes nothing. A model that never reproduces a single protected sentence still, in the course of training, loaded and processed the protected works — and whether that intermediate copying is itself the actionable act, or an unactionable incident of a transformative, non-expressive purpose, is the precise question on which the American courts split in 2025 and on which the EU and UK have legislated. In Bartz v. Anthropic, Judge Alsup held that using lawfully acquired books to train an LLM was “exceedingly” transformative fair use, on the reasoning that authors cannot exclude others from using works to learn and that the training produced something new rather than a substitute — but he treated each step separately and held that downloading and retaining pirated copies to build a permanent library was a distinct, non-transformative act, and it was that — the acquisition, not the training — that drove a settlement reported at roughly $1.5 billion. Two days later in Kadrey v. Meta, Judge Chhabria reached the same fair-use result on the record before him but by a different route, treating downloading and training as one integrated process and stressing that the plaintiffs had failed to prove market harm — while pointedly warning that stronger evidence of market substitution could change the outcome in a future case. The lesson is not that training is settled law; it is that two judges in the same courthouse in the same week agreed on the result and disagreed on almost everything about how to get there.

The second hard feature is the one the creator side states best and the developer side most wants to avoid: a model that copies no single work may still erode the market for an entire class of works. Even if no output reproduces any particular author’s expression, a system that can produce an endless supply of competent substitutes for the kind of thing a class of authors produces may compete away the return to producing it — appropriability harm without verbatim copying. This is exactly the level-versus-margin point the companion essays insisted on: the question is not whether any one work was copied but whether the incentive to produce the class of works survives. The contrast in the 2025 cases is instructive. Where the use was a direct market substitute, fair use failed: in Thomson Reuters v. Ross, a company that took Westlaw’s copyrightable headnotes to build a competing legal-research tool was held to have infringed, the court resting heavily on the fourth fair-use factor — the effect on the work’s potential market — because the product was a substitute for the original. Where the plaintiffs could not show that substitution, fair use succeeded, but the courts signalled that the market-harm question was the one that would decide the next round. The bounded-form test and the appropriability argument meet exactly here: market substitution is the signature of appropriation, and it is the thing the law is converging on as the real question, whether or not any individual work was reproduced.

So the “you still have it” defence fails in the middle of the tree too — the author’s retained file no more answers the erosion of his market than it answered the publisher’s reprint. But it fails on both sides of the slogan war. “All training is theft” is false, because much training takes only the unprotected layer and outputs new expression, which the law has always permitted. “All training is fair use” is false, because training that reproduces the form or floods the market with substitutes is the copyist’s appropriation at scale. The case turns on design and testing — what the model ingests, what it outputs, whether it substitutes — and not on a slogan in either direction.

V. Where the value moves, and what to do about it

Strip the technology away and look at the distribution, because that is where the companion essays’ first argument — the distributive face — reappears at platform scale.

On one side stand the bearers of the fixed cost: the authors, journalists, researchers, illustrators, coders, and photographers, in their thousands, each having paid in time and risk to create one work before any return arrived. On the other stands the party that captures the return: the firm with the compute, the data pipeline, the trained model, and the distribution, which reproduces the corpus’s capability at marginal cost and serves it at scale. The value moves from the first to the second, and it moves for the same reason it moved from the author to the publisher in the companion essay’s production-house scenario — because the party that bore the fixed cost has no way to capture the return once the capability has been extracted and can be served at marginal cost. This is appropriation by scale with a new and much larger defendant.

Figure 3. Where the value moves when the corpus trains the model. The distributive face of the companion essay, at platform scale.

But naming the harm is not the same as prescribing the cure, and the two cures the slogans offer are both wrong. “Abolish intellectual property” — the abolitionist’s answer — would not free the creators; it would remove the only lever they have and complete the transfer the companion essays described, leaving the model’s owner in undisputed possession of a capability built entirely from work it never had to pay for. “Ban the models” — the maximalist’s answer — would forgo the genuine public good of systems that can do what learning from the corpus enables, and would punish the legitimate branch of the tree (the reader that learns and writes newly) to reach the illegitimate one. Both answers fail because both ignore the bounded-form test: they treat training as all-appropriation or all-learning, when it is some of each.

The cure that follows from the test is to price the appropriation of the form while keeping the learning free, and the institutional materials for it already exist and are being built. The European Union’s approach is the clearest template: its text-and-data-mining regime permits reproductions for mining and model-building but subjects the general, commercial exception to a rightsholder opt-out — the right to reserve a work against training by machine-readable means — and the AI Act ties general-purpose model providers to respecting those reservations. The United Kingdom, in its December 2024 consultation, proposed an EU-style opt-out exception underpinned by transparency about training sources, with collective licensing expected to fill the gap. The American cases point the same way from the litigation side: train on lawfully acquired works and take the unprotected layer, and you are likely safe; build your corpus from piracy, or produce market substitutes, and you are not — a roadmap that prices the appropriation (through liability and settlement) while leaving the learning (through the transformative-use holding) free. None of these regimes is finished, and each has real defects — the opt-out puts the burden on the smallest and least-resourced rightsholders, transparency obligations are thin, and the licensing markets are immature. But they are all instances of the same correct move: not the abolition of intellectual property, and not the prohibition of the technology, but the construction of institutions that distinguish the appropriation of the form from the learning of the lesson, and charge for the first while permitting the second.

VI. Conclusion: the principle is old, the machine is new

The model that read everything is not a refutation of the companion essays’ argument; it is its largest confirmation. The appropriability problem they described — fixed cost borne by the creator, near-zero marginal cost captured by the better-capitalised reproducer — is exactly the problem that training on the corpus presents, scaled from one book to a creative economy and concentrated in the hands of whoever owns the compute. The “you still have your file” defence fails against the machine for the same reason it failed against the press: the author’s retained copy was never the point, and the appropriation of the form and its market is.

But the machine is genuinely harder than the press, and the honesty the companion essays demanded of the abolitionist is owed here in the other direction too. A trained model is not a photocopier; it ingests expression and, when it works as designed, outputs new expression, taking the unprotected layer — facts, ideas, style — that the law has always left free. Copyright protects the form, not the act of reading, and whether machine ingestion is copying or use is the live question the courts split on in 2025 and the legislatures are now answering. The slogans on both sides are lying: “all training is theft” ignores the legitimate branch where only the lesson is taken, and “all training is fair use” ignores the illegitimate branch where the form is reproduced and the market is flooded with substitutes.

The bounded-form test cuts the knot the slogans cannot. Training that reproduces the protected form, or produces a market substitute, is the copyist’s appropriation at scale, and the appropriability argument applies with full force. Training that takes only the unprotected layer and outputs new expression is the reader learning, and the law has always permitted it. The hard cases — intermediate copies that infringe no output, models that copy no single work but erode the market for a class of works — are real, and the disciplined position names them instead of pretending either branch swallows the other. And the remedy is the one the companion essays implied: not the abolition of intellectual property and not the banning of the models, but the pricing of the appropriation of the form while the learning is kept free — through licensing, opt-out and text-and-data-mining regimes, collective bargaining, and carve-outs.

The creator who writes and the firm that trains on the writing without paying for it is the author and the production house again, at a scale the production house could never have reached. The issue is not that the creator “still has his file.” The issue is that the capability his work made possible has been appropriated and served back to the market by the party that bore none of his cost. That is the oldest problem in the economics of creation. The machine that read everything is only its newest, and largest, instance — and the answer is the old one: charge for the form, free the lesson, and build the institutions that can tell them apart.

References

Economic foundations

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity, 609–626. Princeton: Princeton University Press (NBER), 1962.

• Landes, William M., and Richard A. Posner. “An Economic Analysis of Copyright Law.” Journal of Legal Studies 18, no. 2 (1989): 325–363.

United States case law (2025)

• Bartz v. Anthropic PBC, No. 3:23-cv-04768 (N.D. Cal. June 23, 2025) (training on lawfully acquired books held “exceedingly” transformative fair use; pirated-library acquisition a distinct, non-transformative act; subsequent class certification and a settlement reported at ~$1.5 billion).

• Kadrey v. Meta Platforms, Inc., No. 3:23-cv-03417 (N.D. Cal. June 25, 2025) (training held fair use on the record; downloading and training treated as one integrated process; plaintiffs failed to prove market harm; court warned stronger market-harm evidence could change the result).

• Thomson Reuters Enterprise Centre GmbH v. ROSS Intelligence Inc., No. 1:20-cv-00613 (D. Del. Feb. 11, 2025) (Westlaw headnotes protectable; use to build a competing tool not fair use; decided on the fourth factor, market substitution; on appeal to the Third Circuit).

Statutory / legislative anchors

• 17 U.S.C. § 107 (fair use; the four factors, including effect on the potential market).

• 17 U.S.C. § 102(b) (idea/expression; exclusion of ideas, procedures, methods, facts).

• Directive (EU) 2019/790 (Copyright in the Digital Single Market), Arts. 3–4 (text-and-data-mining exceptions; Art. 4(3) rightsholder opt-out by machine-readable reservation).

• Regulation (EU) 2024/1689 (AI Act), Art. 53(1)(c) and Recital 106 (general-purpose model providers must respect Art. 4(3) reservations).

• UK Intellectual Property Office, Copyright and Artificial Intelligence consultation (17 December 2024) (proposing an EU-style TDM exception with opt-out and transparency).

Doctrinal anchors (developed in the companion essays)

• Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991) (originality, not effort; facts free).

• Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 (1985) (the bounded form and its exclusivity; first-publication right).

Note on method and scope. All case holdings, settlement figures, and statutory provisions above are drawn from current reporting and the instruments themselves and were verified at the page level before writing; none rests on an abstract. The 2025 American decisions are district-court rulings — one settled, one in continuing litigation, one on appeal — and are described as non-binding and unsettled, which they are; nothing here states that the law of AI training is fixed. Two propositions are flagged as contested rather than settled: whether intermediate copying in training is itself actionable, and whether non-substitutive training nonetheless inflicts cognisable market harm on a class of works. The essay defends the bounded-form principle and the appropriability analysis; it does not endorse any particular litigant, statute, or licensing scheme, and it states plainly that the opt-out and TDM regimes it points to have real and unresolved defects.

Abstract

The fight over intellectual property is usually staged as creators against pirates, but the sharper and more revealing fight is internal to the libertarian right, between two camps that agree on almost everything except this. On one side stands the Randian, Objectivist tradition, which holds that creation is a source of rights: you made the value, the value is a thing, therefore you own it, and intellectual property is simply the recognition of that ownership. On the other stands Hans-Hermann Hoppe — an anarcho-capitalist, not merely an Austrian economist, and it is his anarchism that does the work — together with Stephan Kinsella, who hold that the only legitimate object of ownership is the physical integrity of a scarce resource, that “value” is conferred by the shifting valuations of others and so cannot be owned, and that intellectual property is therefore a fiction. This essay argues that both camps are half-right and both are incomplete, and that the truth lies in a third position neither will occupy. Hoppe is correct that value is not the object of a right and that labour is not, by itself, a source of title — production is the transformation of things one already owns, not the conjuring of property out of effort, and the Randian “creation grounds ownership” move is a genuine non-sequitur. But Hoppe’s own conclusion does not follow from his premises, and the argument he uses to reach it is built out of the very intangibles it forbids: title, contract, consent, and — at the deepest level — the norms of argumentation on which his argumentation ethics rests. His decisive test, that a right must be ascertainable in advance without consulting anyone’s subjective valuation, does not kill intellectual property; it licenses the narrow version of it, because a would-be copier can determine in advance, and in complete ignorance of anyone’s valuation, whether he is reproducing another’s work. And his “no third possibility” dilemma — either all value-effects are aggression or none are — is exhaustive only if all effects on another’s economic position are one undifferentiated category, which they are not: competition by substitution and appropriation by reproduction are different acts with different objects, and only the second is what intellectual property addresses. The defensible account keeps Hoppe’s two concessions — value is not the object, effort is not the title — without his conclusion. Title to a made thing flows from prior ownership of the underlying resource plus first use; the right is in the bounded form and its exclusivity, not in its value and not in the labour that produced it; and that is exactly the interest the law has long protected (the right of first publication in Harper & Row) while withholding protection from effort as such (Feist‘s rejection of “sweat of the brow”). Rand reaches the right destination — that authorship can ground exclusivity — by the wrong route. Hoppe takes the right route — that ownership is not conjured from value or labour — to the wrong destination. The bounded-form view takes the route to the destination, and it is the position both halves of the libertarian world have talked themselves out of seeing.

I. A family quarrel, not a border war

Most writing on intellectual property pits creators against copyists, the industry against the pirates, the maximalist against the abolitionist. That is the public quarrel. The more illuminating one is private — a quarrel inside the libertarian family, between people who agree about almost everything else and discover, on this single question, that they cannot stand in the same room.

On one side is the Randian and Objectivist tradition. It holds that the mind is the root of production, that a created thing is the objectification of a creator’s thought and effort, and that to deny a creator the ownership of what he created is to deny him the product of his mind. On this account intellectual property is not an anomaly requiring special justification; it is the paradigm case of property, because it is the case where the link between the producer and the produced is least mediated by anything he did not make. Adam Mossoff, the most careful modern legal exponent of this line, argues that the creation of value through productive labour is the historical and moral foundation of property rights, and that copyright and patent are not monopolies but ordinary property in the fruits of that labour.

On the other side is Hans-Hermann Hoppe, and here a label matters, because the usual one is wrong. Hoppe is routinely filed under “Austrian economics,” and he is an Austrian economist, but his position on intellectual property is not driven by Austrian economics; it is driven by his anarcho-capitalism — his commitment to a stateless order of pure physical-property rights in which the only wrongs are physical invasions of body and owned scarce resources. It is the anarchism, not the economics, that generates the conclusion. Hoppe holds that what can be owned is the physical integrity of a scarce resource and nothing else; that the “value” of a thing is not a feature of the thing but a function of what other people will give for it; that one cannot own other people’s valuations; and that intellectual property, being a claim to control something other than the physical integrity of a scarce resource, is therefore illegitimate. Kinsella, his most prolific follower on this subject, has built an entire body of work elaborating the consequence.

The two camps are not arguing about a detail. They are arguing about the source of ownership itself — about what, at bottom, makes a thing yours. Rand says: you made its value. Hoppe says: you control its physical integrity, and nothing about value can enter into it. This essay’s claim is that each has hold of one true thing and one false thing, and that the true things, separated from the false ones, point to a position neither side defends.

Figure 1. Three accounts of where ownership of a made thing comes from. Rand and Hoppe are each half-right; the defensible view keeps Hoppe’s concessions without his conclusion.

II. Where Hoppe is right, and Rand is wrong

Concede the strong half of Hoppe’s case first, cleanly, because a refutation that will not grant what is true is worthless — and because what is true here is genuinely fatal to the Randian formulation.

Hoppe is right that value is not the object of a right. The value of a thing is what others will give for it, and what others will give is a fact about them — their wants, their alternatives, their valuations — not a fact about the thing or its owner. You do not, and cannot, own the contents of other people’s heads. If your restaurant loses custom because a better one opens across the street, the fall in your restaurant’s value is real and it is no wrong, because the value was never yours to begin with; it was always a standing bet on what strangers would choose to do. Rothbard put the same point in Man, Economy, and State: a man owns the physical existence of his property, not its monetary value, because monetary value is a function of what others will pay, and he has no vested right in the choices of others.

Hoppe is right, too, that labour is not, by itself, a source of title. This is the deeper and less obvious point, and it is where the Randian formulation breaks. “Creation” feels like it ought to ground ownership — you made it, so it is yours — but the feeling does not survive analysis. Production is not the creation of matter from nothing; it is the transformation of things from one configuration to another, and the things were already owned before the labour was applied. When you carve a statue from your block of marble, you do not come to own the statue because your labour “created” it; you own the statue because you already owned the marble, and the statue is simply the marble in a new shape that is worth more to you. The labour changed the configuration and raised the value; it did not generate a fresh title, because the title was already there, attached to the stuff. Mix your labour with a thing you do not own — carve a statue from someone else’s marble, or from marble owned by no one but lawfully closed to you — and your labour earns you no title at all; at most it earns you a mess. Labour is how you raise the value of what you own. It is not an independent source of owning.

This is precisely the move the Randian tradition makes and cannot justify. From the true premises that you laboured, that the result is valuable, and that the value would not exist but for your effort, it leaps to the conclusion that you therefore own anything you create that has value — and the leap is a non-sequitur. That you made something more valuable does not establish that you own a new thing; it establishes that you made the thing you already owned more valuable. To get from “I created this value” to “I own this value” you need a premise that value is an ownable object, and that premise is exactly what Hoppe correctly denies. Kinsella’s sharpest single observation, in his commentary on Mossoff, is just this: it is a non-sequitur to move from “your labour made your property more valuable” to “you own whatever you create that has value,” because the first is about the value of a thing you already own and the second smuggles in a new and ungrounded kind of property. On that narrow point he is right, and the Randian is wrong.

Figure 2. The creation–ownership inference. The valid chain grounds title in prior ownership; the Randian leap to “you own whatever you create that has value” is a non-sequitur — but so is the opposite leap, that nothing about the produced form can be owned.

So the Randian formula, taken literally, fails. “Creation grounds rights” is false as stated; creation raises value, and value is not a thing you own. If that were the end of the matter, Hoppe would win, and intellectual property would fall. It is not the end of the matter, because Hoppe’s conclusion does not follow from the concessions, and the argument he uses to bridge the gap collapses on inspection.

III. What “you cannot own value” does not prove

Here is the move on which the entire anti-IP case turns, and it is a conflation.

From the true proposition that you cannot own value, Hoppe and Kinsella slide to the conclusion that you cannot own the work whose value you are trying to protect. But those are different propositions, and the gap between them is the whole argument. “You cannot own value” refutes a claim to own the market’s response to your work — its demand, its price, the public’s taste, your expected profit. It says nothing whatever about a claim to control the reproduction of the specific bounded artifact you produced. The first is a claim against the choices of strangers; the second is a claim about a definite object — the form you made — and its exclusivity. To own the second is not to own anyone’s valuations; it is to own the thing, in the only sense in which an authored form can be a thing.

The decisive way to see that these come apart is to disaggregate the cases the anti-IP argument keeps fused. There is competition by substitution: a rival writes a better book, invents a superior machine, sells a cheaper substitute, and your work’s value falls. No wrong — and copyright and patent do not touch it, which is the point. And there is appropriation by reproduction: someone takes the specific expressive or inventive form you produced and runs off copies of it. These are not two intensities of one act called “harming value.” They are different acts with different objects. The first competes against your product; the second reproduces your product. The difference does not turn on value at all, and so “you cannot own value” — true of the first — is simply silent about the second.

This also disposes of Hoppe’s flourish that “a third possibility does not exist”: either every effect on another’s economic position is an actionable wrong (which would make all competition aggression, an absurdity) or none is (in which case copying is permitted). The dilemma is exhaustive only if all economic effects are one undifferentiated category. They are not. Once substitution and reproduction are pulled apart, the third possibility is exactly the one every intellectual-property system has always occupied: there is no right against substitutive value-reduction and there is a right against the reproduction of one’s bounded form, on grounds that never mention subjective valuation. The “no third possibility” rhetoric works by deleting the answer and then observing that nothing is left.

Figure 3. What “you cannot own value” proves and what it does not. The anti-IP case relies on fusing substitution and reproduction; disaggregate them and Hoppe’s dilemma dissolves.

IV. The test that backfires

Hoppe does not merely assert that value cannot be owned; he offers a criterion, and it is worth taking seriously because it is the strongest thing in his argument — and because, applied honestly, it cuts the other way.

The criterion is ascertainability in advance. A right, Hoppe argues, must have a boundary that any person can recognise on his own, before he acts, without first consulting anyone else’s subjective valuation; otherwise no one could know in advance whether a planned action was permitted, and one would have to interrogate the entire population before doing anything at all. Property in physical integrity has this feature — you can see where the fence runs, you can tell whether you are about to strike another’s body or take his marble — and property in “value” lacks it, because whether your act lowers another’s value depends on facts about other people’s heads that you cannot inspect. So far this is a real argument, and it is the part of Hoppe worth defending.

But run copyright through the criterion and it passes — cleanly. Can a would-be copier determine, in advance, and in complete ignorance of anyone’s subjective valuation, whether the thing he is about to reproduce is another’s authored work? Yes. He does not poll the public; he does not estimate anyone’s valuations; he checks a single, objective, in-principle-knowable fact: am I copying someone else’s work, or producing my own? The boundary of a copyright is at least as ascertainable in advance as the boundary of a parcel of land — and arguably more so, because the edge of a parcel is itself a contested normative construction (the law of nuisance, of trespass, of where one estate ends and a neighbour’s air begins, is a thicket of judgment), whereas “did I reproduce this expression or independently make my own” is a comparatively crisp question of fact. The criterion Hoppe deploys to abolish intellectual property, taken at face value, vindicates its narrow form. It rules out a right in value — correctly — and it admits a right in the bounded form, because the bounded form has exactly the advance-ascertainability that value lacks.

There is a deeper irony, which belongs to the anarchism rather than the economics. Hoppe’s ultimate ground for the physical-integrity-only view is his argumentation ethics: the claim that anyone who argues at all must presuppose self-ownership and objective property borders, so that to deny them in argument is to contradict oneself in the act of arguing. Set aside whether the argument succeeds. Notice what it is made of. “The presuppositions of rational argumentation” are not a scarce physical resource; they are an intangible, non-rivalrous, socially-instituted normative order — the very kind of entity Hoppe says cannot ground a right. The deepest foundation of the anti-IP position is built from the material the position outlaws. A theory whose floor is made of intangible norms cannot consistently treat “it is an intangible norm” as a knock-down objection to a right. And the same is true all the way up: title, contract, consent, transfer, the homestead boundary itself — the working vocabulary of Hoppe’s and Kinsella’s own system — are intangible juridical relations, not physical objects. They are accepted without protest when they protect marble and land, and forbidden only when the protected interest is an authored form. That is not a principle doing work. It is a preference deciding which intangibles are allowed.

V. The third position: the bounded form

We can now state the account both camps talk past.

Title to a made thing flows from prior ownership of the underlying resource, plus first use — not from creation, and not from labour. This keeps Hoppe’s two concessions intact. Value is not the object: conceded; the object is the form, and value is merely the reason the form is worth contesting. Labour is not the title: conceded; the labour raised the value of what was already owned, and the title runs to the owner, not to the effort. What Hoppe denies, and what does not follow, is the further claim that nothing about the produced form can be owned. That is the opposite non-sequitur to Rand’s. Rand leaps from “I created value” to “I own value”; Hoppe leaps from “value and labour are not sources of title” to “the produced form is not an object of property.” Both leaps skip the same middle term — the bounded form itself, which is neither value nor labour, but a definite, ascertainable, transferable thing.

What is the “bounded form”? It is the specific expressive or inventive configuration the producer made: this arrangement of words, this claimed mechanism, this particular confidential compilation, this source-identifying mark. It is bounded in the precise sense the law has always insisted on and the abolitionist always ignores — it excludes the idea, the method, the fact, the function, and the effort, and protects only the determinate form. Copyright protects this expression and not the idea behind it; patent protects this claimed and disclosed invention and not the principle it exploits; and both are limited in time, defeated by independent creation or prior art, and silent against anyone who learns from the work and goes on to make his own. The right is not a claim to value and not a reward for toil. It is ownership of a determinate object that the producer brought into a particular shape, on the same logic by which the sculptor owns the statue: he owned the marble, he gave it a form, and the form is now a thing he holds.

And the law has been operating on exactly this account for a very long time, against both camps’ descriptions of it. Against Rand: copyright does not reward creation-as-such or effort-as-such. In Feist the Supreme Court threw out a laboriously assembled telephone directory and held that originality, not sweat of the brow, is the constitutional touchstone, and that copyright’s purpose is to promote knowledge, not to compensate labour. The labour theory the Randian leans on is the very theory copyright disclaimed. Against Hoppe: the protected object is a bounded form whose appropriation is a recognised wrong even when nothing physical has been taken and no “value” has been seized. In Harper & Row, a magazine that obtained President Ford’s unpublished manuscript and printed his account of the Nixon pardon ahead of the authorised serialisation was held to have appropriated the right of first publication — a definite, marketable interest in the form and its exclusivity — and the fact that Ford still had every word of his manuscript was true and beside the point. The wrong was not the loss of a physical token and not the seizure of “value.” It was the taking of the bounded form’s exclusivity, which is exactly the object the third position names and both camps miss.

VI. Why each camp resists the third position

It is worth asking why neither side simply adopts the view that resolves their quarrel, because the resistance is instructive.

The Randian resists because the bounded-form view gives up the thing the Randian most wants to say: that the creative act itself is the moral source of the right, that the author deserves the work because he made it from his mind. The bounded-form view declines that. It says the author owns the work the way the sculptor owns the statue — through prior ownership and the giving of form — not through a special creative desert that the law is obliged to honour. For a tradition built on the heroism of the producing mind, this feels like a demotion, a reduction of authorship to mere configuration. But the demotion is the price of coherence: the moment “creation deserves ownership” is stated as a general principle, it either proves too much (every value you generate becomes yours, including the value competition strips from your rivals) or it collapses into the bounded-form account on inspection. Rand’s instinct — that authorship can ground exclusivity — is sound. Her theory of why is the part that does not survive.

Hoppe and Kinsella resist for the mirror-image reason. The bounded-form view concedes their two best points — value is not the object, labour is not the title — and then declines their conclusion, and the conclusion is the thing the system is built to reach. An anarcho-capitalist order of pure physical-property rights has no room for an intangible right in an authored form, because such a right binds strangers who are using their own presses and their own paper, and binding strangers who commit no physical invasion is precisely what the system exists to forbid. So the bounded form must be denied not because the denial follows from the concessions — it does not — but because admitting it would require admitting an intangible in rem right into a system whose whole architecture is the refusal of intangible in rem rights. The refusal is structural, not argued. And it is purchased, as Section IV showed, only by forgetting that the system’s own foundations — title, contract, the norms of argument — are themselves the intangibles it claims to do without.

VII. Conclusion: right route, wrong destination; right destination, wrong route

The libertarian civil war over intellectual property is not a clash between a party that is right and a party that is wrong. It is a clash between two parties that are each half-right and have each mistaken their half for the whole.

Rand reaches the right destination by the wrong route. She is correct that authorship can ground a right to exclude others from the work — that there is something there to own, and that the maker is the one who owns it. But her route, “creation is a source of rights” and “you own the value you create,” is a non-sequitur that proves too much and dissolves on contact with the fact that production is transformation and value is conferred by others. She is right about the conclusion and wrong about the argument.

Hoppe takes the right route to the wrong destination. He is correct that ownership is not conjured out of value or out of labour, and his demand that a right be ascertainable in advance is a real and valuable test. But his route does not arrive where he thinks. The ascertainability test admits the bounded form rather than excluding it; the “no third possibility” dilemma is exhaustive only because it deletes the third possibility; and the deepest ground of the view, argumentation ethics, is built from the intangible norms the view forbids. He is right about the route — that value and labour are not sources of title — and wrong about where it leads.

The bounded-form view takes the right route to the right destination. Title flows from prior ownership of the underlying thing plus first use; the object of the right is the determinate form the producer made, not its value and not the effort that made it; the right is bounded, ascertainable in advance, limited in time, and defeated by independent creation — and it is exactly the interest the law has long protected, rewarding originality of form while refusing to reward effort as such. It keeps everything true that Hoppe says and discards only his conclusion. It keeps everything true that Rand says and discards only her argument.

What neither camp will say, and what the quarrel between them obscures, is the plain thing in the middle: a made form is a thing, the person who made it from what he owned is the person who owns it, and “you cannot own value” — true, and conceded — was never an argument against that, because the form is not its value, any more than the statue is the price it would fetch. Rand mistook the value for the object. Hoppe mistook the absence of a value-right for the absence of any right. The object was the form all along.

References

The Objectivist / creation-based tradition

• Mossoff, Adam. “Saving Locke from Marx: The Labor Theory of Value in Intellectual Property Theory.” Social Philosophy and Policy 29, no. 2 (2012): 283–317.

• Mossoff, Adam. “What Is Property? Putting the Pieces Back Together.” Arizona Law Review 45 (2003): 371–443.

• Rand, Ayn. “Patents and Copyrights.” In Capitalism: The Unknown Ideal. New York: New American Library, 1966.

The anarcho-capitalist anti-IP tradition

• Hoppe, Hans-Hermann. A Theory of Socialism and Capitalism. Boston: Kluwer, 1989, esp. 139–141 (physical integrity vs value).

• Hoppe, Hans-Hermann. The Economics and Ethics of Private Property. Auburn, AL: Ludwig von Mises Institute, 2006 [1993] (the value argument; argumentation ethics).

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (2001): 1–53.

• Kinsella, N. Stephan. “Hoppe on Property Rights in Physical Integrity vs Value.” StephanKinsella.com, 12 June 2011; and “Objectivist Law Prof Mossoff on Copyright; or, the Misuse of Labor, Value, and Creation Metaphors.”

• Rothbard, Murray N. Man, Economy, and State, with Power and Market. Scholar’s ed. Auburn, AL: Ludwig von Mises Institute, 2009 [1962], ch. 2 §12 (no vested right in value); ch. 10 (defending copyright, opposing patents).

Philosophy of social institutions

• Searle, John R. The Construction of Social Reality. New York: Free Press, 1995 (institutional facts; status functions).

Doctrinal anchors

• Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991) (originality, not “sweat of the brow”; copyright is not a labour-desert regime).

• Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 (1985) (right of first publication as a bounded, marketable interest).

• Baker v. Selden, 101 U.S. 99 (1879) (idea/expression; copyright protects the form, not the method).

• 17 U.S.C. §102(b) (excludes ideas, procedures, processes, systems, methods, concepts, principles, discoveries).

Note on method and scope. Sources are cited as engaged, against their authors’ own statements; where this essay reports what a source argues — Mossoff’s labour theory, Hoppe’s value argument and argumentation ethics, Kinsella’s non-sequitur point, Rothbard’s value/physical-existence distinction, the case holdings — the report reflects the verified content of that source. Two labels are used precisely: Hoppe is identified as an anarcho-capitalist, because it is the anarchism and not Austrian economics that drives the physical-integrity-only conclusion; and the “bounded-form view” is offered as the position that survives both camps’ objections, not as either camp’s own doctrine. The essay defends the principle that an authored form can be owned; it does not defend any particular statute, term, or scope, and it concedes the genuine force of Hoppe’s two negative claims (value is not the object; labour is not the title) while denying the conclusion he draws from them.

1. The slogan and the move

There is a slogan that does a great deal of work in the argument against intellectual property: you cannot own an idea. It is repeated as though it ended a discussion, and it sounds profound because it asks you to picture property as a physical thing — a fence around a field, a lock on a door, a hand closed around an object. Once property is reduced to that picture, intellectual property looks absurd. No one loses an idea when another person copies it. No one is physically dispossessed when a pattern is repeated. No molecule is removed when a design is imitated. The conclusion seems to follow by itself: there is nothing there to own.

This is not economic analysis. It is a child’s ontology of property dressed in the language of natural law. Mature legal systems have never recognised property only where a person can clutch the thing in his fist. They recognise timed rights, divided rights, future rights, contingent rights, security interests, leaseholds, equitable interests, contractual expectancies, debts, and choses in action — rights that exist precisely because value has to be ordered across time and against other people. The anti-property argument begins by treating those institutions as curiosities at the edge of property law. They are not the edge. They are the bulk of it.

The move I want to expose is a substitution. Confronted with a question that is at bottom institutional — which costly, conflict-prone interests should the law give structured, enforceable, time-bounded form, and on what terms — the abolitionist reframes it as a question of metaphysics: is an idea the kind of thing that can be owned? The reframing feels rigorous. It is the opposite. It replaces an economic inquiry that has answers with a definitional stipulation that forecloses them. The argument is not too rigorous to permit intellectual property. It is insufficiently rigorous to reach the question that intellectual property actually raises.

A word on what this essay is and is not. It is not a defence of intellectual property as a natural right radiating from the act of creation. That theory — that one owns one’s ideas because one made them, as one owns the crops one plants — is weak, and its weaknesses are real. It is not a defence of perpetual copyright, of evergreened patents, of patent thickets, or of the proposition that every existing right is wise; those are miscalibrations, and several are indefensible. This essay attacks one specific thing: the claim that there is a clean, principled, natural-law demonstration that intangible, time-bound, value-bearing interests cannot be property. There is no such demonstration. There is a slogan, and underneath the slogan there is a physicalist reduction that cannot survive contact with a lease.

2. What the argument actually says

The most rigorous version of the anti-property case is Stephan Kinsella’s, and it deserves to be stated exactly, not caricatured, because the caricature is easy to beat and the real argument is not.

Kinsella does not, in the main, defend intellectual property as natural law and then get caught out. He does the reverse: he attacks both the creation-based natural-rights defence of IP (Rand, Spooner, Galambos) and the utilitarian defence, and he offers a positive theory of his own. That theory begins with a question worth taking seriously: why is anything property at all? His answer is scarcity. Property rights exist, he argues, because physical resources are rivalrous — my use of a thing excludes yours — so conflict over them is possible, and a system of exclusive control is needed to avoid that conflict peacefully. In his words, a property right is “simply the exclusive right to control a scarce resource”, and the function of such rights is “to prevent interpersonal conflict over scarce resources.” Allocation, on this view, runs by the Lockean first-use or homesteading rule, which gives property visible, objective borders that strangers can see and avoid.

From this premise the conclusion is meant to be inexorable. Ideas, Kinsella argues, are not scarce. Using my technique does not take it from me: “your use does not exclude my use; we could both use my technique.” From the absence of rivalry he infers that there is no economic scarcity, no possibility of conflict, and therefore no need for exclusivity. Copy my book and I still hold the original and the pattern of words alike. He invokes Jefferson’s image of lighting one taper at another without darkening the first, and he quotes Arnold Plant approvingly: property rights in patents and copyrights, Plant wrote, do not arise from the scarcity of objects but are the deliberate creation of statute that manufactures a scarcity which would not otherwise exist (Plant, 1934). Patents and copyrights, on Kinsella’s account, therefore do something sinister: by granting rights in non-scarce ideal objects, they hand the holder a measure of control over everyone else’s tangible property. If I own the pattern of a book, then you may not arrange your own paper and ink into that pattern; Kinsella calls this a “negative servitude” in the property of others, a transfer of partial ownership of material things from their natural owners to authors and inventors.

He adds two further arguments that the essay must answer rather than dodge. The first concerns contract. A defender might say: fine, forget natural rights — let the author sell each copy on the condition that the buyer not reproduce it, and reproduce intellectual property privately, by agreement. Kinsella’s reply is that contract binds only the parties to it. He runs a thought experiment: an author sells one copy without restriction and one copy “reserving” the right to reproduce; the two books are physically identical, yet on the reserved-rights theory one carries an invisible “right to copy” and the other does not. Worse, consider a stranger who never bought the book and never agreed to anything — who merely learns what is in it from gossip or a review. No contract reaches him. As Kinsella puts it, there is no warrant for the view that reserved rights can bind such third parties. A bare agreement cannot generate a right good against the world.

The second is his attack on value. The natural-rights defender, he notes, often slides into saying that something is property if it holds value. But, drawing on Hoppe, Kinsella argues that one cannot have a property right in the value of one’s property — only in its physical integrity. Others may compete with you, undersell you, and invent substitutes that destroy the market value of what you own, and none of that is aggression, because they have not touched your things. Value is a function of other people’s choices; to claim a right in it is to claim a right over their minds and actions.

This is a serious, internally disciplined argument, and three of its planks are simply correct. I concede them at the outset, because a fair argument states what its opponent gets right. Ideas, as mental contents, are non-rivalrous: your knowing what I know does not unknow it for me. You cannot own value in the abstract: competition that lowers the worth of your property is no wrong, and a right against being competed with would indeed be monstrous. A bare contract does not bind strangers: privity is real, and the reserved-rights manoeuvre cannot, by itself, reach a non-contracting third party. Grant all three. The argument still does not reach its conclusion, because the conclusion was never derived from these premises. It was derived from a fourth proposition, never defended, that does all the actual work: that the only interests capable of being property are rivalrous physical objects. That proposition is not a finding. It is the whole question, asserted as an answer.

3. Natural law is not physicalism

Begin with the philosophical sleight, because it sets up everything that follows. The argument advertises itself as natural law. Natural law, in any of its serious historical forms, is a theory about human action, reason, and justice — about promise-keeping, honest dealing, desert, reliance, exchange, restitution, the wrong of fraud, and the ordering of liberty among persons. It is emphatically not the doctrine that only objects one can physically hold can ground a right. A tradition that runs from the Roman law of obligations through the common law of contract and equity has always protected things you cannot touch: the enforceable promise, the debt, the expectancy, the relationship of confidence.

So when the abolitionist says “natural law” but means “only tangible rivalry counts,” he has not deepened natural law; he has hollowed it out and kept the name. A natural law that can recognise trespass on land but cannot recognise the betrayal of a confidence, the deliberate passing-off of one trader’s goods as another’s, the fraudulent misappropriation of what someone has built, or the calculated destruction of a contractual expectancy is not a more rigorous natural law. It is merely a thinner one. Thinness is being sold as rigour. The physicalist has not earned his austerity by argument; he has simply declared that the hard cases are not cases.

And the declaration is self-undermining in a way worth naming early. Kinsella’s own foundation is not a brute physical fact but a rule: the first-use homesteading principle, a chosen convention for converting unowned scarce things into owned ones, justified because it reduces conflict and gives visible borders. That is an institution — a human-made rule for assigning exclusive control to achieve a social end. Once you have conceded that property in land is constituted by a conflict-reducing rule rather than discovered in nature, you have conceded the only thing that matters. The question is no longer natural versus artificial; all property is rule-constituted, his homestead included. The question is which rule, for which interests, on what terms. That is an economic and institutional question, and it is exactly the question the scarcity slogan is designed to avoid reaching.

4. The physical-exclusion fallacy: scarcity is not the same as touch

Here is the economic heart of the matter, and the first and most important fallacy.

The argument treats physical exclusion as if it were the whole of scarcity. It is not. Scarcity, in economics, is not the question can two people touch the same object at once? It is the question are there valuable means that cannot simultaneously serve all the competing ends to which people would put them? On that definition — the real one — a great many things attached to an idea are scarce even though the idea itself, as information, is not. Time is scarce. Priority is scarce. Attention is scarce. A market window is scarce. The position of first mover is scarce. Reputation and provenance are scarce. The condition of secrecy, once broken, cannot be restored. The option to be the one who decides how a thing is first used has a present value and can be held by only one party at a time.

Run the distinction through cases. Two firms can use the same chemical formula; they cannot both be the first to bring it to market, and the first-mover position is worth money precisely because it is exclusive. Two authors can possess the identical text; they cannot both be the authenticated origin of it, and authorship and provenance command a premium for exactly that reason. Two manufacturers can run the same process; they cannot both hold the single exclusive supply position in a given market at a given time. The information is non-rivalrous and can be copied without limit. The commercial position attached to it is rivalrous, depletable, and conflict-prone. The slogan proves the first and then behaves as though it had proved the second. It has not. It has performed a category substitution: it establishes that information is not consumed by use and then helps itself to the very different proposition that no economically scarce interest exists around information. That is not a proof. It is a switch of subjects in the middle of the sentence.

Notice that this defeats the argument on its own chosen test. Kinsella’s criterion is not, in the end, physical depletion; it is the possibility of conflict — he says it is the possibility of conflict over a resource that renders it scarce and calls property into being. Very well. Is conflict possible over the first-mover position, the exclusive licence, the undisclosed formula, the market window? Not merely possible — routine. Firms litigate, spy, race, and bargain over exactly these positions, because two parties genuinely cannot both occupy them. By Kinsella’s own definition of scarcity as the possibility of conflict, these economic positions are scarce, and therefore are candidates — on his terms, not mine — for the kind of conflict-avoiding rule he says property is. The scarcity test, taken seriously and applied consistently, does not exclude the interests around information. It includes them. The only way to keep them out is to redefine scarcity, at the last moment, as physical rivalry — which is the thing that was supposed to be shown, not assumed.

There is a further category the slogan never registers, and it is the one that matters most in a networked economy: coordination scarcity. A standard, a protocol, a ledger, a trademark, a reference implementation — these are valuable not because the underlying information is scarce but because they coordinate the expectations of strangers, and a coordinating position is intensely rivalrous. Only one arrangement can be the standard that everyone builds to; two incompatible specifications cannot both occupy that slot, which is why standard-setting is fought over so bitterly. A trademark is the cleanest legal case: its entire function is to let a buyer rely on the source of goods, and two producers cannot both authentically signal the same origin — if they could, the signal would carry no information and the coordination would collapse. The law that protects the mark is not protecting a scarce symbol, since symbols are infinitely copyable; it is protecting a scarce coordinating position, the reliable link between a name and a source. Authentication is the same phenomenon from the buyer’s side: provenance, attribution, and a verified chain of custody are rivalrous because there can be only one true origin of a given thing, and the worth of knowing it depends on others being unable to counterfeit the claim. None of this is scarce in the sense of a depletable object. All of it is scarce in the sense that decides whether conflict is possible — and all of it is conflict-prone, which is the only sense Kinsella’s own theory says matters.

This is why Plant, whom the abolitionist quotes, is a treacherous ally for the abolitionist. Plant was a copyright sceptic; he did argue that the law manufactures a scarcity in patents and copyrights that the objects do not have (Plant, 1934). But Plant also documented, carefully, how authors were paid in the era before international copyright protected them: nineteenth-century American publishers, free to reprint any English book without paying anyone, nonetheless paid English authors handsomely for early “advance sheets,” because being first to the American market was worth paying for (Plant, 1934). That is the point in miniature. The information was free to copy; the priority was not, because priority is scarce. Plant’s own evidence is a demonstration that a commercial position can be valuable and appropriable even when the underlying information is not — which is precisely what the non-scarcity slogan denies. Arrow made the structural version of the same observation in the founding paper of the economics of invention: information can be reproduced at little or no cost, which is exactly why a competitive market tends to underinvest in producing it (Arrow, 1962). Non-rivalry is not the abolitionist’s discovery that ends the inquiry. It is the orthodox premise that opens it.

5. The permanence fallacy: leaseholds, easements, and the death of the tangibility test

The second fallacy is the assumption that real property must look like fee-simple ownership of a physical object: tangible, perpetual, and physically possessed. Strip that assumption out and the argument loses its metaphysical simplicity, because actual property law has never honoured it.

Take the entry point the abolitionist himself cannot avoid: the lease. A leasehold, in black-letter law, is a property interest. The Legal Information Institute states it plainly: a lease between landlord and tenant “creates both a contractual interest and a property interest” — the latter called a leasehold (Cornell LII, Leasehold). It is time-bound — a term of years, not forever. It is alienable and, in many systems, mortgageable. During its term it gives the tenant a present right to exclude even the owner; afterward it evaporates. A thirty-year lease is, on any honest accounting, property: valuable, transferable, defended by law. And it is not perpetual. So temporality is not fatal to property. That single, banal institution refutes the permanence requirement outright.

Press on through the catalogue, and the physicalist criteria fall one by one. An easement is a right to use another’s land without owning or possessing it — non-possessory by definition, and it runs to subsequent owners. So possession is not fatal either: the law recognises a property interest that gives no possession at all. A life estate lasts a lifetime and then ends — temporal again, and Kinsella himself, drawing on the standard bundle-of-rights conception, gives the example of a landowner carving out a life estate for his mother and an easement for a neighbour while selling the minerals to an oil company. He accepts, in his own text, that one tangible parcel can be sliced into divided, timed, non-possessory interests, each a legitimate piece of property. That concession is fatal to half his rhetoric, because it means the dispute was never really about permanence or possession. It was only ever about scarcity — which Section 4 has already met on its own terms.

Now go further than land, to the interest that does the most damage to the physicalist picture: the chose in action. A chose in action is a property right whose entire content is an enforceable claim — a debt, a contractual right, an insurance claim, a right to sue. It is intangible: there is no object. It is non-possessory: you cannot hold a debt in your hand. It is enforceable only through the legal system. And it is, unquestionably, property — bought, sold, assigned, securitised, inherited. The Uniform Commercial Code classifies a “general intangible” as “any personal property, including things in action” (U.C.C. § 9-102), and the Legal Information Institute defines intangible property — listing stocks, bonds, trademarks, copyrights, and trade secrets — as property that “cannot be physically held” and yet “is still a form of property” (Cornell LII, Intangible property). The law’s own definition of property is not “a tangible thing.” It is “anything (items or attributes / tangible or intangible) that can be owned”, and it expressly names “the goodwill of a company” as intangible property (Cornell LII, Property). A debt is a pure value-claim, owned and traded as property, with no physical integrity whatsoever — and no one calls it a metaphysical fraud or a monopoly.

The figure sets the whole landscape out at once.

Read down the rows. Fee simple in land — the physicalist’s archetype — is the only interest that is tangible, perpetual, possessory, and good against the world all at once. Every other interest the law recognises fails at least one of those tests, and most fail two or three. Leaseholds and life estates are not perpetual. Easements and covenants are not possessory and attach to no object. Options, licences, debts, and goodwill are intangible, time-bound, and non-possessory. And then patents and copyrights sit at the bottom, failing exactly the criteria that easements, covenants, options, and debts already fail — and binding third parties exactly as easements, covenants, and perfected security interests already do. Intellectual property is not an anomaly that a physicalist test cleanly excludes while leaving the rest of property law intact. It is unremarkable. The test that would void it would void the lease, the mortgage, the easement, the covenant, the option, the share, and the debt along with it. A theory of property that cannot survive contact with a thirty-year lease is not a theory of property. It is an aesthetic preference for objects, presented as ontology.

6. The privity trick: contract is not the whole of private law

Kinsella’s contract argument is his strongest, and it is half right, which is what makes it dangerous. He is correct that a bare contract binds only its parties, and that the “reserved rights” manoeuvre cannot, by itself, reach a stranger who never agreed — least of all the third party who merely overhears the information. If that were the entire toolkit of private law, the abolitionist would have a point. But it is not, and the inference from contract alone cannot bind the world to no non-possessory, intangible interest can bind the world is a leap across a canyon.

The law has, for centuries, known how to make certain obligations run beyond the original pair — not by magic, and not by pretending a contract reaches strangers, but through structured, notice-based property doctrines. The clearest is the covenant that runs with the land. A real covenant is a promise about the use of land that binds not only the promisor but future owners who were never parties to it. The Legal Information Institute is explicit that such covenants run with the land, so that “future owners of the property are bound by the covenant” (Cornell LII, Real covenant), and that they “differ from personal covenants, which bind only the original parties” and do not transfer to successors (Cornell LII, Covenant that runs with the land). This is not done casually. The burden runs only when a set of conditions is satisfied: the parties intended it to run, the covenant touches and concerns the land, the requisite privity exists, and — decisively — the successor took with notice of it (Cornell LII, Vertical privity). A later owner is bound because the restriction was intended to run, related to the land, and was knowable to him when he acquired it.

Sit with what that establishes. The law already possesses a mature, disciplined mechanism for making a non-possessory restriction binding on a person who never signed anything — and it treats this as ordinary property doctrine, not as a metaphysical fraud. The same is true across the field: a perfected security interest binds third-party creditors and purchasers, not because they agreed but because the interest was filed and made knowable; an easement binds successors who took with notice; an in-rem right is precisely a right held against the world subject to public notice. The recurring structure is notice plus a recognised form. The law makes a claim opposable to strangers when it is rendered visible and falls within an established category. This is exactly the “objective, visible borders” condition that Kinsella himself says good property rights must have — borders a stranger can see and avoid. Registration, publication, and the patent and copyright records are the intangible analogue of the fence: the mechanism by which a non-possessory claim is made legible to the world.

So the honest statement is narrow but devastating to the rhetoric. Kinsella is right that contract alone cannot bind strangers; he is wrong to convert that truism into a theory of property. Property law is the institutional technology by which some claims — not all — are made legible, transferable, and opposable beyond the contracting pair, under notice rules, for limited durations, with defences and exceptions. The real question is never the false binary the slogan offers — bare contract or universal metaphysical ownership. Real legal systems live in the enormous space between: they build forms, notice rules, priority rules, terms, limits, and remedies. To say “contract does not bind the world, therefore intellectual property is illegitimate” is to pretend that the entire law of property — which exists precisely to make selected claims run against the world — does not exist.

7. The no-value-in-property fallacy

The fourth plank is the Hoppean claim that one may own the physical integrity of one’s property but not its value. As I said at the outset, the core of this is correct and I concede it without reservation: there is no right against being competed with, no right against a rival’s better product or lower price, no right in the mere market worth of what you hold. A right of that kind would be a licence to suppress competition, and it would be indefensible.

But the concession does not reach the conclusion, and the chose in action shows why with unusual clarity. A debt is not a right against “value in the abstract.” It is a specific, defined, legally structured claim — this obligor owes this sum under this obligation — and its entire content is a claim to value, enforceable only through law, with no physical integrity at all. The law recognises it as property, lets you sell it, lets you use it as collateral, and protects it against third parties who knowingly interfere with it. If “you cannot own value, only physical integrity” were a true and general principle of property, the entire law of obligations-as-property would be impossible. It is not impossible; it is enormous, ancient, and uncontroversial. The principle, stated as a universal, is simply false as a description of the legal order it claims to describe.

What is true is narrower and more useful: you cannot own undifferentiated value, the kind that fluctuates with every rival’s choices. What you can own — what the law constructs and enforces every day — is a defined, bounded, structured interest whose value is the reason for defining it. That is what a lease, an option, a debt, a covenant, a security interest, and a patent all are: structured legal expectations about future use, exclusion, transfer, yield, priority, and remedy. Demsetz, Coase, and the whole property-rights tradition put the point bluntly — what is exchanged in any transaction is a bundle of rights, and it is the value of the rights that determines the value of the thing. Value is not an ornament hung on property after the fact. Value is why the right is specified at all. The physicalist treats value as a contaminant that disqualifies an interest from being property. The law treats value as the entire motivation for recognising the interest in the first place. One of these is a description of the legal world. The other is a stipulation about which the legal world has never been consulted.

8. “Artificial scarcity” is not the objection

The most rhetorically effective move remaining is the charge that intellectual property creates artificial scarcity — that it manufactures, by statute, an exclusivity the objects do not naturally have. This is true and it is supposed to sound damning. It is not damning, for a reason the abolitionist’s own framework concedes.

The law creates “artificial scarcity” constantly, and we do not flinch. A lease creates a temporary, exclusive, legally enforced scarcity of access to a building. A security interest creates a priority — an artificial ranking of one creditor’s claim above another’s. A restrictive covenant manufactures a legally enforced scarcity of permitted uses on a neighbour’s land. A corporate share is a bounded, artificial claim on a residual. A court’s injunction creates an enforceable restraint that did not exist in nature. Spectrum licences, water rights, and franchise territories are all deliberate, statutory creations of exclusive position. None of these arises from the physical scarcity of an object; all of them are institutional artifacts. If “artificial scarcity” were a sufficient objection, it would condemn the lease, the mortgage, the covenant, and the corporate share with the same stroke — and it would condemn Kinsella’s own homesteading rule, which is itself an artificial convention for converting non-owned resources into exclusively controlled ones.

So “artificial” cannot be the objection, because every property right is artificial in the only sense that matters: it is constituted by a rule rather than found in nature. The honest question is not whether a legally created scarcity is artificial. It is whether it is justified — whether the institution reduces conflict, supports the investment and disclosure that would otherwise not be financed, lowers transaction costs, protects reliance, disciplines opportunism, and orders production better than the available alternatives. That is an institutional and economic question, and it is the one Coase put at the centre of the analysis of rights. Coase showed that there can be no market transactions to allocate resources until rights have first been delimited — entitlements that have not been defined cannot be traded — and that once transaction costs are positive, which in the real world they always are, the assignment of those rights determines how resources are used and who captures the gains (Coase, 1960). The choice is never between an “artificial” right and a state of natural freedom. It is between one assignment of rights and another, each with its own consequences for conflict, investment, and value. “It’s artificial” is not an analysis. It is a refusal to do the analysis.

9. The sophistic structure

Step back from the individual planks and look at the shape of the argument, because the shape is the tell. It advances by a series of substitutions, each of which trades a true and narrow claim for a false and broad one while keeping the same words:

• “Ideas are not depleted by use” becomes “no economic interest around an idea is harmed by copying.” The first is true; the second ignores priority, market position, disclosure, reputation, and option value.

• “A bare contract does not bind strangers” becomes “no non-possessory intangible right can bind strangers.” The first is true; the second ignores covenants, easements, security interests, and the whole in-rem architecture of property.

• “You cannot own value in the abstract” becomes “no value-bearing claim can be property.” The first is true; the second ignores every debt, option, and chose in action.

• “Intellectual property is often overbroad” becomes “all intellectual property is illegitimate.” The first is true and urgent; the second does not follow from it.

• “State grants can be abused” becomes “legal form is itself coercion.” The first is true; the second would dissolve all law.

• “Some exclusivity is monopoly” becomes “all exclusivity is monopoly.” The first is true; the second is a definitional trick.

• “Scarcity calls property into being” becomes “only physical rivalry counts as scarcity.” The first is Kinsella’s own premise; the second is the unstated conclusion he needs and never earns.

This is the anatomy of the thing. It is not rigorous; it is tidy. It achieves its cleanliness by discarding, at each step, the facts that make the problem hard — the time-dimension of value, the in-rem machinery of property, the law of obligations, the difference between a bad instance and an illegitimate institution. A tidy argument that has thrown away the difficulties is more dangerous than a messy one that has kept them, because it persuades by the comfort of its symmetry. The proper response is not contempt for the conclusion but precision about the mechanism: at each substitution, a true premise is quietly swapped for a stronger one that the premise does not support. Pull the substitutions apart and there is no argument left — only a definition of property narrow enough to guarantee the result, asserted before the inquiry begins.

10. A better question, and the calculation the slogan never reaches

Critique is not enough; a serious position has to say what it would put in place. Here is the framework the scarcity-stop refuses to reach.

A rigorous theory of property does not ask the metaphysical question is the object physically rivalrous? and stop there. It asks a sequence of institutional questions. What is the scarce economic interest actually in play — the position, the priority, the disclosure, the secret, the expectation? Who created, acquired, financed, or relied upon it? Would exclusion be required to protect it, or would a weaker liability rule suffice? For how long is protection justified before the incentive saturates and the access cost dominates? What notice must be given so that strangers can see and avoid the right? What are the transaction costs of recognising it, and the deadweight costs of the exclusivity it confers? What exceptions and limits keep it from swallowing adjacent liberties? What remedy is proportionate? And, finally: does recognising a structured right of this shape order production, reliance, and exchange better than refusing to recognise it?

That sequence can be compressed into a single decision rule. Let the relevant quantities be defined as follows:

ΔV_create   = additional creation, investment, and disclosure the rule induces
R_coord     = reliance and coordination gains (standards, provenance, secure exchange)
ΔF_reduced  = appropriation / free-riding losses the rule prevents
E           = enforcement and administration cost of the right
M           = monopoly / deadweight cost of the exclusivity conferred
N           = notice and information cost imposed on third parties

The institution is justified when the gains it produces exceed the costs it imposes:

Recognise a structured right where:

    ( ΔV_create  +  R_coord  +  ΔF_reduced )   >   ( E  +  M  +  N )

This is not a formula that delivers a single global verdict on “intellectual property,” and that is its virtue. Each term varies by domain. Where the fixed cost of creation is enormous and imitation after disclosure is near-free — a new medicine, say — ΔV_create and ΔF_reduced are large and the inequality can hold comfortably for a bounded term. Where the fixed cost is small and the protected object is an input into a great deal of downstream work, M and the follow-on burden can swamp the gains and the inequality fails; the rule should then be narrower, shorter, or absent. The framework is the same; the answer is local. It also explains, without any appeal to metaphysics, why the limited term of a patent is a feature rather than a refutation: the right is bounded because the incentive effect saturates while the access and follow-on costs do not, so welfare peaks at an interior term and the right is written to expire — which is why the international architecture fixes that term at twenty years from filing (TRIPS, Art. 33) and why copyright protects only expression and never the idea, leaving the teaching free (17 U.S.C. § 102(b)). The bound is the design working, not the design failing (Landes & Posner, 1989). Teece’s work on appropriability makes the institutional stakes concrete from the other side: when an innovation cannot be protected, the returns do not vanish into a commons — they flow to whoever controls the complementary assets, the manufacturing and the distribution, rather than to whoever bore the cost of creating (Teece, 1986). Non-recognition is not neutrality. It is a different allocation, with its own winners.

The framework also exposes a design choice the slogan cannot even state: the choice between an exclusion rule and a liability rule. To protect an interest, a legal system can grant the holder a right to exclude — an injunction, the power to refuse — or merely a right to be paid, a claim in damages when another uses it. The two are not interchangeable. Exclusion fits interests where bargaining before use is cheap and the holder’s valuation is hard for a court to reconstruct, so the parties should be made to strike their own price. A liability rule fits interests where pre-use bargaining is impractical — users too many, too dispersed, or too urgent — and a court can approximate the price after the fact. Much of what presents itself as a binary war over whether intellectual property should “exist” is, on inspection, an unstated argument about which rule fits which interest: compulsory licensing, statutory damages, fair-use carve-outs, and standard-essential-patent commitments are all liability-rule machinery bolted onto a property-rule frame, precisely to fix the cases where pure exclusion would over-block. The scarcity-only test cannot reach this question, because it has terminated the inquiry at the threshold; it never asks how an interest should be protected, having wrongly concluded that it cannot be protected at all. A theory that cannot distinguish an injunction from a damages award is not a theory of remedies. It is the absence of one.

The decisive point against the scarcity-only test is that it never performs this calculation at all. It halts at the threshold, asks whether the object is physically rivalrous, answers no, and declares the inquiry closed. But that is not economic analysis; it is an exclusionary definition wearing the costume of economics. The whole substance of the problem — the costs, the benefits, the duration, the notice, the externalities, the comparison to alternatives — lies on the far side of a door the slogan refuses to open. To call that rigour is to mistake the act of stopping early for the act of being careful.

11. Conclusion

Kinsella’s argument does not defeat intangible property. It defeats a caricature of intangible property — the perpetual, free-floating ownership of ideas as such — that no competent legal system asserts and that the statutes, treaties, and cases expressly disclaim. Against the real thing, the argument fails, and it fails technically, at identifiable joints. It substitutes physical rivalry for economic scarcity, and so never sees that priority, exclusivity, disclosure, and market position are scarce on its own definition of scarcity as the possibility of conflict. It demands that property be tangible, perpetual, and possessory, and so would void the lease, the easement, the covenant, the option, the share, and the debt — the bulk of property law — if it were applied with a straight face. It treats privity as the whole of private law, and so ignores the in-rem, notice-based machinery by which the law has always made selected claims run against the world. It denies that value can be owned, and so cannot account for the chose in action, the purest property right the law contains. And it brands legal scarcity “artificial” as though that were an objection, when its own homesteading rule is artificial in exactly the same sense.

What survives, once the slogan is pulled apart, is not the abolitionist’s conclusion but the abolitionist’s true premise, restored to its proper modesty: scarcity and the possibility of conflict are indeed why property exists. Quite so. And the economic interests that gather around costly creation, disclosure, secrecy, priority, reputation, and exchange are scarce, conflict-prone, costly to produce, and valuable to hold. They are, on the abolitionist’s own test, exactly the kind of interest for which property, contract, and equity exist. Whether and how far the law should give any particular one of them structured form is a hard institutional question with answers that differ by domain — answers that require the calculation, not the slogan. The non-scarcity argument is an elaborate device for never having to do that work. It is tidy. It is clean. It is too clean to survive contact with a lease, and the cleanliness was always purchased by throwing away the difficulty that is the entire point.

References

Arrow, K. J. (1962). Economic welfare and the allocation of resources for invention. In R. R. Nelson (Ed.), The Rate and Direction of Inventive Activity: Economic and Social Factors (pp. 609–626). Princeton University Press / NBER.

Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1–44. https://www.jstor.org/stable/724810

Cornell Law School, Legal Information Institute. Covenant that runs with the land. Wex. https://www.law.cornell.edu/wex/covenant_that_runs_with_the_land

Cornell Law School, Legal Information Institute. Intangible property. Wex. https://www.law.cornell.edu/wex/intangible_property

Cornell Law School, Legal Information Institute. Leasehold. Wex. https://www.law.cornell.edu/wex/leasehold

Cornell Law School, Legal Information Institute. Property. Wex. https://www.law.cornell.edu/wex/property

Cornell Law School, Legal Information Institute. Real covenant. Wex. https://www.law.cornell.edu/wex/real_covenant

Cornell Law School, Legal Information Institute. Vertical privity. Wex. https://www.law.cornell.edu/wex/vertical_privity

Kinsella, N. S. (2008). Against Intellectual Property. Ludwig von Mises Institute. (Originally published as Kinsella, N. S. (2001), Against intellectual property, Journal of Libertarian Studies, 15(2), 1–53; argument restated in Kinsella, N. S. (2013), The case against intellectual property, in C. Luetge (Ed.), Handbook of the Philosophical Foundations of Business Ethics, Springer.)

Landes, W. M., & Posner, R. A. (1989). An economic analysis of copyright law. The Journal of Legal Studies, 18(2), 325–363. https://doi.org/10.1086/468150

Plant, A. (1934). The economic aspects of copyright in books. Economica, 1(2), 167–195. https://doi.org/10.2307/2548748

Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6), 285–305. https://doi.org/10.1016/0048-7333(86)90027-2

Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement), Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, Art. 33 (term of protection).

Uniform Commercial Code § 9-102(a)(42) (definition of “general intangible,” including things in action).

United States Copyright Act, 17 U.S.C. § 102(b) (idea/expression distinction).

Abstract. This essay examines the case for abolishing intellectual property — stated most rigorously in libertarian form by Stephan Kinsella and in economic form by Michele Boldrin and David Levine — and argues that, whatever its force against the overgrown rights we actually have, it fails as a general program for a reason its proponents rarely confront. The standard anti-IP arguments are not economic refutations; they are category errors. That a right which expires cannot be property; that copying takes nothing because the original remains; that enforcement is violence; that markets will reward creators unaided; that pre-modern genius proves modern rights unnecessary; that open source disproves copyright. Each survives only by beginning the analysis after the work exists, and by treating the non-rivalry of the finished idea as decisive while ignoring the costly, failure-ridden process that produced it. The deeper objection is political-economic, and it is the one the abolitionist least expects. Where a creator cannot hold an enforceable right, the surplus from creation does not disperse to the public; it flows to whoever controls the complementary assets — distribution, capital, manufacturing scale, platforms, legal capacity — exactly as the appropriability literature documents. Abolition does not dissolve monopoly power. It relocates it, from the party who bore the cost to the largest incumbent. The defensible position is neither maximalism nor abolition but calibration: intellectual property is an institutional instrument for financing fixed-cost creation where imitation is cheap, justified only to the extent that limited, enforceable, competition-aware rights raise dynamic welfare net of deadweight loss, enforcement cost, and harm to follow-on innovation. The argument is built on full-text evidence and refuses any appeal to natural law — which is the trap the abolitionist sets and the maximalist walks into.

1. The move

There is a characteristic move in the argument against intellectual property, and once you have seen it you cannot unsee it. Confronted with a question that is fundamentally about institutions — what set of rules produces the most creation, disclosure, investment, competition, and access over time — the abolitionist reframes it as a question about metaphysics. Is an idea scarce? Does copying remove anything from the original mind? Is a right that expires really a right at all? Is the enforcement of any rule simply violence wearing a robe? The reframing is elegant, it feels rigorous, and it is almost entirely beside the point.

It is beside the point because the relevant test is not whether a right “feels natural.” It is whether the legal rule, in a given industry, under given conditions of cost and imitation, raises total welfare across time. Write that object down honestly and it has six terms, not one:

W  =  CS  +  PS  +  B_F  −  DWL  −  E  −  C_F

where CS is consumer surplus from access, PS is the return to the creator or producer, B_F is the future benefit from creation and disclosure that the rule induces, DWL is the deadweight loss from temporary exclusivity, E is enforcement and transaction cost, and C_F is the cost the rule imposes on follow-on innovation. The case for any particular intellectual-property rule is that it raises W. The case against it is that the access and follow-on gains from weakening it exceed the creation it would forgo. Both are empirical claims. Neither is settled by an axiom about the nature of ideas.

The abolitionist wins only if abolition improves this function. The striking thing about the literature is how rarely it even attempts the calculation. It establishes, correctly, that information is non-rivalrous, and then treats that single fact as decisive. But non-rivalry is not the abolitionist’s discovery; it is the orthodox starting premise. Kenneth Arrow built the modern economics of invention on it in 1962, observing that information, once produced, can be reproduced “at little or no cost,” that legal protection can make it only partially appropriable, and — crucially — that for exactly this reason a competitive market will tend to underinvest in producing it. The non-rivalry of ideas is the premise of the problem, not the solution to it. It tells you why access is valuable. It tells you nothing about how the costly, uncertain process that generates the idea gets financed when anyone may copy the result.

That is the gap this essay works in. I will take the strongest anti-IP arguments in turn — the ones Kinsella and others actually deploy — and show that each is a category error that depends on starting the story after the fixed cost has been sunk. Then I will make the argument the abolitionist least expects: that the world without enforceable creator rights is not a commune of equal creators but a feeding ground for scale. And I will be explicit throughout about what I am not claiming. I am not defending perpetual copyright, evergreened drug patents, patent thickets, or the proposition that every right we have is justified. The defensible position is that intellectual property is a calibrated instrument, and most real disputes are about how badly it has been miscalibrated — not about whether to throw the instrument away.

This is not a theology of intellectual property. It is economics. And economics, unlike theology, gives different answers in different places.

2. What is actually protected — and the toy world that hides it

The most-repeated anti-IP slogan is that “ideas cannot be owned.” It is repeated because it is unanswerable, and it is unanswerable because no serious legal system claims otherwise. It is a refutation of a position nobody defends.

Copyright does not protect ideas. United States law could not be more explicit. Section 102(b) of the Copyright Act states that “in no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work” (17 U.S.C. § 102(b)). What copyright reaches is original expression fixed in a tangible medium — the sentences, the arrangement, the developed characters — and nothing underneath them. The idea of star-crossed lovers from feuding families is free. The plot device, the genre, the historical fact, the method described in the text: all free. Anyone may write a competing book, review it, parody it, quote it, build on its ideas, and use every fact it contains. What they may not do is reproduce the author’s expression and sell it as their own.

Patents do not protect idle thoughts either. A patent reaches a claimed invention that satisfies the legal standards — novelty, an inventive step, industrial applicability — and only in exchange for disclosure: the applicant must describe the invention clearly enough that a skilled person can carry it out, which is why the teaching enters the public record and, at term’s end, the public domain. The thing protected is not “an idea.” It is a specific, examined, disclosed technical solution that someone bore the cost of developing and proving.

This distinction matters because the abolitionist’s favourite illustrations quietly violate it. Consider the island-net example, deployed to make patent law look self-evidently absurd: I weave a fishing net, you watch and weave your own, and the patent system supposedly entitles me to beat you and take your fish. The example is rhetorically effective and analytically worthless, for four reasons. First, it removes the entire institutional setting — no courts, no filing, no examination, no claims, no disclosure requirement, no commercial scale, no market — and then expresses astonishment that legal categories do not apply to a world from which law has been deleted. Removing the institution and asking why the institution matters is not an argument; it is the absence of one. Second, a visible rope net woven in an afternoon is the antithesis of the patentable case: obvious, costless, undisclosed. The real analogue is a costly, non-obvious, disclosed technical advance whose developer bore years of risk and whose imitator arrives only after the uncertainty has been resolved at someone else’s expense. Third, the example smuggles in assault. Patent enforcement is not “I beat you and take your fish.” It is injunctions, damages, licensing, validity challenges, claim construction, and judicial process — the same machinery that enforces contracts and debts. Fourth, all of that machinery is exactly as “coercive” as the law of trespass or fraud, which brings us to a point worth stating in its own right.

The net analogy fails for a deeper reason still: it silently collapses three situations that law and economics keep separate, and that must be kept separate for the argument to make any sense. The first is independent creation — you invent the same net with no knowledge of mine. That is not the problem; it is competition, and no defensible regime punishes it. The second is copying the disclosed result after someone else bore the fixed cost — you wait until I have proven the design works, then reproduce it without having borne the discovery, the risk, or the failures. That is the problem intellectual property exists to address, and it is the only one of the three the abolitionist actually needs to defeat. The third is theft or misuse of confidential material — you take my unpublished plans, my unreleased manuscript, my undisclosed formula — which is a different wrong again, governed by trade-secret and breach-of-confidence law rather than by patent or copyright. The anti-IP argument advances precisely by sliding between these: it borrows the innocence of the first to excuse the second, and treats the third as if it never happens. Hold them apart and the net loses its charm. The hard case is never the man who independently wove his own net; it is the man who waited for mine to fill with fish and then copied it exactly.

It is also worth noticing, before leaving this section, that the case for patents is not uniform across the economy — and the honest defender should say so first. When Cohen, Nelson and Walsh surveyed 1,478 United States manufacturing R&D labs, they found that firms in most industries rank patents the least important of their appropriation mechanisms, leaning instead on secrecy, lead time, and complementary capabilities. Patents are decisive, they found, in only a small set of industries, “most notably pharmaceuticals.” That fact will return as the spine of the calibrated view. For now it does one job: it shows that anyone who reasons about intellectual property from a single toy example — a net, a song, a drug — has already gone wrong, because the economics differs by sector.

3. Expiry, enforcement, and the natural-law trap

Two of the most popular arguments are really one mistake stated twice: the belief that intellectual property must be defended, or refuted, as a metaphysical natural right.

Take the expiry argument first. If a patent lapses after twenty years, it cannot be a real property right, because real rights do not expire. This assumes its own conclusion — that the only genuine property interest is a perpetual one — and the assumption is simply false as a description of how legal rights work. A leasehold is not a counterfeit because it ends. A licence is not fraudulent because it has a term. Options expire, easements are limited, security interests terminate, usufructs are conditional, and corporate charters are revocable. The legal world is full of rights that are time-bound, defeasible, conditional, and contingent, and we do not conclude from their limits that they were never rights at all.

More importantly, the patent is badly defended as a mystical perpetual dominion and well defended as something more modest and more powerful: a disclosure-and-investment institution. The inventor teaches the public how the thing works; in return the inventor receives a limited period of exclusivity; then the teaching belongs to everyone. The twenty-year term is not arbitrary theatre. It is written into the international architecture: Article 33 of the TRIPS Agreement provides that “the term of protection available shall not end before the expiration of a period of twenty years counted from the filing date” (TRIPS, Art. 33). The right is bounded by design, because the design is aimed at dynamic welfare, not at dominion.

Once you see the patent as a calibrated instrument, the abolitionist’s question — “why not forever?” — answers itself, and so does the maximalist’s silent assumption. Why not forever? Because perpetual protection generally overprotects: it piles deadweight loss and follow-on obstruction on top of an incentive effect that saturated long ago. Why not zero? Because zero can underproduce, where the fixed cost is large and imitation cheap. The right term is the one that maximises welfare:

T*  =  argmax_T [ B_F(T) + CS(T) − DWL(T) − C_F(T) − E(T) ]

This is Landes and Posner’s (1989) framing of copyright, turned into a general point: protection is an optimisation problem balancing the incentive to create against the cost of restricted access and constrained follow-on use. The dynamic benefit B_F rises with protection but saturates; the access and follow-on costs rise without saturating; net welfare therefore peaks at an interior T* and declines on either side (Figure 3, later). The fact that a patent expires is not evidence against patents. It is evidence that patents are calibrated instruments rather than absolute claims — which is exactly the point.

The enforcement argument collapses the same way. Isn’t intellectual-property enforcement just violence? If “violence” means the ultimate coercive backing of any legal rule, then yes — and so is the enforcement of a lease, a loan, a fraud judgment, or the boundary of a field. The objection, taken seriously, is not to patents or copyright; it is to law as such. If you are willing to enforce a contract or a property line, you cannot consistently treat the enforcement of a creator’s right as a unique act of aggression. And if you are not willing to enforce any rule, you have not produced a theory of markets; you have produced anarchy, in which, as the next sections show, the strong appropriate from the weak.

The general lesson is to refuse the natural-law framing from both directions. Labour, authorship, and invention may be morally relevant, but the scope, term, exceptions, and remedies of intellectual property are institutional questions answerable only by their consequences. The abolitionist sets the trap by demanding that you justify IP as eternal dominion and then pointing out that it expires. The maximalist walks into the trap by accepting the framing. The escape is to stop treating the question as metaphysics and start treating it as design.

4. “Copying takes nothing”: the trick of starting after the fixed cost

The emotional core of the anti-IP case is the observation that copying is not theft because, when you copy my work, I still have it. Nothing has been removed. The barn is still full.

This is true and it is a sleight of hand. Physical dispossession is not the only way to destroy an economic interest, and it is not the interest that creation depends on. If someone duplicates a manuscript, clones a dataset, lifts a drug formulation, or reproduces a textbook, the creator does still “have” the artifact — and may simultaneously lose exclusivity, market return, sequencing advantage, bargaining power, and any prospect of recovering what the work cost to make. The harm is appropriation of value, not disappearance of an object.

The model that makes this precise is the one the abolitionist never writes down. Let a work earn expected revenue R at marginal production cost C_m, against a fixed cost of creation F:

Π  =  R  −  C_m  −  F

The work gets made only if it can clear its fixed cost: R ≥ F + C_m. Now let copying push the price toward marginal cost before F is recovered. Revenue collapses toward C_m, the condition fails, and at the margin the work is not made — not because anyone took the finished object, but because the prospect of the finished object can no longer pay for its own creation. The same logic, in present-value form, governs investment that pays out over time:

NPV  =  −F  +  Σ_t  π_t / (1+r)^t

If imitation drives the post-launch contribution π_t toward zero before F is amortised, the net present value goes negative and the project is never begun. The break-even can be read off directly: a creator who nets margin m per legitimate sale must clear

Q*  =  F / m

units to recover the fixed cost, and every sale captured by an unpaid copy raises that threshold. The copyist’s trick is to discourse fluently about marginal cost — which is indeed near zero — while behaving as though fixed cost did not exist. But the fixed cost is the whole problem. It is the years of writing, the failed compounds, the studio, the trials, the editing, the development. Marginal-cost pricing is wonderful for distributing a thing that already exists. It is fatal for financing a thing that does not yet exist.

How fatal depends entirely on the size of F and the cheapness of imitation, which is why the answer is sector-specific. For a self-published blog post, F is small and the argument is weak. For a new medicine, F is enormous — a median capitalised development cost of $985 million, with a mean nearer $1.34 billion, both in 2018 dollars (Wouters et al., 2020), or about $2.56 billion in 2013 dollars on the higher, confidential-survey estimate (DiMasi et al., 2016), in each case with the cost of failures folded into the cost of the successes — and imitation after approval is close to free (Figure 1). (The figure plots the $985 million median and the $2,558 million estimate; some summaries of Wouters round the median to “$1.1 billion,” but $985.3 million is the paper’s stated base-case median.) There the constraint binds with full force, and the post-expiry evidence shows just how far price falls once exclusivity ends: across eight high-income countries, Serra-Burriel and colleagues estimate that prices drop by around a third within a year of patent expiry and by more than four-fifths by year eight (Serra-Burriel et al., 2024). Generic competition is a triumph after the bargain has financed the molecule. The question is what happens when it arrives before.

5. The copyist’s Eden is a corporation: why abolition feeds scale

Here is the argument the abolitionist does not see coming, and it is the most important one in this essay.

A word on the term first, since I will use it without apology. By plutocracy I do not mean a conspiracy of wealthy men, and I am not name-calling. I mean a structural outcome: an institutional arrangement in which legal weakness for creators causes economic control to migrate toward those who already hold capital, distribution, manufacturing scale, platform control, and legal capacity. It is plutocratic not because anyone intends it but because, once the creator’s enforceable right is gone, those are the only assets left that can capture the value of what gets created.

The naïve market line holds that creators do not need enforceable rights because the market will reward them anyway — through reputation, prestige, originals, track record, first-mover advantage, and consumer goodwill. The reply is not that these mechanisms are worthless; sometimes they are decisive. The reply is that “the market” is not an agent that rewards desert. A market is a structure of incentives that exists only because institutions — property, contract, enforcement, courts — give it shape. This is the lesson of Coase’s analysis of social cost: even in his frictionless benchmark, market transactions presuppose a prior “delimitation of rights,” because entitlements that have not first been defined cannot be traded or recombined at all; and once transaction costs are positive — which, with copying and enforcement, they always are — the assignment of those rights determines how resources are actually allocated and who captures the gains (Coase, 1960). Strip away the rules and the market does not reward creators. It rewards whoever is positioned to capture the value, and that is rarely the author, the inventor, the small studio, or the small manufacturer.

Who, then, captures the value when the creator’s right is removed? The appropriability evidence answers cleanly. Cohen, Nelson and Walsh found that firms protect the returns to innovation through a portfolio of mechanisms, and that where formal patents are weak the dominant ones are secrecy, lead time, and “complementary marketing and manufacturing capabilities” (Cohen et al., 2000). Read that phrase slowly, because it is the hinge of the whole political economy. The fallback appropriation mechanisms are precisely the ones that scale: manufacturing capacity, distribution reach, marketing budgets, sales channels. When the law removes the creator’s enforceable right, it does not abolish appropriation; it transfers appropriation to whoever owns those complementary assets. This is not a novel worry — it is the central finding of David Teece’s 1986 analysis of why innovating firms so often fail to profit from their own innovations. When imitation is easy, or in his terms when the “appropriability regime” is weak, the profits flow not to the developer of the technology but to the owners of the complementary assets required to commercialise it; and in the limiting case he names explicitly, where incumbents control specialised complementary assets and the innovator cannot protect the technology, all of the profit from the innovation can accrue to the asset-holders rather than to the innovator (Teece, 1986). His canonical illustration is the EMI CAT scanner: EMI developed computed tomography — the greatest advance in radiology since the X-ray — and within roughly eight years had exited the business entirely, while better-positioned late entrants took the market. Arrow had seen a version of the same point in 1962, observing that the firm best able to bear the risk of invention is the large corporation with many projects, acting as its own insurer (Arrow, 1962). Scale was already advantaged in who could afford to create; strip away enforceable rights and scale becomes decisive in who gets to keep the proceeds of creation.

Figure 2 states the structure. With an enforceable right, the party who bore the fixed cost holds something they can license, sell, withhold, or defend, and the surplus is shared with them. Without it, the work is disclosed and freely copyable, and the surplus flows to whoever controls distribution, capital, manufacturing scale, platforms, and legal capacity. Abolition does not remove the monopoly problem. It relocates the monopoly rent from the creator to the incumbent.

Walk it through the industries the abolitionist likes to wave at.

Books. Strip copyright and a dominant retailer or platform can wait until an independent author has borne the cost and risk of writing and proving demand, then copy the successful title, issue its own edition, rank that edition above the author’s in its own search results, bundle it into a subscription, and underprice the original toward the marginal cost of a file. The platform bears no exploration risk; it lets the author bear it and harvests the proven winners. Prestige does not pay rent, and a track record does not stop appropriation by the entity that owns the channel. That copying genuinely displaces paid sales is not speculation: Reimers (2016), studying private anti-piracy enforcement on e-books with a difference-in-differences design, found that protection raised legal e-book sales by more than fourteen percent, mostly by deterring casual rather than professional infringement. The substitution is real and measurable; remove the right entirely and the substitution runs to whoever controls discovery.

Pharmaceuticals. This is the cleanest case, because the fixed cost is impossible to hide. A generic or large manufacturer can let the originator pay for discovery, trials, and approval, then copy the result the moment the law allows — efficient after exclusivity, ruinous before it. The investment response is not hypothetical. Gaessler and Wagner (2022), exploiting patent invalidations at the European Patent Office as an exogenous shock to expected exclusivity, found that the loss of a single year of market exclusivity lowers the likelihood that a drug project reaches approval by about 4.9 percentage points against a base approval rate of 30.8 percent — firms abandon the project the moment the protection falls. And the pull runs the other way too: Acemoglu and Linn (2004) found that a one percent increase in a drug category’s potential market raised the number of new drugs entering it by roughly six percent. Reward draws entry; removing the reward removes the entry. In a world of immediate copying, the firm that bore the fixed cost cannot recover it, and the firm with the manufacturing scale to copy fastest takes the market.

Software. Open-source and source-available licensing is not a refutation of copyright; it runs on copyright, a point developed in Section 7. Remove the enforceable right and a large firm can take a small developer’s source, integrate it, rebrand it, host it behind its own platform, and use lock-in and distribution to dominate, with no licence the developer can enforce. Source availability without an enforceable licence is not openness. It is surrender to whoever has the bigger platform.

Music and film. When digitisation collapsed the appropriability of recorded music, the surplus did not vanish into a creators’ commons; it migrated toward whoever controlled distribution and recommendation. The actor with the largest catalogue, the best ranking system, and the most attention captures the audience.

The political conclusion is the one Kinsella’s admirers should sit with. The anti-IP position presents itself as a blow against corporate power. Economically it is closer to the opposite. It removes the one asset a small creator can hold against a large firm — an enforceable right in the work — and leaves the field to the firms that already own capital, manufacturing, platforms, distribution, and legal departments. Anti-IP is not anti-corporate. In its abolitionist form it is a gift to the incumbents who own the channels. This does not make every existing right wise; it makes the removal of all rights a transfer of power upward, not downward.

6. “People created before copyright”: the argument that proves the opposite

A favourite move is to invoke the towering creators who worked before modern intellectual property — the dramatists, painters, and composers of the early-modern world — as living proof that rights are unnecessary. The premise is true and the conclusion does not follow.

That people created before modern IP proves only that creation can occur under alternative incentive systems. It does not prove that modern mass markets can finance costly creation without enforceable rights, because the alternative systems were doing real work. Pre-modern and early-modern creation leaned on patronage, court and church appointment, royal and printing privileges, guild control, commissions, restricted literacy, and — decisively — the sheer physical slowness and cost of copying. A fresco on a ceiling is hard to reproduce. A manuscript takes a copyist months. The frictions that protected the early-modern creator were technological, not legal, and they have largely been removed.

Look closely at the actual cases the abolitionist invokes, and every one turns out to be creation inside an institutional scaffold, not outside one. Shakespeare’s living came not from selling texts but from the economics of a playing company — a sharer’s stake in the Lord Chamberlain’s, later the King’s, Men, the takings of the Globe, and royal and aristocratic patronage — while the printed playbook was governed by the Stationers’ Company, whose register and Crown-granted printing privileges controlled who could lawfully reproduce a text long before modern copyright existed (the book-trade machinery Plant, 1934, documents in detail). Cervantes published Don Quixote under a royal printing privilege; Molière worked under royal patronage as head of the king’s own troupe. Michelangelo, Raphael and Leonardo painted on commission — the Sistine ceiling was a contract with Pope Julius II, not a speculative work sold into an open market — sustained by popes, princes, and the Medici. Bach held salaried church and court appointments at Köthen and Leipzig; Mozart assembled a living from a court post, commissions, subscription concerts, and publishers; Beethoven was carried by an annuity from aristocratic patrons and by his publishing income. Gutenberg, the very man who made cheap copying possible, financed his press with borrowed capital, was sued by his backer Johann Fust, and lost the press and the printed Bibles to him — an early demonstration that controlling the means of reproduction, not inventing it, is where the value lodges. And Rembrandt, at the height of his fame, was driven to insolvency in 1656 and had his possessions inventoried and sold. The moral is not that these figures show creativity needs no support; it is the reverse. Each was held up by a specific institution — patronage, privilege, appointment, guild, commission, or the sheer slowness of the copy — and the modern argument quietly assumes that if you removed today’s institution, creativity would carry on unaided, when the record shows it carrying on precisely because some institution bore its fixed cost. Pre-modern creativity proves creativity under alternative institutions, not creativity without institutions.

Arnold Plant, in the first sustained economic analysis of book copyright, made exactly this case in 1934 — and he made it as a copyright skeptic, which is precisely why it is useful here. Plant documented that authors were paid before and outside copyright: nineteenth-century American publishers, free to reprint any foreign work without paying anyone, nonetheless paid English authors handsomely for “advance sheets,” and authors sometimes earned more from their uncopyrighted American sales than from their protected British royalties. His point was that priority in the market — lead time — secured the return where copyright did not. He is right, and the honest reader must concede what follows: in the predigital book trade, lead-time appropriation worked, which is one reason books are a weaker case for copyright than medicines are.

But notice what Plant’s mechanism depended on. Lead time is worth something only when copying is slow and costly and the first mover enjoys a real head start; inferior, late, expensive copies tap a different market rather than destroying the original’s. That is the world of 1934. It is not the world of the costless, instantaneous, perfect digital copy, where the marginal cost of reproduction goes to zero and the head start evaporates. Landes and Posner saw the trajectory: improving copying technology steadily raises the need for copyright over time, because it erodes the very frictions — inferior copies, the head start — that once substituted for legal protection. Plant’s evidence proves that creativity does not require copyright in all conditions. It does not prove that creativity survives the collapse of the frictions his own mechanism relied on. Historical genius proves human creativity. It does not prove that a market of zero-marginal-cost copies can finance the expensive end of modern creation without rules.

7. “It’s socialism” and “open source proves you wrong”

Two further slogans deserve dismantling because they are so often treated as decisive.

The first is that intellectual property is a form of socialism, because it restricts what others may do with their own physical property — their paper, their presses, their servers. This inverts the actual economic structure of the abolitionist’s own proposal. Consider what the abolitionist asks for: the author bears the cost of writing; the inventor bears the cost of inventing; the laboratory bears the cost of discovery; the developer bears the cost of building — and then everyone else may copy the finished output and monetise it. That is not market liberalism. It is, in functional terms, collectivisation of the completed product while leaving the cost of production private. I would not call it socialism as a legal classification, because that is sloppy; I would call it what it is. If the creator must privately fund creation but any copyist may freely appropriate the finished work, the system redistributes from the producer to the imitator. The label “socialism” is not merely wrong when aimed at copyright; it more accurately describes the regime the abolitionist is proposing.

The second slogan is that open-source software proves intellectual property unnecessary. It proves the reverse. Open-source and source-available licensing exists because copyright gives the author the legal power to set the terms on which others may use the code. The GPL, the MIT licence, the Apache licence, dual-licensing arrangements, paid binaries, hosted services, and support contracts all presuppose an enforceable right in the work — without that right, there is no licence to grant and no condition to enforce, only code that anyone may strip, rebrand, and sell. Open source is not anti-copyright. It is copyright deployed as governance: the author uses the right to require sharing, attribution, or reciprocity, conditions that evaporate the instant the underlying right does. Source-visible is not ownerless. Making code auditable — for inspection, reproducible builds, signed releases, public verification — is a genuine and valuable security practice, but auditability is a property of disclosure, not a surrender of the licence. The honest claim about open source is the opposite of the abolitionist’s: it is one of the clearest demonstrations that an enforceable intellectual-property right is what makes a sophisticated commons possible.

8. The economics, calibrated: where the case is strong and where it is not

Having dismantled the abolitionist’s arguments, I owe the reader the other side of the ledger, because the calibrated view earns its credibility by refusing to defend bad intellectual property. The case against over-broad rights is strong, and it rests on the same economics.

Set the two welfare states against each other. Without protection, price falls toward marginal cost and current access rises, but future creation may fall where fixed costs cannot be recovered:

W_0  =  CS_0  +  PS_0  +  B_0

With protection, access falls during the term but future creation may rise through investment and disclosure, at the cost of deadweight loss, enforcement, and follow-on constraint:

W_1  =  CS_1  +  PS_1  +  B_1  −  DWL  −  E  −  C_F

The question is whether W_1 > W_0, and the answer is not uniform. It depends on the size of the fixed cost, the cheapness of imitation, whether the protected object is an end-product or an input into later work, and whether alternative appropriation mechanisms exist. This is why a single verdict on “intellectual property” is incoherent. It is not one thing doing one job.

On the creation side, the evidence that protection induces real output is strong where the fixed cost is real. Giorcelli and Moser (2020), studying Italian opera across the long nineteenth century, found that the introduction of basic copyright caused measurably more and better work — and, just as importantly, that extensions beyond the composer’s lifetime added nothing, which is the saturation of B_F made visible. In pharmaceuticals the incentive-sensitivity is sharp from both directions, as already seen: Acemoglu and Linn (2004) show reward drawing entry, Gaessler and Wagner (2022) show lost exclusivity driving exit, and the magnitude of the fixed cost (Wouters et al., 2020; DiMasi et al., 2016) and the depth of the post-expiry cliff (Serra-Burriel et al., 2024) make immediate copying incompatible with private financing. And the reason patents matter so much here specifically is the one Cohen, Nelson and Walsh (2000) supply: a regulator-disclosed, chemically reverse-engineerable molecule cannot be protected by secrecy or lead time at all, so in pharmaceuticals the formal right is not one appropriation mechanism among many — it is the only one that works.

But the same framework indicts protection when it is too strong, and the honest defender must say so plainly. When the protected object is an input into later work, rights can suppress the follow-on. Williams (2013), using the natural experiment of the human genome, estimated that gene-level intellectual property held by a private firm reduced subsequent research and product development on those genes on the order of 20 to 30 percent, persistently — though she is careful that she does not measure net welfare, since the firm’s entry was itself spurred by the prospect of the right. The copyright analogue is just as clean: Biasi and Moser (2021), studying a wartime program that licensed cheap reprints of enemy-owned science books, found that the resulting price drop of about a quarter raised follow-on citations by English-language authors by roughly 45 percent for each ten percent fall in price. Weaker protection of an input — patent or copyright — raised the cumulative science built on it. The fixed-term patent can also bend the direction of research, not merely its level: Budish, Roin and Williams (2015) showed that because the patent clock runs at a fixed length regardless of how long development takes, long-horizon projects are penalised — eight drugs approved to treat advanced lung cancer over a recent window, and not one ever approved to prevent it, because prevention trials run too long to leave effective patent life at launch.

Put the two halves together and the shape of the answer appears. Net welfare peaks at an interior optimum — neither zero nor infinite protection, but a finite, sector-specific T* (Figure 3). Protection below it is too little; protection above it is too much; and the entire substance of intellectual-property policy is locating that point and asking which side of it a given rule sits on.

The honest reading of the evidence places much of what we actually have on the wrong side of the optimum — life-plus-seventy copyright terms, evergreened drug patents, vague software patents, dense thickets — while placing basic copyright and basic patents on the right side, where they induce real creation. That is the conclusion that condemns both camps at once. The abolitionist who claims only that current intellectual property is far too strong is largely correct, and the mainstream evidence supports him. The abolitionist who claims that protection is therefore never productive is refuted by the same evidence, and so is the maximalist who treats every right as sacred.

9. What a serious case against IP would have to show

The way to end the slogans is to state the burden of proof and then notice that the abolitionist almost never carries it. To establish that abolishing intellectual property would raise welfare, the argument would have to show at least one of the following, with evidence rather than assertion: that fixed-cost creation would not fall materially without enforceable rights; that some alternative institution would finance it better; that creators can in fact recover their investment through lead time, secrecy, reputation, services, or contract under modern conditions of costless copying; that the static access gains exceed the dynamic creation losses; that the follow-on gains from weaker rights exceed the original creation forgone; that large-firm appropriation would not dominate small creators once the right is removed; or that enforcement costs and deadweight losses exceed the creation benefits.

These are answerable questions. Some of them, in some sectors, plausibly resolve in the abolitionist’s favour — the follow-on and access arguments are strong against over-broad rights, as Section 8 conceded. But they have to be shown, sector by sector, against the cost structure and the imitation cost, and the rhetorical devices this essay has dismantled do not show them. They evade them.

There is one more requirement, and it is the one the word “abolition” hides. Abolishing a financing mechanism is not a policy unless it names the replacement. If exclusivity is removed from a high-fixed-cost activity, something must take its place — public research funding, public trials, prizes, advance market commitments, procurement contracts, or regulated royalties — and each of those is a real institution with its own costs and failure modes, not the spontaneous generosity of an unregulated market. Some of these may even outperform patents in particular cases: antibiotics, where good stewardship means low sales volumes, and neglected diseases, where the market is too small, are obvious candidates for prizes or procurement rather than exclusivity. But that is an argument for replacing the instrument where it fails, not for abolishing the function it serves. “Abolish intellectual property and the market will provide” is not an institution. It is a hope, and the appropriability evidence says the hope is misplaced.

10. Conclusion

Strip away the metaphysics and a short list of conclusions survives, each of them institutional rather than theological.

Intellectual property is not sacred, and it is not infinite. It is not the ownership of ideas, and it is not justified by the mere assertion of a natural right. It is not refuted by the observation that it expires — that is evidence of calibration, not of fraud — and it is not refuted by island hypotheticals that delete every institution that makes the hard case hard. It is not socialism; if anything, the regime that privatises the cost of creation and socialises the finished work is the better fit for that label. And open source does not disprove it; open source is what an enforceable right looks like when an author chooses to govern a commons with it.

What intellectual property is is an institutional mechanism for aligning incentives where creation is costly and copying is cheap — a calibrated instrument whose right strength is an empirical, sector-specific question, strong where imitation would otherwise destroy fixed-cost recovery and the alternatives fail, weak where the protected object is mainly an input into further creation or where other mechanisms already work. The serious debate is not abolition versus maximalism. It is the location of the optimum, and the honest verdict is that much of our actual law sits well past it while the basic rights sit short of it.

The abolitionist’s deepest error is not a mistake about ideas. It is a mistake about when the story begins. He starts the analysis at the moment the work already exists, when the only remaining question is how cheaply it can be copied, and at that moment copying does indeed look like pure gain and exclusion like pure loss. Economics begins earlier — with the person who bore the cost, the risk, and the failures to bring the work into existence, and with the institutions that determine whether anyone will do so again. The copyist calls his world an Eden of liberty. It is an Eden only because he arrives after the harvest, and only for those already large enough to gather it.

This is not a theology of intellectual property. It is economics — and economics, unlike theology, gives different answers in different places.

References

Acemoglu, D., & Linn, J. (2004). Market size in innovation: Theory and evidence from the pharmaceutical industry. Quarterly Journal of Economics, 119(3), 1049–1090. https://doi.org/10.1162/0033553041502144

Arrow, K. J. (1962). Economic welfare and the allocation of resources for invention. In R. R. Nelson (Ed.), The Rate and Direction of Inventive Activity: Economic and Social Factors (pp. 609–626). Princeton University Press / NBER.

Biasi, B., & Moser, P. (2021). Effects of copyrights on science: Evidence from the WWII Book Republication Program. American Economic Journal: Microeconomics, 13(4), 218–260. https://doi.org/10.1257/mic.20190113

Boldrin, M., & Levine, D. K. (2008). Against Intellectual Monopoly. Cambridge University Press.

Budish, E., Roin, B. N., & Williams, H. (2015). Do firms underinvest in long-term research? Evidence from cancer clinical trials. American Economic Review, 105(7), 2044–2085. https://doi.org/10.1257/aer.20131176

Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1–44. https://www.jstor.org/stable/724810

Cohen, W. M., Nelson, R. R., & Walsh, J. P. (2000). Protecting their intellectual assets: Appropriability conditions and why U.S. manufacturing firms patent (or not) (Working Paper No. 7552). National Bureau of Economic Research. https://doi.org/10.3386/w7552

DiMasi, J. A., Grabowski, H. G., & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: New estimates of R&D costs. Journal of Health Economics, 47, 20–33. https://doi.org/10.1016/j.jhealeco.2016.01.012

Gaessler, F., & Wagner, S. (2022). Patents, data exclusivity, and the development of new drugs. Review of Economics and Statistics, 104(3), 571–586. https://doi.org/10.1162/rest_a_00987

Giorcelli, M., & Moser, P. (2020). Copyrights and creativity: Evidence from Italian opera in the Napoleonic age. Journal of Political Economy, 128(11), 4163–4210. https://doi.org/10.1086/710534

Kinsella, N. S. (2008). Against Intellectual Property. Ludwig von Mises Institute. (Originally published as Kinsella, N. S. (2001), Against intellectual property, Journal of Libertarian Studies, 15(2), 1–53.)

Landes, W. M., & Posner, R. A. (1989). An economic analysis of copyright law. The Journal of Legal Studies, 18(2), 325–363. https://doi.org/10.1086/468150

Moser, P. (2005). How do patent laws influence innovation? Evidence from nineteenth-century world’s fairs. American Economic Review, 95(4), 1214–1236. https://doi.org/10.1257/0002828054825501

Plant, A. (1934). The economic aspects of copyright in books. Economica, 1(2), 167–195. https://doi.org/10.2307/2548748

Reimers, I. (2016). Can private copyright protection be effective? Evidence from book publishing. The Journal of Law and Economics, 59(2), 411–440. https://doi.org/10.1086/687521

Serra-Burriel, M., Martin-Bassols, N., Perényi, G., & Vokinger, K. N. (2024). Drug prices after patent expirations in high-income countries and implications for cost-effectiveness analyses. JAMA Health Forum, 5(8), e242530. https://doi.org/10.1001/jamahealthforum.2024.2530

Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6), 285–305. https://doi.org/10.1016/0048-7333(86)90027-2

Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement), Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, Art. 33 (term of protection), Art. 27.1 (patentable subject matter), Art. 29 (disclosure).

United States Copyright Act, 17 U.S.C. § 102(a)–(b) (1976) (idea/expression distinction).

Waldfogel, J. (2017). How digitization has created a golden age of music, movies, books, and television. Journal of Economic Perspectives, 31(3), 195–214. https://doi.org/10.1257/jep.31.3.195

Waldfogel, J. (2018). Digital Renaissance: What Data and Economics Tell Us about the Future of Popular Culture. Princeton University Press.

Williams, H. L. (2013). Intellectual property rights and innovation: Evidence from the human genome. Journal of Political Economy, 121(1), 1–27. https://doi.org/10.1086/669706

Wouters, O. J., McKee, M., & Luyten, J. (2020). Estimated research and development investment needed to bring a new medicine to market, 2009–2018. JAMA, 323(9), 844–853. https://doi.org/10.1001/jama.2020.1166

Abstract. Intellectual property converts a non-rivalrous good — an idea, a molecule, a sequence of words — into a temporary, excludable, rivalrous one, accepting a static welfare loss in exchange for a dynamic incentive to create; whether that trade is worth making is not a matter of principle but of measurement, and the measurements differ sharply by sector, so this essay builds the trade explicitly and tests it against the evidence in the two industries where the stakes are clearest. For pharmaceuticals it shows, using the published cost estimates ($2,558M capitalised per approval in DiMasi, Grabowski and Hansen, 2016; a contested median of $985M in Wouters, McKee and Luyten, 2020), the industry cost of capital (10.5%), and measured post-expiry price collapse (US prices falling 32% within a year and 82% within eight years of patent loss, Serra-Burriel et al., 2024), that the appropriability gap is genuinely fatal: without exclusivity the asset that repays a billion-dollar fixed cost stops paying the instant generics enter, and a stylised break-even model demonstrates how steeply the required return rises as effective patent life shortens. For creative works it shows, via the Landes–Posner (1989) cost-of-expression model and the causal evidence (Giorcelli and Moser, 2020), that basic copyright did raise both the quantity and quality of output — but that protection beyond the creator’s life added nothing, that digitisation collapsed recorded-music revenue by more than half in real terms without collapsing output (Waldfogel, 2017, 2018), and that the contractual alternative favoured by the abolitionist position fails precisely where mass distribution makes appropriability hardest because it cannot bind the third party who never signed. The conclusion is neither that IP is a swindle nor that more of it is better: Kinsella’s (2008) non-rivalry premise is correct and is in fact the starting point of the orthodox welfare analysis, but his universal conclusion is falsified by the pharmaceutical and historical-copyright evidence, while the maximalist position is falsified by the same evidence pointing the other way — the welfare curve has an interior peak, and most of the live policy disputes are about how far we have wandered to the right of it.

1. The thing that makes ideas different

Begin with the property that every serious treatment of this subject begins with, because everything follows from it. An idea is non-rivalrous. If I teach you a method for harvesting cotton, I still have the method; your use does not consume mine. If you copy this essay, the original arrangement of words is undiminished. Kenneth Arrow put the point with characteristic economy in 1962: information, as an economic commodity, is peculiar because its use by one party does not deplete it, and once produced it can be reproduced at a cost approaching zero. Arrow framed this as the inappropriability of information — absent legal protection, any purchaser can reproduce it “at little or no cost,” so a competitive market cannot reward its creation — and drew the conclusion the rest of this essay must reckon with: a free-enterprise economy will therefore tend to underinvest in the production of knowledge. This is not a libertarian observation or a statist one. It is a physical fact about information, and it is the foundation of the entire field.

From that single fact, two opposite-looking conclusions can be drawn, and the whole debate is a quarrel between them.

The first conclusion is the orthodox welfare-economics one. Precisely because an idea can be copied at near-zero cost, a competitive market will drive the price of copies down to that near-zero marginal cost. But the first copy is not cheap. Discovering a drug, writing a novel, composing an opera, debugging a compiler — these carry a large fixed cost that does not vary with the number of copies sold. Landes and Posner (1989) gave this fixed cost its standard name: the cost of expression. If competition pushes price to marginal cost, the cost of expression is never recovered, and the rational creator, foreseeing this, does not create. The market under-provides the good. Intellectual property is the legal device that restores appropriability — that lets the creator charge above marginal cost for a limited period, so the wedge between price and cost can repay the fixed cost. We accept a static inefficiency (some buyers who value the good above its marginal cost are priced out) to buy a dynamic benefit (the good exists at all).

The second conclusion is the abolitionist one, and its sharpest modern statement is Stephan Kinsella’s Against Intellectual Property (2008). Property rights, Kinsella argues, exist for one reason only: to resolve conflict over scarce resources. Two people cannot both eat the same loaf, so we need rules about who owns the loaf. But two people can both use the same recipe. There is no scarcity, hence no possibility of conflict, hence no warrant for a property right. Intellectual property, on this view, does not protect property — it manufactures an artificial scarcity in something naturally abundant, and enforces it with state coercion that necessarily overrides real property rights in tangible things. Your copyright in a pattern of words restricts what I may do with my own paper, my own ink, my own printing press. The deadweight loss is real; the justification, Kinsella says, is not.

Notice what has happened. Both camps start from the identical premise — non-rivalry — and reach opposite destinations. That tells you the disagreement cannot be settled at the level of premises. It is an empirical question: does the absence of appropriability actually suppress the creation of valuable works below the social optimum, and if so, by how much, and in which industries? The rest of this essay is an attempt to answer that question with numbers rather than postures. I take the two industries where the answer matters most and where it turns out to differ most: pharmaceuticals, where the case for protection is at its strongest, and creative works, where it is at its most ambiguous.

2. The static cost, drawn precisely

Before the evidence, the mechanism — because you cannot weigh a trade-off you have not drawn. Figure 1 is the standard monopoly diagram, and it contains, in one picture, both the libertarian’s complaint and the economist’s reply.

Let inverse demand be P = a − bQ and let the marginal cost of reproducing a copy be a constant c. Two prices matter. The competitive price is P = c: this is what a market without intellectual property delivers, because free entry by copiers competes the price down to the cost of making one more copy. The monopoly price P_m is set where marginal revenue equals marginal cost, and it is strictly higher.

Three regions follow. The blue rectangle is producer surplus — the wedge (P_m − c) × Q_m. This is the only region that can repay the cost of expression F. Knock it out and the fixed cost is unrecoverable. The red triangle is the deadweight loss: the buyers who value the good above c but below P_m, who would have been served under competition and are not served under monopoly. This is the static cost of intellectual property, and — this is the point worth sitting with — Kinsella and Boldrin and Levine are simply correct that it exists. There is no honest version of the pro-IP argument that denies the red triangle. The green region is the consumer surplus that survives.

The orthodox claim is not that the red triangle is illusory. It is that, in the absence of the blue rectangle, the good is not produced, so the relevant comparison is not “monopoly versus competition” but “monopoly versus nothing.” If the work would exist either way, IP is pure deadweight loss and the abolitionist wins on that work. If the work exists only because of the blue rectangle, then the consumer surplus it generates — the entire area under the demand curve above price — is a benefit that would otherwise be zero, and it must be set against the triangle. Which case obtains is, again, empirical. So let us go and measure.

3. Pharmaceuticals: where the gap is fatal

3.1 The size of the fixed cost

The pharmaceutical industry is the cleanest natural experiment in the appropriability problem that exists, because it combines two features in the extreme: a colossal fixed cost of creation and a trivial marginal cost of imitation. A small-molecule drug that costs a fortune to discover and prove safe can, once its structure is public, be synthesised by a competent generic manufacturer for cents per dose.

How large is the fixed cost? The most-cited figure comes from DiMasi, Grabowski and Hansen (2016), who surveyed ten firms on a random sample of 106 compounds first tested in humans between 1995 and 2007. Their headline numbers, which I have read in the original rather than from any summary: the average out-of-pocket cost per approved compound is $1,395 million in 2013 dollars; capitalising those outlays to the point of approval at a real cost of capital of 10.5% — to account for the decade or more between spending and any revenue — yields $2,558 million; adding post-approval research raises it to $2,870 million. The capitalisation matters enormously. More than half of the headline figure is not cash spent on chemistry; it is the time-value of money tied up for years in projects that mostly fail. Indeed the engine of the estimate is failure: the overall probability that a compound entering human trials is eventually approved was estimated at just 11.83%, down from 21.5% in the authors’ earlier work. You are not paying for one drug. You are paying for the eight that died so that one could live.

That figure is contested, and intellectual honesty requires saying so plainly rather than leaning on the number that flatters the argument. The DiMasi study is funded in part by the industry whose costs it estimates, uses proprietary data that cannot be independently audited, and has drawn sustained methodological criticism. The most credible rival estimate, Wouters, McKee and Luyten (2020), used publicly available data — principally US Securities and Exchange Commission filings — for the 63 of 355 drugs and biologics approved by the FDA between 2009 and 2018 for which development costs could be reconstructed, applied the same 10.5% cost of capital, and found a median capitalised cost of $985 million (95% CI, $684–1,229 million) and a mean of $1,336 million — roughly a third to a half of DiMasi’s figure. The gap is partly explained by sample (Wouters’s public-data firms skew smaller and leaner) and partly by methodology. I do not adjudicate the dispute here. I carry both numbers through the model, because a conclusion that survives only under the larger figure is not a conclusion, it is special pleading.

Either way, the order of magnitude is the same and it is the only thing the argument needs: the fixed cost of a new drug is measured in the high hundreds of millions to low billions of dollars, and the marginal cost of a copy is measured in cents.

3.2 The cliff

Now watch what happens to the asset when its legal protection ends. Figure 2 traces the price of a branded drug over its commercial life.

During the protected period the brand sustains a price far above marginal cost; this is the blue window, the only interval in which the fixed cost F can be repaid. Then the patent expires, generics enter, and the price does not drift downward — it falls off a cliff. The two marked points are measured, not assumed. Serra-Burriel et al. (2024), in a difference-in-differences study of 505 drugs losing protection across eight high-income countries between 2011 and 2020, found that in the United States prices fell by 32% within the first year of patent loss and by 82% within eight years (95% confidence intervals 24–39% and 71–89%). The decline varied across countries — from 64% over eight years in Australia down to 18% in Switzerland — which is itself informative: the speed of the collapse is a function of how competitive the generic market is allowed to become. (One caveat the authors flag and I repeat: these are list prices, not net of the confidential rebates that pervade US drug pricing, so the true net collapse may differ in level though not in direction.)

For specific blockbusters the collapse is starker still. When atorvastatin (Lipitor) lost US protection in late 2011, the price of a month’s supply fell by more than 80% within roughly a quarter once multiple generic manufacturers were cleared to compete. The general pattern is well established in the health-economics literature: a single generic entrant produces a discount of perhaps 20–40%; the steep declines arrive as the second, third and later entrants pile in, because the post-entry market for a chemically identical molecule is close to the textbook ideal of perfect competition, and perfect competition does exactly what the textbook says — it drives price to marginal cost and economic profit to zero.

This is the appropriability problem of Figure 1, made concrete and given a clock. The generic firm did not pay the $985 million, or the $2,558 million, or whatever the true figure is. It free-rides on a fixed cost it did not bear, and in doing so it competes away the very wedge that was supposed to repay that cost. If generics could enter immediately, on the day of approval, the blue window would have zero width, the wedge would never open, and — by the logic of Figure 1 — the drug would not be developed in the first place. The patent does not enrich the innovator at the consumer’s expense in some gratuitous way. It is the device that makes the blue window wide enough to clear the fixed cost. Take it away and you do not get cheap drugs; you get no drugs.

3.3 The break-even, and why effective life is everything

We can be quantitative about how wide the window must be. Suppose, for the sake of a transparent illustration — and I flag every assumption rather than smuggling it, because an unstated assumption is a defect — that a drug earns a constant annual gross margin M (price minus marginal manufacturing cost, times volume) for L years of effective on-market exclusivity, after which the cliff takes the margin to zero. The fixed cost F is already risk-adjusted in the DiMasi and Wouters methodologies, since both link the cost of failures to the successes, so I treat the cash flow as deterministic. To break even at the industry real cost of capital r = 10.5%, the present value of the margin stream must equal the fixed cost:

F  =  M · [ 1 − (1 + r)^(−L) ] / r

which rearranges to the required annual margin

M  =  F · r / [ 1 − (1 + r)^(−L) ].

Figure 3 plots this hurdle against effective patent life, for both cost estimates.

The lesson is in the convexity. With a ten-year effective life, a $2,558 million drug must clear about $425 million in gross margin every year, simply to return its cost of capital; a $985 million drug must clear about $164 million. Stretch the effective life to fourteen years and the hurdles fall to roughly $357 million and $137 million. Compress it toward the shaded region on the left — a long clinical programme that eats years off the patent term before launch, or an early cliff from a successful patent challenge — and the required margin climbs steeply, because there are fewer years over which to amortise the same fixed cost.

This is not an abstraction. The patent term is twenty years from filing, but filing happens early, before the decade-plus of trials; what remains at launch is the effective life, and it is the only thing that pays. Budish, Roin and Williams (2015) showed in the American Economic Review that this design has a perverse consequence the cliff model predicts. Because the patent clock starts before clinical development and runs at a fixed length regardless of how long that development takes, drugs that take longer to bring to market — preventives and treatments for early-stage disease, whose trials must run for years to show a mortality benefit — enjoy a shorter effective life and a correspondingly higher break-even hurdle. The result is a documented distortion in the direction of research: over a recent five-year window, eight new drugs were approved to treat advanced lung cancer, every one of them for the most late-stage patients whose trials read out fastest, while not a single drug has ever been approved to prevent lung cancer, and only six have ever been approved to prevent any cancer at all. The market is not failing to invest because prevention is scientifically impossible. It is failing to invest because the fixed-term patent makes the long-horizon project uneconomic. Note what this establishes: patents matter so much in this industry that the precise shape of the patent rule bends the trajectory of medical research. That is the strongest possible evidence that the appropriability constraint binds.

So for pharmaceuticals the verdict is unambiguous, and it cuts against the abolitionist. The fixed cost is real, enormous, and audited (if disputed); the marginal cost of imitation is trivial; the post-expiry price collapse to near marginal cost is measured; and the historical record contains no example of a research-intensive pharmaceutical industry sustaining itself on free-ridden molecules. Here the blue rectangle of Figure 1 is load-bearing. Knock it out and the building falls.

Two pieces of causal evidence close off the obvious escape — the hope that the break-even arithmetic is a paper constraint firms would simply absorb. They do not absorb it; entry tracks the expected reward on both sides of the equation. On the demand side, Acemoglu and Linn (2004) used exogenous variation in market size driven by US demographic shifts and found that a one percent increase in the potential market for a drug category raised the number of new drugs entering that category by about six percent — roughly four percent for new non-generic drugs and four to six percent for new molecular entities, the most research-intensive class. Bigger expected markets pull more medicines into existence. On the supply side, Gaessler and Wagner (2022) exploited a natural experiment — drugs whose underlying patents were invalidated in European Patent Office opposition proceedings, an exogenous shock to expected exclusivity — and found that the loss of a single year of market exclusivity lowers the likelihood that a drug development project reaches approval by about 4.9 percentage points, against an unconditional approval rate of 30.8 percent; firms overwhelmingly abandon the project the moment the patent falls. The first result says reward draws entry; the second says shortened exclusivity drives exit. Both are exactly the sign the break-even model predicts, measured rather than assumed.

4. Creative works: where the gap is real but smaller, and shrinking

4.1 The cost of expression, and its discontents

Now the harder case, and the one Kinsella’s argument fits more comfortably. Landes and Posner (1989) — whose model I have read in the original and which remains the most careful economic statement of the pro-copyright position — apply exactly the Figure 1 logic to books. The author and publisher incur a cost of expression (the author’s time, editing, typesetting) that is independent of the print run. The marginal cost of a copy is the cost of printing, binding and distribution. Without copyright, a reprinter buys one copy and undersells the originator down to marginal cost; foreseeing this, the originator does not write. The structure is identical to the drug case.

But here is the crucial difference, and Landes and Posner are scrupulous about it in a way that maximalist copyright advocates who invoke them are not. They devote pages to the qualifications that limit the need for copyright, and the qualifications are substantial:

• Inferior copies. When a copy is a poor substitute — a blurry photostat, a cheaply bound reprint — the originator can price above marginal cost without legal protection.

• The head start. Copying takes time. There is an interval during which the originator faces no competition, and for faddish works whose demand is front-loaded, that interval alone may suffice.

• Indirect appropriability. The originator can sometimes capture the value of copies by charging more for the original that enables them — the mechanism by which a publisher of academic journals prices in the photocopying it cannot prevent.

• Non-pecuniary and indirect returns. Many authors are paid for writing in coin other than royalties: prestige, a professorship, consulting income, the platform a book confers. Where these dominate, the royalty stream — and hence the copyright that protects it — is not what motivates creation.

This sceptical line is not new, and it is worth registering its pedigree because it shows the appropriability-without-copyright argument is older than the platforms now accused of self-interest in making it. Arnold Plant (1934), in the first sustained economic analysis of book copyright, pressed it with historical evidence: through the nineteenth century anyone in the United States was free to reprint a foreign work, yet American publishers paid English authors handsomely for “advance sheets,” and authors sometimes earned more from their uncopyrighted American sales than from their protected British royalties — priority in the market, not legal monopoly, secured the return. Plant’s verdict was that copyright raises the variety of titles published while reducing access to them — more books are written, but fewer copies of the ones readers want reach them — and that, beyond a point, it bends authors toward writing whatever maximises monopoly profit rather than what they would otherwise write. That last point anticipates, in the creative domain, the direction-distortion this essay has already met in pharmaceuticals.

And then they make the observation that detonates the maximalist position from within. Beyond some level, more copyright protection is counterproductive, because creating a new work requires borrowing from old ones — characters, situations, chord progressions, prior results — and the stronger and longer the protection on existing works, the higher the cost of expression for every future creator who must license, search around, or substitute. Every author is simultaneously a borrower from the past and a creator for the future; the borrower wants weak protection on others’ work, the creator wants strong protection on her own, and the social optimum balances the two. This is not a critique imported by opponents of copyright. It is built into the foundational economic defence of copyright. It means the policy question was never “copyright: yes or no.” It was always “copyright: how much,” and the answer is an interior optimum, not a maximum.

4.2 The causal evidence: basic copyright works; extensions do not

Theory can establish that an optimum exists without telling you where it sits. For that you need an experiment, and creative-industry experiments are rare because copyright is usually changed by lobbying, not by chance. Giorcelli and Moser (2020), in the Journal of Political Economy, found a clean one. Napoleon’s military campaigns brought French copyright law — France’s copyright statute of 1793, granting composers exclusive rights tied to their own lifetimes — to two Italian states, Lombardy and Venetia, in 1801, while the remaining Italian states, sharing the same language and culture and exposed to the same French influence, did not adopt copyright. Comparing the production of new operas across eight Italian states using newly assembled data on 2,598 operas created between 1770 and 1900 in a difference-in-differences design, they found three results, and all three matter.

First, the adoption of basic copyright caused a significant increase in the number of new operas — their estimates imply a roughly five-fold increase in the number of historically popular operas in the states that gained protection. Second, it raised their quality, measured by how popular and how durable they proved — copyright induced not just more work but better work, because it let composers profit from a piece’s continued performance rather than a single commissioning fee. Third — and this is the result the maximalists never quote — extensions of copyright beyond the composer’s life produced no measurable benefit at all. The incentive to create responds to basic protection; it does not respond to the prospect of one’s heirs collecting royalties decades after one’s death.

Read those three findings against Figure 1 and Landes–Posner and they lock together perfectly. Basic copyright moved output up the rising part of the welfare curve: the blue rectangle was load-bearing for opera as it is for drugs. But term extension is the flat-to-falling part: it adds nothing to the dynamic incentive (a composer’s behaviour in 1805 is not altered by his estate’s revenues in 1860) while it does add to the static cost and the cost of expression for later borrowers. Figure 4 is the unifying picture.

The dynamic gain from inducing new works rises with protection but saturates — past a point, more protection induces no further creation, exactly as Giorcelli and Moser found for extensions. The static and access costs rise without saturating. Net welfare therefore peaks at an interior T* and declines thereafter. The question “is intellectual property good?” is, on this picture, malformed. Protection below T* is too little; protection above T* is too much; and the entire substance of intellectual-property policy is locating T* and asking which side of it we are on. The causal evidence says basic copyright and basic patents sit to the left of T* (they bind, they induce real creation), while life-plus-seventy copyright terms, evergreened drug patents, and dense patent thickets sit to the right (they impose cost without buying incentive). Both the naive abolitionist and the reflexive maximalist are answering the wrong question.

4.3 The Amazon problem: when the marginal cost of a copy goes to zero

The reader asked, specifically, what happens when a distributor like Amazon — or anyone with a server — can copy and distribute a work at will. The model answers precisely. Everything in Figure 1 turns on the marginal cost of a copy, c. For a physical book, c is positive but modest. For a digital file, c is approximately zero. As c → 0, the competitive price goes to zero, the blue appropriability rectangle collapses, and — on the pure theory — the work cannot recover its cost of expression at all. Digital reproduction does not weaken the appropriability problem; it perfects it. This is why Landes and Posner, writing in 1989, already saw the trajectory: improving copying technology, they noted, steadily raises the need for copyright over time, because it erodes every one of the practical frictions (inferior copies, the head start) that once substituted for legal protection. By their logic, the digital era should have been an extinction event for creative production.

It was not. And the reason it was not is the single most important empirical finding in this whole literature, because it complicates the tidy theory in a way the theory did not anticipate. Joel Waldfogel (2017, in the Journal of Economic Perspectives, and at book length in 2018) documented what actually happened when digitisation — Napster and its successors — devastated the appropriability of recorded music. United States recorded-music revenue began a precipitous slide in 1999 and, on the series Waldfogel uses — the RIAA’s reported value of US music shipments, inflation-adjusted to constant 2016 dollars — fell to roughly 25 percent below its 1999 peak by 2005 and, in real terms, by more than half by 2012, with international sales off by a similar fraction (Waldfogel, 2018). By the orthodox model, that revenue collapse should have produced a collapse in the quantity and quality of new music. It did not. On balance, Waldfogel finds, digitisation increased the number of new products created and made available to consumers; and because product quality is unpredictable ex ante, more releases mean more high-quality draws, so the quality of the best new music — measured by critics’ best-of lists and by what listeners actually consume — held up or improved rather than declining. The shift shows up structurally too: the share of top-selling albums released by independent labels grew from roughly 12% to 35% between 2000 and 2010. The same pattern, he found, held for film, books and television.

Why? Because the very technology that destroyed appropriability also destroyed the cost of expression. The fixed cost F that copyright exists to protect was itself a creature of the predigital production and distribution system — the studio, the pressing plant, the distributor, the shelf space, the gatekeeper. Digitisation let a musician record at home, a novelist self-publish, a filmmaker distribute globally without any of that apparatus. F fell at the same time as c fell. And if F falls far enough, the appropriability problem shrinks even as appropriability itself evaporates, because there is simply less fixed cost that needs recovering. Waldfogel’s “gatekeepers” — the studios and labels whose high fixed costs the old copyright bargain subsidised — turned out to have been, in significant part, a cost the system imposed rather than a value it created.

This is the right place to guard against over-reading that result, because the aggregate survival of creative output is not the same as copying being costless to the individual creator, and the honest evidence cuts both ways. Reimers (2016), in the Journal of Law and Economics, isolated the substitution effect directly: using a difference-in-differences design on the e-book and physical sales of RosettaBooks titles between 2010 and 2013, she found that private anti-piracy enforcement raised legal e-book sales — the closest substitute for an unauthorised download — by more than fourteen percent, with most of the gain coming from deterring casual rather than professional infringement. So copying does displace paid sales where the copy is a genuine substitute; the displacement is measurable and not small. What the music evidence shows is not that θ, the substitution rate, is zero, but that the fall in F outran it in the aggregate. Both facts are true at once, and a serious account holds them together rather than choosing the convenient one: protection still does real work for the marginal creator whose costs are not trivial, even as digitisation lowered the costs that made protection necessary in the first place.

This finding must be handled carefully, and not oversold, because it has been weaponised. The result is genuinely inconvenient for the maximalist case: it shows that creative output is far more robust to the loss of appropriability than the simple model predicts, at least where production costs have fallen in step. But Waldfogel’s argument has also been seized upon by parties — large technology platforms among them — with a direct commercial interest in weak copyright, and it has fair critics. It measures aggregate quantity and average quality over a window of years; it does not measure the long-run effect on the professional creative class of removing the income floor, nor what happens to expensive, slow, fixed-cost-heavy forms (the big-budget film, the multi-year investigative book) for which the production cost did not fall. The honest reading is narrow and it is this: the digital era falsifies the claim that strong copyright is necessary for a flourishing creative output in the forms where production costs collapsed alongside reproduction costs. It does not establish that copyright is unnecessary everywhere, and it says nothing about pharmaceuticals, where the fixed cost has done the opposite of fall.

4.4 The international and out-of-contract problem, and why it sinks the contractual alternative

The reader also asked about international distribution, online works, and the case where a copier is “outside of contract.” This is not a footnote; it is the precise point on which the most sophisticated version of the abolitionist position breaks.

Intellectual-property rights are territorial. A copyright or patent is granted by a state and enforceable within its jurisdiction; a work freely copied in a country that does not recognise or enforce the right escapes the creator’s protection entirely. Online distribution makes every work instantly transnational, so the weakest enforcement regime anywhere on earth tends to set the effective ceiling on appropriability everywhere. This is a real and serious erosion, and it is one reason the empirical appropriability of digital works is so much lower than the law on the books would suggest.

Now, the standard libertarian rejoinder — Kinsella’s, and the broader Rothbardian tradition’s — is that creators do not need state-granted monopolies at all, because they can protect their works by contract. Sell the book or the film under a licence that forbids copying; enforce the licence against violators; no artificial scarcity, no state IP, just ordinary freedom of contract. The proposal is coherent in a world of few buyers and durable relationships. It collapses in a world of mass digital distribution, for a reason that is purely economic and that Landes and Posner identified in 1989: contracts bind only the parties to them. This is the doctrine of privity, and it is not a technicality. If I sell you a file under a no-copying licence, that licence binds you. It does not bind the stranger who later downloads the file from a server, because the stranger never agreed to anything. There is no privity between the creator and the marginal copier. Contract reaches the second party; it cannot reach the third. And in mass distribution the third parties are the entire problem — they are millions, anonymous, and contractually untouched.

This is why only a property-style right — a right good against the world, which is precisely what intellectual property is and what a contract is not — can reach the unconsenting copier. The contractual alternative does not avoid the need for IP; it quietly presupposes a world small enough that everyone who touches the work is in privity with the creator, which is exactly the predigital world whose frictions made copyright less necessary in the first place. Remove those frictions — go online, go global, go to zero marginal cost — and the contractual solution fails in the same breath and for the same reason that the appropriability problem becomes acute. The abolitionist cannot have it both ways: the conditions under which contracts suffice are the conditions under which copyright was least needed, and the conditions under which copyright is most needed are the conditions under which contracts fail.

5. Answering Kinsella on the merits

We can now state, precisely and without either deference or dismissal, what is right and what is wrong in the case for abolition — because the reader asked for the economics of it, not the rhetoric.

What is right. The non-rivalry premise is correct, and it is not a minor concession. Kinsella is right that ideas are not scarce in the physical sense, right that intellectual property manufactures an excludability that does not occur in nature, and right that this manufacture imposes a real cost — the red triangle of Figure 1, which no honest economist denies. He is also right, directionally, that much of the intellectual-property edifice we actually have is unjustified: the causal evidence (Giorcelli and Moser on copyright extensions; Budish, Roin and Williams on the distortions of fixed patent terms) places a great deal of existing protection to the right of T*, where it imposes cost without buying incentive. Boldrin and Levine (2008) marshalled the broader case that patent systems frequently retard rather than promote innovation in industries characterised by cumulative, follow-on invention — they go further than this essay will, calling intellectual monopoly “an unnecessary evil,” a consequence of innovation rather than its cause — and on those industries the case has force. Moser (2005) bears on the same question from the opposite end of history. Using roughly fifteen thousand innovations exhibited at the world’s fairs of 1851 and 1876, she shows that patent laws shaped the direction of innovation rather than merely its level: inventors in countries without patents clustered in industries where secrecy did the work patents would have done — scientific instruments, food processing, dye stuffs — while patent-granting countries spread across the spectrum, and when the Netherlands abolished its patent system in 1869 the share of Dutch innovation in food processing rose from 11 to 37 percent. That is the same lesson as Budish, Roin and Williams, read backwards: patents do not simply turn invention up or down; they bend what gets invented. There is a second strand the abolitionist gets right, and it is the strongest empirical case against over-broad rights: when the protected object is itself an input into later work, protection can measurably suppress the follow-on. Williams (2013), in the Journal of Political Economy, used the natural experiment of the human genome — genes sequenced first by the private firm Celera and held under its contractual IP, against genes sequenced by the public effort and placed immediately in the public domain — and estimated that Celera’s gene-level IP reduced subsequent scientific research and product development on those genes on the order of 20 to 30 percent, an effect that persisted even after the IP lapsed. (Williams is scrupulous that she does not measure net welfare: Celera’s entry was itself spurred by the prospect of IP and may have sped sequencing, so the follow-on loss is one side of a ledger, not the whole of it — a caveat the honest reader keeps.) The copyright analogue is just as clean. Biasi and Moser (2021), in American Economic Journal: Microeconomics, studied the wartime Book Republication Program, under which the United States licensed cheap reprints of enemy-owned German science books: the average title’s price fell about 25 percent, and they find that a ten percent decline in price raised follow-on citations by English-language authors by roughly 45 percent, with treated books drawing on the order of 80 percent more citing works than comparable controls. Weaker protection of an input — patent or copyright — raised the cumulative science built on it. An abolitionist who claimed only that current intellectual property is far too strong would be defending a position the mainstream evidence substantially supports.

What is wrong. The universal claim — that intellectual property never generates net social benefit, that creation would proceed undiminished without it — is falsified, and not by theory but by measurement. The premise that non-rivalry implies no warrant for exclusion does not survive contact with the fixed-cost problem, because the orthodox argument was never that ideas are scarce; it was that the incentive to produce them is, and that incentive is scarce precisely because ideas are non-rivalrous and therefore unappropriable in a competitive market. Kinsella’s argument treats the non-rivalry of the finished idea as decisive, when the economically relevant scarcity is in the costly, uncertain, failure-ridden process that produces it. The pharmaceutical evidence shows that where that process is expensive enough and imitation cheap enough, the absence of exclusion does not merely reduce output at the margin — it can eliminate the category, as the eleven-percent success rate and the billion-dollar fixed cost and the eighty-percent price cliff together demonstrate. And it is no accident that pharmaceuticals are the sector where the constraint bites hardest: Cohen, Nelson and Walsh (2000), surveying 1,478 US manufacturing R&D labs, found that firms in most industries rank patents the least important of their appropriation mechanisms, relying instead on secrecy and lead time — but that patents are decisive in a small set of industries, “most notably pharmaceuticals,” precisely because a regulator-disclosed, chemically reverse-engineerable molecule cannot be protected by secrecy or a head start at all. This is the empirical hinge of the whole essay: appropriability mechanisms are sector-specific, so the case for patents is strong exactly where the alternatives fail and weak where they do not. The historical-copyright evidence shows that basic protection caused measurably more and better creative work. You may argue about how strong protection should be; you may not argue, against this evidence, that it is never productive.

The deepest error in the abolitionist argument is methodological, and it is worth naming because it is the error this essay has tried throughout to avoid. It derives a universal empirical conclusion — IP never helps — from an a-priori premise about the nature of ideas. But whether a particular institution, in a particular industry, produces net benefit is not the kind of question an axiom can answer. It is the kind of question only data can answer, and the data answer it differently in different industries. That is the finding. There is no single correct verdict on intellectual property, because intellectual property is not one thing doing one job. It is a deadweight loss imposed on opera that paid for itself, an extension imposed on opera that did not, a twenty-year monopoly on a molecule without which the molecule would not exist, and an evergreened thicket on a molecule that distorts research away from the patients who need it most — all at once.

6. What the models actually license

Strip away the rhetoric on both sides and the four pictures in this essay license a small number of conclusions, stated flatly.

The static cost of intellectual property is real and is correctly identified by its critics (Figure 1). It is justified only where the dynamic benefit exceeds it, which is an empirical condition, not a presumption.

In pharmaceuticals the condition is met decisively. A fixed cost in the high hundreds of millions to low billions (DiMasi et al., 2016; Wouters et al., 2020), a marginal imitation cost near zero, and a measured post-expiry price collapse of 32% in a year and 82% in eight (Serra-Burriel et al., 2024) together make exclusivity the precondition of the asset’s existence (Figures 2 and 3). The live problem here is not too much protection in the aggregate but its design: the fixed patent term bends research toward fast-reading late-stage treatments and away from slow-reading prevention (Budish, Roin and Williams, 2015), a distortion that better-designed protection — endpoint reform, term structured around effective rather than nominal life, targeted subsidy for long-horizon work — would relieve. Stronger protection is not the answer; smarter protection is.

In creative works the condition is met for basic protection and fails for extended protection. Copyright at its founding level raised both the quantity and quality of output (Giorcelli and Moser, 2020); copyright beyond the creator’s life did not (the same study); and the welfare curve’s interior peak (Figure 4) tells you that the long terms and broad scope of contemporary copyright sit on its falling side. Digitisation, by collapsing the cost of expression at the same time as the cost of copying, has shrunk the appropriability problem in the forms where production costs fell, without abolishing it and without touching the forms where they did not (Waldfogel, 2017, 2018). The contractual alternative that abolitionists offer in copyright’s place fails on privity exactly where mass distribution makes it most needed.

And the unifying conclusion, the one the reader asked for and the one the evidence actually supports rather than the one any prior would prefer: the right quantity of intellectual property is neither zero nor unbounded. It is an interior optimum that differs by industry, and the principal pathology of the actual system is not that it exists but that, in most of the domains where we can measure, it has wandered to the right of its own peak — long after the marginal work it was meant to call into being has already been written, and long past the point where the molecule it was meant to fund has already been found.

References

Acemoglu, D., & Linn, J. (2004). Market size in innovation: Theory and evidence from the pharmaceutical industry. Quarterly Journal of Economics, 119(3), 1049–1090. https://doi.org/10.1162/0033553041502144

Arrow, K. J. (1962). Economic welfare and the allocation of resources for invention. In R. R. Nelson (Ed.), The Rate and Direction of Inventive Activity: Economic and Social Factors (pp. 609–626). Princeton University Press / NBER.

Biasi, B., & Moser, P. (2021). Effects of copyrights on science: Evidence from the WWII Book Republication Program. American Economic Journal: Microeconomics, 13(4), 218–260. https://doi.org/10.1257/mic.20190113

Boldrin, M., & Levine, D. K. (2008). Against Intellectual Monopoly. Cambridge University Press.

Budish, E., Roin, B. N., & Williams, H. (2015). Do firms underinvest in long-term research? Evidence from cancer clinical trials. American Economic Review, 105(7), 2044–2085. https://doi.org/10.1257/aer.20131176

Cohen, W. M., Nelson, R. R., & Walsh, J. P. (2000). Protecting their intellectual assets: Appropriability conditions and why U.S. manufacturing firms patent (or not) (Working Paper No. 7552). National Bureau of Economic Research. https://doi.org/10.3386/w7552

DiMasi, J. A., Grabowski, H. G., & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: New estimates of R&D costs. Journal of Health Economics, 47, 20–33. https://doi.org/10.1016/j.jhealeco.2016.01.012

Gaessler, F., & Wagner, S. (2022). Patents, data exclusivity, and the development of new drugs. Review of Economics and Statistics, 104(3), 571–586. https://doi.org/10.1162/rest_a_00987

Giorcelli, M., & Moser, P. (2020). Copyrights and creativity: Evidence from Italian opera in the Napoleonic age. Journal of Political Economy, 128(11), 4163–4210. https://doi.org/10.1086/710534

Kinsella, N. S. (2008). Against Intellectual Property. Ludwig von Mises Institute. (Originally published as Kinsella, N. S. (2001), Against intellectual property, Journal of Libertarian Studies, 15(2), 1–53.)

Landes, W. M., & Posner, R. A. (1989). An economic analysis of copyright law. The Journal of Legal Studies, 18(2), 325–363. https://doi.org/10.1086/468150

Moser, P. (2005). How do patent laws influence innovation? Evidence from nineteenth-century world’s fairs. American Economic Review, 95(4), 1214–1236. https://doi.org/10.1257/0002828054825501

Plant, A. (1934). The economic aspects of copyright in books. Economica, 1(2), 167–195. https://doi.org/10.2307/2548748

Recording Industry Association of America. (2024). U.S. recorded music revenue database (1973–2024). RIAA. https://www.riaa.com/u-s-sales-database/

Reimers, I. (2016). Can private copyright protection be effective? Evidence from book publishing. The Journal of Law and Economics, 59(2), 411–440. https://doi.org/10.1086/687521

Serra-Burriel, M., Martin-Bassols, N., Perényi, G., & Vokinger, K. N. (2024). Drug prices after patent expirations in high-income countries and implications for cost-effectiveness analyses. JAMA Health Forum, 5(8), e242530. https://doi.org/10.1001/jamahealthforum.2024.2530

Waldfogel, J. (2017). How digitization has created a golden age of music, movies, books, and television. Journal of Economic Perspectives, 31(3), 195–214. https://doi.org/10.1257/jep.31.3.195

Waldfogel, J. (2018). Digital Renaissance: What Data and Economics Tell Us about the Future of Popular Culture. Princeton University Press.

Williams, H. L. (2013). Intellectual property rights and innovation: Evidence from the human genome. Journal of Political Economy, 121(1), 1–27. https://doi.org/10.1086/669706

Wouters, O. J., McKee, M., & Luyten, J. (2020). Estimated research and development investment needed to bring a new medicine to market, 2009–2018. JAMA, 323(9), 844–853. https://doi.org/10.1001/jama.2020.1166

Keywords: intellectual property, copyright, patents, trade secrets, trademarks, Kinsella, Hoppe, Rothbard, property rights, scarcity, rivalry, idea–expression dichotomy, appropriability, public goods, fixed costs, marginal cost, free-riding, in rem rights, institutional facts, law and economics, Arrow, Landes and Posner, Machlup, authorship, first publication, appropriation by scale.

Abstract

Stephan Kinsella’s case against intellectual property does not argue for its central premise; it borrows it. The premise — that only rivalrous physical resources can be owned — is taken on the authority of Hans-Hermann Hoppe and Murray Rothbard and treated as settled, when it is precisely the proposition in dispute; and at least one of the borrowings is a misuse, since Rothbard, the authority enlisted, defended copyright as a legitimate market right, while the Holmes dissent quoted in support rests on a premise (”property, a creation of law”) that destroys the natural-rights framework it is meant to serve. Once the premise is examined rather than asserted, the argument fails on its own terms. It commits a straw man by abstraction (treating intellectual property as ownership of “ideas” or of “value,” when the operative legal categories protect bounded objects — original expression, claimed inventions, confidential information, source-identifying marks — and exclude ideas, methods, facts, and effort); it equivocates on “property” (denying that intangible juridical relations can attach to authorship while building its own system entirely out of intangible juridical relations — title, contract, consent, aggression, transfer, remedy, and the very norms of argumentation); it deploys a false dichotomy (physical integrity versus subjective value) that omits the real category, defined legal interests not reducible to anyone’s valuation; and the control criterion it wields against “value-rights” actually licenses copyright, because a would-be copier can determine in advance, independently of anyone’s subjective valuation, whether he is reproducing another’s work. The economic core of the matter is the one the theory cannot represent: expressive and inventive goods carry high fixed costs of creation and near-zero marginal costs of reproduction, so that in the absence of any right of exclusion the creator bears the cost of resolving uncertainty while the better-capitalised copier harvests the resolved product — a classic appropriability problem identified by Arrow and modelled by Landes and Posner. A world without authorial rights is therefore not a world of liberated competition but one in which wealthy production houses copy finished works, undersell originators who alone bore the fixed cost, flood distribution, capture derivative formats, and diversify the residual risk across many appropriated authors. That outcome is not the abolition of monopoly; it is the transfer of the market to whoever owns the press. The honest qualifications are stated rather than hidden: the empirical magnitude of under-provision is genuinely contested (Plant, Boldrin and Levine, and Machlup’s own verdict of indeterminacy are engaged, not suppressed), and some of what Hoppe and Rothbard say about value and about production-as-transformation is correct and is conceded; but neither qualification rescues Kinsella’s argument, because his conclusion is asserted a priori from a premise that is question-begging, self-undermining, and — when it touches real commerce — a recipe for appropriation by scale.

I. The method before the argument

Begin with how the piece is built, because the construction is the first defect.

Kinsella’s article on Hoppe is, in form, an anthology. It assembles passages from Hoppe’s A Theory of Socialism and Capitalism and The Economics and Ethics of Private Property, a passage from Rothbard’s Man, Economy, and State, and a one-line dissent from Justice Holmes, and binds them with connective sentences asserting that one may not own the “value” of property, only its physical integrity, and that creative labour increases wealth without generating any right. The foundational proposition — that legitimate ownership reaches only rivalrous physical objects — is never independently defended in the post. It is quoted into place. The reader is invited to treat it as established because distinguished men have said it.

This matters because that proposition is the entire question. The dispute between the defender and the critic of intellectual property is exactly the dispute over whether ownership is confined to rivalrous physical resources. To answer that dispute by citing authors who assume the confining definition is not to argue; it is to restate the conclusion in someone else’s voice. An argument earns a premise. An anthology assigns it.

Three of the borrowings deserve particular attention, because they are not merely unargued — they are misapplied.

First, Rothbard. The post enlists Rothbard’s statement that an owner has “no vested right in the value of his property, only in its physical existence.” Read in context, that sentence concerns competitive substitution: a rival who lures away your customers lowers the monetary value of your factory and commits no wrong, because he has not touched it. True, and conceded below. But the post lets Rothbard’s authority bleed from that narrow and correct point onto the broad anti-IP conclusion — and on the broad conclusion Rothbard is a witness for the other side. In the very same treatise, Rothbard defended copyright and attacked only patents, holding that <cite index=”90-1”>copyright is a logical attribute of property right on the free market, while patent is a monopoly invasion of that right</cite>. He endorsed a “common-law copyright” reaching both books and inventions. One may think Rothbard’s contractual theory of copyright inadequate — it is, as we will see, vulnerable on third parties — but one may not quote him against copyright. The authority invoked for the conclusion held the opposite of the conclusion.

Second, Holmes. The post closes with the line from his dissent in International News Service v. Associated Press (1918): “Property, a creation of law, does not arise from value.” Two problems. It is a dissent; the majority recognised a quasi-property interest in the fruits of the wire service’s labour and enjoined the competitor. And the half-sentence relied upon contains a premise — property is a creation of law — that is fatal to the natural-rights edifice the post is defending. If property is whatever the legal order constitutes as property, then the legal order’s constitution of copyright is property on identical footing. A natural-law abolitionist cannot lean on a legal positivist’s epigram without sawing through the branch.

Third, Hoppe’s value argument itself, which supplies the post’s intellectual spine and which we take up directly in Part IV. Its conclusion is sound and uncontested; its relevance to intellectual property is nil, because it refutes a claim no serious proponent of intellectual property makes.

So the post’s method is: borrow a contested premise from authority; misread one authority who in fact disagrees; quote a second whose own words undercut the framework; and rest the spine on a third argument aimed at a target that is not there. None of that is yet a refutation on the merits. It is a description of why the merits were never reached. We now reach them.

A note on fairness, since this essay will insist on it throughout. Kinsella’s longer monograph, Against Intellectual Property (2001), is a genuine argued text, not an anthology, and it does confront the economic case head-on — it states the public-goods and free-rider argument accurately and then rejects the entire consequentialist frame on principled grounds. The charge of arguing-by-borrowed-authority is levelled at the post under examination, which is what we were asked to refute. The deeper claims of the monograph — the scarcity premise and the deontological objection that intellectual property is force against people using their own physical property — are addressed on their strongest form below, not dodged.

II. The debate is not about owning ideas

The opening move of nearly every anti-IP argument is to reduce intellectual property to the ownership of ideas, because once that reduction is granted the rest is downhill. No one owns the idea of a chair, or of love, or of a detective story, or of a payment network. Therefore — the argument runs — intellectual property is metaphysical extravagance.

It is theatre, and it survives only by refusing to read the statutes it condemns.

Intellectual property is not, and has never been, the ownership of ideas. Every operative category is defined precisely to exclude the floating abstraction the caricature attacks.

Copyright protects expression and withholds protection from ideas. The United States Code says so in terms: protection extends to original works of authorship, and in no case to “any idea, procedure, process, system, method of operation, concept, principle, or discovery” (17 U.S.C. § 102). This is not a local American quirk. The principle is written into the international architecture. The WIPO Copyright Treaty provides that <cite index=”73-1”>copyright protection extends to expressions and not to ideas, procedures, methods of operation or mathematical concepts as such</cite>, and Article 9.2 of the TRIPS Agreement says the identical thing. English law is the same: copyright subsists under the Copyright, Designs and Patents Act 1988, and as the Court of Appeal restated in the Da Vinci Code litigation, Baigent v The Random House Group (2007), <cite index=”77-1”>copyright does not subsist in ideas; it protects the expression of ideas, not the ideas themselves</cite>.

The distinction is old and judicially foundational, not a modern patch. In Baker v. Selden (1879), the Supreme Court held that copyright in a book explaining a bookkeeping system gave the author no exclusive right in the system itself; the method, as opposed to the prose describing it, was free to all, and could be captured if at all only by patent. The exclusive right attached to the expression; the idea ran free.

Patent law likewise does not protect ideas. It protects a claimed invention — a new and useful process, machine, manufacture, or composition of matter, or improvement thereof (35 U.S.C. § 101) — and only after the applicant satisfies novelty, non-obviousness, utility, and the disclosure requirement, for a bounded term. A vague thought claims nothing and receives nothing.

Trade-secret law protects neither ideas nor value as such, but specified information that derives independent economic value from not being generally known or readily ascertainable and that is the subject of reasonable measures to keep it secret, with liability tied to improper acquisition, disclosure, or use (18 U.S.C. § 1839). Lose the secrecy and you lose the right. Reach the same information honestly and you owe nothing.

Trademark protects a word, name, symbol, or device used to identify and distinguish goods and to indicate their source (15 U.S.C. § 1127) — source identity and goodwill, the prevention of consumer confusion — not “value,” and not language as such.

None of these gives anyone ownership of an idea in the broad, conversational sense the caricature requires. The author does not say “I own sadness”; he says “I wrote this novel.” The engineer does not say “I own motion”; she says “I disclosed this mechanism.” The merchant does not say “I own these words”; she says “this mark identifies my goods.” The anti-IP argument is rhetorically powerful exactly in proportion to its willingness to ignore this, because it must collapse expression into idea, invention into information, information into pattern, and pattern into a non-rivalrous abstraction, and then pronounce the abstraction unownable. That is not analysis of intellectual property. It is the substitution of a cartoon for the subject.

III. The value straw man

The Hoppe passage that anchors the post frames the issue as physical integrity versus value. You do not own the value of your property, because value is conferred by the changing valuations of others, over which you have no control. If a competitor opens a better restaurant and your custom falls, no right of yours is violated. If fashion turns and your stock is worthless, no one has robbed you. All of that is correct. It is also beside the point.

Intellectual property is not a property right in value. No serious proponent asserts that an author owns public taste, or that an inventor owns the profit he hoped for, or that a studio owns attention. Value is not the object of the right. Value is the consequence of the protected object. The author’s right is in the expression; the inventor’s in the claimed invention; the secret-holder’s in the confidential information; the mark-holder’s in source-identifying goodwill. Value is the reason the right is worth having. It is not the thing owned. The post repeatedly trades on the slide from “you cannot own value” to “you cannot own the thing whose value you are trying to protect,” and the slide is the whole trick.

The decisive illustration is the case the law has already decided. In Harper & Row v. Nation Enterprises (1985), a magazine obtained the unpublished manuscript of President Ford’s memoirs and rushed roughly three hundred words of his account of the Nixon pardon into print, timed to scoop the authorised serialisation that Time had paid for. Time cancelled. The Supreme Court held this was not fair use: by appropriating the unpublished expression, the magazine had, in the Court’s words, <cite index=”50-1”>effectively arrogated to itself the right of first publication, an important marketable subsidiary right</cite>. Observe what the wrong was not. Ford still had his manuscript. Harper & Row still had every word. Nothing physical had vanished. The loss was the appropriation of a specific, bounded, commercially decisive interest — the timing and exclusivity of first publication — which the copier could capture precisely because expression is reproducible. “You still have it” was true and irrelevant, which is the point.

Generalise the structure and the slogan collapses everywhere it is deployed. If a trade secret is leaked, the owner still knows it. If source code is copied, the developer still has the repository. If confidential bid data is disclosed, the firm still knows its numbers. If a manuscript is pirated, the author still holds the file. In each case the dispossession is not of the physical token but of control, exclusivity, secrecy, priority, timing, and the recoverability of organised investment. The landowner does not own “value” either; yet when a stranger builds an unauthorised road across his field we do not dismiss the trespass on the ground that market prices are subjective. The chattel owner does not own “value”; yet when a stranger takes his machine for a weekend and returns it we do not say the conversion is illusory because it left the metal intact. The object of the right is control; value is merely the evidence that control is worth contesting. Shout “value” at copyright and the entire apparatus is supposed to vanish. It does not vanish; it was never there.

IV. The equivocation on “property,” and the intangibles the theory cannot do without

Here is the deepest defect, and the one the post never confronts: its theory of rights is built, from the foundation up, out of exactly the kind of intangible it forbids others to own.

The argument depends on treating “property” as physical control over matter. But rights are not matter, and the system Kinsella defends is a system of rights. Consider its own vocabulary: ownership, title, contract, consent, aggression, transfer, homesteading, the boundary of a parcel, the claim, the remedy, liability, restitution, the right to exclude. Not one of these is a physical object. A deed is paper; the title is not the paper. A signature is ink; consent is not the ink. A fence is timber; the boundary is not reducible to the timber. A contract is a sheet of words; the obligation is not the cellulose. These are, in John Searle’s phrase, institutional facts — status functions that exist because a community of persons recognises and enforces them, not because they are inscribed on the physical world. They are non-rivalrous, non-physical, socially instituted normative relations. They are, in short, precisely what the post says cannot ground a right when the protected interest happens to be expression, invention, secrecy, or a mark.

The selectivity is the contradiction. The argument does not reject intangible rights. It accepts an entire apparatus of them whenever they protect land and chattels, and rejects them only when they protect creators. Title binds strangers; the post is content. Contract obligates a promisor who handed over no atom; the post is content. The homestead boundary is a normative line no one can see; the post is content. But let the same kind of bounded, enforceable, transferable normative relation attach to a novel or a mechanism and suddenly the theorist becomes a crude materialist who can recognise only what he can stub his toe against. That is not a principle. It is a preference wearing the costume of a principle.

The point becomes inescapable at the deepest layer. The post’s spine is Hoppe’s argument that one cannot own value because, while you control whether your acts change the physical properties of another’s thing, you do not control whether they change its value; a right whose boundary you cannot ascertain in advance is no right, since you would have to “interrogate the entire world population” before acting. Set aside that this argument, even if sound, refutes only ownership-of-market-price and is silent against ownership-of-a-bounded-expression. Notice what the argument is made of. It is the argumentation-ethics manoeuvre: to argue at all, Hoppe says, one must presuppose objective borders recognisable by everyone independently of subjective valuation, and the very act of arguing therefore proves those borders exist. But “the presuppositions of rational discourse” are an intangible, non-rivalrous, socially-constituted normative order. The entire metaethical foundation of the anti-IP position is the kind of non-physical, non-scarce, instituted thing the position declares cannot ground a right. A framework whose floor is made of the material it has outlawed cannot use the outlawing as a load-bearing wall.

And the ascertainability criterion — the very test Hoppe wields to kill value-rights — turns out to vindicate copyright. The objection to value-rights was that you cannot know, in advance, whether your act will cross the line, because the line depends on others’ valuations. Copyright has no such defect. A would-be copier can determine, in advance, with certainty, and in complete ignorance of anyone’s subjective valuation, whether the thing he is about to reproduce is another’s authored work. He does not poll the population. He checks whether he is copying. The boundary of a copyright is at least as objectively ascertainable as the boundary of a field — arguably more so, since the field’s edge is itself a contested normative construction (as the law of nuisance, trespass, and pollution thresholds attests, the line between a “physical invasion” and a mere externality is drawn by the legal order, not read off nature). The criterion offered to destroy intellectual property, applied honestly, protects it.

V. The false dichotomy and the missing third category

Hoppe’s value passage ends with a flourish: either the diminution of another’s property value is an actionable wrong — in which case all competition is aggression, which is absurd — or it is not, in which case there can be no right against it; “a third possibility does not exist.”

A third possibility exists, and the dichotomy survives only by suppressing it.

The argument treats all effects on another’s economic position as a single undifferentiated category called “value.” But the category is not single. There is competition by substitution — building a better or cheaper rival, inventing a superior product, persuading customers to switch — which lowers a competitor’s value and is no wrong, and which copyright and patent do not touch. And there is appropriation by reproduction — copying the specific expressive or inventive artifact another produced — which is the only thing intellectual property actually addresses. These are different acts with different objects, and the difference does not turn on “value” at all. The first competes against a producer’s product; the second reproduces that very product. Disaggregate them and Hoppe’s dilemma dissolves: one may consistently hold that there is no right against truthful, substitutive value-reduction and a right against the reproduction of one’s bounded work, on grounds that never mention subjective valuation. The “third possibility” the post declares non-existent is the category every intellectual-property system has always occupied: defined legal interests that are not reducible to anyone’s valuation. Copyright is not “my valuation of my book”; it is the protected expression. A patent is not “my valuation of my machine”; it is the claimed invention. The dichotomy is exhaustive only if you first delete the answer.

VI. Creation is not “mere rearrangement” — and copyright never claimed otherwise

A further stock move holds that creation generates no property because all production merely rearranges matter already owned; the author rearranges words, the engineer components, the chemist molecules, and since no one creates matter from nothing, no new right can arise. Hoppe and Rothbard are right that production is transformation rather than creation ex nihilo, and the point is worth conceding cleanly: mixing your labour with a thing you already own yields you a more valuable configuration of what you owned, not a fresh title conjured by the labour itself. Carve your stone into a statue and you own the statue because you owned the stone, not because creativity is a title-generating force.

The trouble is the inference, which is a non-sequitur in two directions.

First, “everything is rearrangement” proves too much. A house is rearranged timber and stone and steel; a microprocessor is rearranged silicon; a watch is rearranged metal and motion. If rearrangement defeated property, most of the economy would dissolve into dust. The economic question was never whether matter was rearranged. It is whether labour, judgment, risk, capital, and organisation produced a new and bounded good — and a finished book, a working mechanism, a maintained program, a verified database manifestly is one. A topic is not a book. “Write about Rome” is not Gibbon. “Write about whales” is not Melville. The topic is raw possibility; the book is the produced good, with structure, sequence, selection, evidence, prose, and a marketable form, and it converts an open field into a consumable work at real cost.

Second — and this is the part the post never registers — copyright does not in fact rest on the labour-desert theory the post is busy refuting. It rests on originality, and it has explicitly rejected reward-for-effort. In Feist Publications v. Rural Telephone Service (1991), the Supreme Court held that a telephone directory, however laborious its compilation, was not copyrightable, because facts are not original to anyone and effort is not the touchstone. The Court repudiated the “sweat of the brow” doctrine outright: <cite index=”48-1”>originality, not “sweat of the brow,” is the touchstone of copyright protection</cite>, and the purpose of copyright is to promote the progress of knowledge, not to reward labour as such. So when the post attacks intellectual property as a confused labour-theory of value applied to ideas, it is attacking a theory copyright disclaimed in 1991. Hoppe’s correct observation that labour alone is not a source of title is fired at a target that conceded the ground a generation ago. The error is doubled: the labour theory is rejected, and copyright never asserted it. What copyright protects is not the sweat but the original expression — a bounded, identifiable, transferable, infringeable object — and Feist is the proof that the category is disciplined enough to throw out the very effort-claims the post pretends are its essence.

VII. The economics the theory cannot represent

Now the matter the framework is constitutionally unable to see, because its ontology forbids it from treating production cost as relevant to rights: the appropriability problem.

Expressive and inventive goods characteristically combine high fixed costs of creation with low marginal costs of reproduction. The author spends months or years; the publisher pays editors, designers, lawyers, marketers, printers, distributors; the software firm pays developers, testers, security engineers, documentation teams; the pharmaceutical firm funds research, trials, and regulatory compliance. Once the good exists, copying it is cheap, and in digital form the marginal cost of an additional copy approaches zero. This is the textbook public-good structure of information, and it is the reason the economics of intellectual property is a serious field rather than a slogan.

Kenneth Arrow stated the core in 1962. The competitive market under-invests in the production of knowledge for three linked reasons — increasing returns, inappropriability, and uncertainty — and inappropriability is the decisive one: because information can be used without being used up, and revealed without being surrendered, its producer cannot capture its social value. Arrow named the paradox at the centre: the value of a piece of information cannot be judged by a buyer until he has it, but once he has it he has no need to pay. Sell the information by disclosing it and you have given it away; withhold it and you cannot sell it. A producer facing that structure, with no right of exclusion, under-produces relative to the social optimum.

Landes and Posner translated the point into the law of copyright in 1989. The distinguishing feature of intellectual property, they wrote, is its public-good aspect: the cost of creating the work is high, the cost of reproducing it low, and if the creator’s copies are priced near the marginal cost of reproduction, his total revenue may not cover the cost of creation — so copyright trades the costs of limiting access against the benefit of preserving an incentive to create the work in the first place. State the same logic as a ledger. Let C be the cost of creation, M the marginal cost of a copy, R the revenue a finished work can earn, and E the availability of exclusion. The originator’s expected profit is roughly R(E) − C; the copier’s is roughly R′ − M. Where M is near zero and C is substantial, and where E is absent, the copier’s position is structurally superior to the originator’s: the copier waits until the originator has borne C and resolved the uncertainty, then reproduces the resolved product at M and competes the price toward M, at which the originator cannot recover C. The asymmetry is not moral panic. It is arithmetic.

The crucial conceptual error the post commits — and it is fatal — is to read the low reproduction cost as proof that there is no protectable production cost. A book may be cheap to copy after it exists and expensive to create before it exists; an invention cheap to imitate after disclosure and expensive to discover before it; a database cheap to scrape after compilation and expensive to assemble before it. The copier always enters after the uncertainty has been resolved; the creator always acts before. To treat the cheapness of the copy as dispositive is to confuse the cost of reproduction with the cost of production. Those are different magnitudes, and the gap between them is the economic heart of the whole subject.

VIII. The honest qualifications, stated rather than hidden

A polemic that suppressed the strongest contrary evidence would be committing, against its readers, exactly the offence it charges against Kinsella. So the qualifications are placed in the open.

The empirical magnitude of under-provision is genuinely contested, and the contest is not frivolous. As early as 1934, Arnold Plant argued that copyright was less necessary than its defenders supposed, observing that authors are moved by much besides royalties and that nineteenth-century American publishers profited handsomely while freely reprinting uncopyrighted British works; he conceded only that some <cite index=”97-1”>authors write books because copyright exists, and a greater variety of books is published</cite>. Michele Boldrin and David Levine, in Against Intellectual Monopoly (2008) and related work, mount a comprehensive economic case that intellectual property functions as a costly and dangerous monopoly rather than a necessary spur, and that competitive returns to first-mover advantage and complementary sales often suffice. And the most thorough mid-century economic review, Fritz Machlup’s 1958 study for the United States Senate, reached a verdict of candid indeterminacy: no economist, on present knowledge, could say whether the patent system confers a net benefit or a net loss; if we had no patent system it would be irresponsible to recommend instituting one, and since we have long had one it would be irresponsible to recommend abolishing it.

Two things follow, and they cut in a single direction.

First, Machlup’s indeterminacy is far more damaging to Kinsella than to the defender of intellectual property, because Kinsella’s posture is confident abolition. If the careful economist cannot say whether the system is net-beneficial, then no one can claim its abolition is economically costless or obviously efficient. Indeterminacy is fatal to confidence, and the confidence here is all on the abolitionist side. The honest defender of intellectual property need claim only that there is a real appropriability problem that the institution addresses imperfectly; the abolitionist must claim that dismantling the institution is safe, and the evidence does not license that claim.

Second — and this is the cleaner point — Kinsella’s central argument does not rest on the empirics at all, and so cannot be saved by them. His deontological claim is that intellectual property is illegitimate in principle, because enforcing it means using force against people who are using their own physical property (their presses, their paper, their machines) in ways that invade no one physically. On that view the wealth consequences are simply irrelevant: rights trump utility, and even a regime that demonstrably increased social wealth would be unjust if it violated physical-property rights. Very well — then the refutation must be, and has been, at the level of the premise: the scarcity-only theory of legitimate ownership is a stipulative definition, not a demonstrated truth; it is self-undermining, because the theory’s own foundations are intangible normative relations; and it does not even demarcate the cases it wants, because “physical invasion” is itself a normative construction. The empirics are not the load-bearing objection. They enter for a different and narrower purpose, which is the subject of the next two parts: to show what Kinsella’s prescription actually produces in the world, and thereby to answer the rhetorical claim — the one piece of consequentialism the abolitionist does smuggle in — that abolition amounts to liberty.

And the concession to Hoppe and Rothbard is made without reservation, because a fair argument states what its opponents get right. Value is not ownable, and competition that lowers it is no wrong: conceded. Production is transformation rather than creation from nothing, and labour alone is not a source of title: conceded. Intellectual property, as actually administered, frequently overreaches — copyright terms are too long, many patents should never have issued, enforcement is often abusive: conceded, and urged. None of these concessions reaches the conclusion, because the conclusion was derived from a premise that is false, not from observations that are true.

IX. The demonstration: a world without authorial rights

Run the world the post recommends, and watch what happens to a working market.

An unknown author spends four years on a serious work of non-fiction. It required travel, archival access, interviews, expert review, subscriptions, editing, indexing, typesetting, and design. The author has no celebrity brand and no independent distribution channel. The value is in the book.

Before publication, a large production house obtains a copy. In Kinsella’s world there is no copyright, so there is no claim to make. The house breaks into nothing; it uses its own scanners, its own servers, its own presses, its own warehouses, its own advertising accounts. It produces a polished edition, buys distribution, lists the work across every platform, prices it below the originator, commissions an audiobook with hired narrators, and generates summaries, classroom editions, and translations. The author objects. The house replies, with perfect fidelity to the theory: “You still have your manuscript. We used our own property. We copied only non-rivalrous information. You do not own value.” Under the theory, and absent a contract or a physical trespass, that is the end of the matter.

The commercial injury, however, is total and specific, and the law has already named it. This is Harper & Row with the protection switched off. The interest destroyed is the right of first publication — “an important marketable subsidiary right,” in the Court’s phrase — together with the author’s pricing power, his licensing options, his bargaining position, and his capacity to recover the fixed cost C. The house bore none of C. It enters after the uncertainty is resolved, reproduces the resolved product at M, and undersells the only party who paid to create the market. “You still have it” pays no editor, no researcher, no printer, and no rent.

Now iterate across the economy, because the structure is general. A novelist releases a book; a platform clones a cheaper edition. A historian publishes a specialist study; a larger house repackages it. An educational author writes a textbook; a corporation bundles a copy with its software. A translator produces a translation at the cost of months; a distributor copies it on release. A composer records music; a service copies and monetises it. A photographer builds an archive; an agency scrapes and resells it. A developer ships a tool; a larger firm folds a clone into its suite. In each case the theory’s answer is identical and identically beside the point: the originator retains his physical copy.

And the advantages compound on one side. The production house has lower unit costs, established retail relationships, capital reserves, marketing teams, the ability to cross-subsidise and absorb losses, the reach to flood channels and capture search and placement, the option to exploit every derivative format — and, decisively, the ability to copy many authors at once and so diversify the residual risk that any single appropriated work underperforms. The originator has one advantage: the original act of creation. That is precisely the advantage the theory declines to protect.

The predictable equilibrium is not a flowering of liberated competition. It is a migration of production toward whoever holds complementary assets. Publication is delayed; more work is kept private; collaboration narrows because disclosure invites copying. Creators who can monetise something other than the work — celebrity, a platform, a salaried post, a consulting practice, a touring schedule, a patron — survive; the writer who has only the work does not. Expensive categories thin out first: investigative books, scholarly syntheses, technical manuals, large translations, anything whose recovery depends on exclusivity. Patronage returns, and with it the patron’s control over what gets written. And the cost of all this is invisible in the way Bastiat taught us to expect: the cheap copy is seen; the book that was never written, the invention never disclosed, the translation never commissioned, the archive never built, is not. A public domain is not enriched by works that were never made.

This is the outcome the post markets as freedom. It is the opposite of its advertised politics. The argument poses as anti-monopoly and delivers, in practice, the consolidation of the market in the hands of whoever already owns the machinery of reproduction. The man with the press wins; the author is told, generously, that he still has his file.

X. In rem rights are ordinary, and the slippery slope is already fenced

Two residual objections, briefly, because each is answered by the structure of property itself.

The post objects that intellectual property binds strangers — that someone who signed no contract may be enjoined from copying. So it does. But binding strangers is the ordinary architecture of property, not a pathology peculiar to intellectual property. Land title binds strangers. Chattel title binds strangers. Trespass and conversion bind strangers. Easements bind successors who never agreed to them. Security interests reach third parties. Passing-off and trademark restrain merchants who never dealt with the claimant. A right in rem is, by definition, a right good against the world; that is what distinguishes property from contract. The objection that intellectual property binds non-contracting strangers therefore reduces, once again, to the bare assertion that intellectual property is not “real” property — which is the conclusion, restated, not an argument for it. (It is also the precise pressure point on Rothbard’s contractual copyright, which tried to bind third parties through a theory of title-transfer and reservation; the difficulty is real, and it is a reason to ground exclusivity in a property-like in rem right rather than in contract — not a reason to deny the interest exists.)

The post also suggests that admitting intellectual property opens the floodgates: allow it, and every intangible becomes ownable. The premise is false, and its falsity is the category’s strength rather than its weakness. The system is already fenced, comprehensively and on purpose. Bare ideas are not ownable, nor facts, nor laws of nature, nor mathematical truths, nor general concepts, nor public-domain material, nor generic words. Independent creation is a complete answer to a charge of copyright infringement. Prior art and obviousness defeat patents; lack of utility and inadequate disclosure defeat them; loss of secrecy defeats trade secrets; lack of distinctiveness defeats trademarks; fair use and fair dealing limit copyright; exhaustion limits the distribution right; and every exclusive right expires. Feist threw out a laborious directory; Baker threw out a useful method; § 102(b) and the WIPO Copyright Treaty throw out ideas and procedures by their terms. A category defined by what it excludes is not incoherent; it is coherent, in exactly the way a boundary makes a parcel.

What the anti-IP argument does, finally, is judge intellectual property by its abuses while judging physical property by its ideal. Overlong copyright terms are said to disprove copyright; patent trolls to disprove patents; censorship-by-takedown to disprove authors’ rights. But abusive land seizures do not disprove land title, predatory litigation does not disprove contract, and malicious trespass claims do not disprove real property. We refine the rules; we do not declare the category metaphysical nonsense. The asymmetry is the tell. Held to a single standard, intellectual property is a bounded, limited, transferable, enforceable interest with a coherent object — which is all any property right has ever been.

XI. What the theory actually distributes

Strip away the vocabulary and ask what the theory allocates, because every theory of property is, in the end, an allocation of power.

The constitutional and economic purpose of copyright was never to enrich authors as an end. As the Supreme Court put it in Twentieth Century Music Corp. v. Aiken (1975), the statute strikes a balance — assuring the creator an adequate return for the value of his work while protecting the public from oppressive monopoly — because the immediate device of a private incentive serves an ultimate public end: the production and eventual dissemination of works that would otherwise not be made, or not be made as well. The bargain is criticisable at every margin — too long, too broad, too readily abused — and it should be criticised there. But it has a logic, and the logic is the appropriability problem: without some right of exclusion, certain works are not written, certain inventions are kept secret rather than disclosed, certain investments are not made, and the public domain that the abolitionist promises to enlarge is instead filled with absences.

Kinsella’s theory inverts this. It removes the right from the creator and confers the practical advantage on the better-capitalised copier, and it calls the transfer “liberty” on the ground that the copier used his own machines. But commerce is not the mere operation of machines. It is the organisation of effort, risk, trust, disclosure, timing, and recoverable return; and a theory of property that can see the press but not the authorship, the server but not the work, the warehouse but not the years, is not a theory of liberty. It is a theory of appropriation, optimised for the party who arrives after the cost has been paid by someone else.

XII. Conclusion: not liberty, but appropriation by scale

Kinsella’s argument looks rigorous because it opens with a truth: physical goods are rivalrous, and property in them exists to avert conflict over their use. The error is the inflation of that truth into a complete theory of rights. Scarcity of physical resources explains why chairs need owners. It does not explain why books, inventions, programs, designs, confidential information, and marks should be surrendered to whoever can copy them once they exist.

We can now say exactly where the argument fails. It does not defend its central premise; it borrows it from Hoppe and Rothbard and treats the borrowing as proof, while misreading Rothbard (who defended copyright) and leaning on a Holmes dissent whose premise dissolves the natural-rights frame. It strawmans intellectual property as the ownership of ideas, when the statutes, the treaties, and the cases define it as bounded expression, claimed invention, secrecy, and source — and exclude ideas, methods, facts, and effort. It strawmans intellectual property as the ownership of value, when value is the consequence and not the object, as Harper & Row shows by protecting first publication against a copier who left every physical token intact. It equivocates on “property,” accepting an entire apparatus of intangible juridical relations for land and chattels while declaring the same relations illegitimate for authorship — and its own deepest foundation, the norms of argumentation, is made of the very intangible it forbids. Its ascertainability test, honestly applied, licenses copyright rather than killing it. Its dichotomy is exhaustive only because it deletes the real category, the defined legal interest. And when the argument finally touches working commerce, it cannot represent the appropriability problem at all, and so cannot see that its prescription produces not the abolition of monopoly but its concentration — the migration of the market to whoever owns the means of reproduction.

The honest qualifications stand, and they do not save it. The empirical magnitude of under-provision is contested; conceded — and the contest, ending in Machlup’s indeterminacy, is fatal to confident abolition, not to the cautious defender. Some of what Hoppe and Rothbard say is correct; conceded — and none of it reaches the conclusion. The conclusion was never derived. It was asserted from a premise that begs the question, undermines itself, and, pressed into the real economy, becomes a charter for the man who already owns the press.

A theory of rights that cannot see the author, the inventor, the engineer, the translator, and the unknown writer against the machinery of capital is not, whatever it calls itself, a theory of liberty. It is a theory for the better-capitalised — and it defines its central problem out of existence rather than answering it, which is the one move no theory of property is permitted to make.

References

Primary text under examination

• Kinsella, N. Stephan. “Hoppe on Property Rights in Physical Integrity vs Value.” StephanKinsella.com, 12 June 2011.

• Kinsella, N. Stephan. “Against Intellectual Property.” Journal of Libertarian Studies 15, no. 2 (Spring 2001): 1–53. (Republished as a monograph, Auburn, AL: Ludwig von Mises Institute, 2008.)

Austrian / libertarian sources invoked or engaged

• Hoppe, Hans-Hermann. A Theory of Socialism and Capitalism. Boston: Kluwer, 1989, esp. 139–141.

• Hoppe, Hans-Hermann. The Economics and Ethics of Private Property: Studies in Political Economy and Philosophy. Auburn, AL: Ludwig von Mises Institute, 2006 [1993].

• Hoppe, Hans-Hermann, and Walter Block. “Property and Exploitation.” International Journal of Value-Based Management 15, no. 3 (2002): 225–236.

• Rothbard, Murray N. Man, Economy, and State, with Power and Market. Scholar’s ed., 2nd ed. Auburn, AL: Ludwig von Mises Institute, 2009 [1962], ch. 2 §12 and ch. 10 (patents and copyrights — defending copyright; opposing patents).

Economics of intellectual property

• Arrow, Kenneth J. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity: Economic and Social Factors, 609–626. Princeton: Princeton University Press (NBER), 1962. (Increasing returns, inappropriability, uncertainty; the disclosure paradox.)

• Landes, William M., and Richard A. Posner. “An Economic Analysis of Copyright Law.” Journal of Legal Studies 18, no. 2 (1989): 325–363. (Public-good structure; cost of expression versus cost of copies; the access/incentive trade-off.)

• Machlup, Fritz. An Economic Review of the Patent System. Study No. 15, Subcommittee on Patents, Trademarks, and Copyrights, Committee on the Judiciary, U.S. Senate. Washington: U.S. Govt. Printing Office, 1958, esp. 79–80. (Verdict of indeterminacy.)

• Plant, Arnold. “The Economic Aspects of Copyright in Books.” Economica, n.s., 1, no. 2 (1934): 167–195. (Skeptical case; honest counter-authority.)

• Plant, Arnold. “The Economic Theory Concerning Patents for Inventions.” Economica, n.s., 1, no. 1 (1934): 30–51.

• Boldrin, Michele, and David K. Levine. Against Intellectual Monopoly. Cambridge: Cambridge University Press, 2008. (Comprehensive economic case against IP; honest counter-authority.)

• Demsetz, Harold. “Toward a Theory of Property Rights.” American Economic Review 57, no. 2 (1967): 347–359. (Property rights as institutional responses to externality and appropriability.)

Philosophy of social institutions

• Searle, John R. The Construction of Social Reality. New York: Free Press, 1995. (Institutional facts and status functions — the framework for the “intangible juridical relations” argument.)

Statutes and treaties

• 17 U.S.C. § 102 (subject matter of copyright; § 102(b) excludes ideas, procedures, processes, systems, methods, concepts, principles, discoveries).

• 35 U.S.C. § 101 (inventions patentable).

• 18 U.S.C. § 1839 (trade-secret definition).

• 15 U.S.C. § 1127 (trademark definition).

• Copyright, Designs and Patents Act 1988 (UK), ss. 1, 16.

• WIPO Copyright Treaty (1996), Art. 2 (”Copyright protection extends to expressions and not to ideas, procedures, methods of operation or mathematical concepts as such”).

• Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS, 1994), Art. 9.2 (idem); Berne Convention (Paris Act, 1971).

Cases

• Baker v. Selden, 101 U.S. 99 (1879). (Copyright in a book describing a method gives no exclusive right in the method; idea/expression.)

• International News Service v. Associated Press, 248 U.S. 215 (1918) (Holmes, J., dissenting at 246 — “Property, a creation of law…”).

• Twentieth Century Music Corp. v. Aiken, 422 U.S. 151 (1975). (Balanced purpose of copyright: adequate return to the creator while protecting the public from oppressive monopoly.)

• Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 (1985). (Unpublished manuscript; right of first publication as “an important marketable subsidiary right”; not fair use.)

• Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). (Originality, not “sweat of the brow”; facts uncopyrightable; copyright rewards originality, not effort.)

• Baigent v The Random House Group Ltd, [2007] EWCA Civ 247; Designers Guild Ltd v Russell Williams (Textiles) Ltd, [2000] 1 WLR 2416 (UK; idea/expression and substantial-part doctrine).

Note on sources: statutory and treaty language is quoted from the instruments themselves. Where this essay states what a source argues (Arrow, Landes and Posner, Machlup, Plant, Boldrin and Levine, Rothbard, the cases), the claim reflects the verified content of that source, not a summary of an abstract. Two propositions are flagged as contested rather than settled: the empirical magnitude of creative under-provision absent IP, and the net welfare effect of the patent system — on both, the literature genuinely divides, and the essay relies on that division rather than papering over it.

Property is not the allocation of pre-existing matter but a designed system for securing produced values and the conflictable resources over which control can be exclusive. The argument that only rivalrous material objects can be owned is correct about pure information — which is non-rivalrous and unownable — yet false as a theory of property, because rivalrousness can be engineered. The recognition of rivalrous digital tokens as property under the Property (Digital Assets etc) Act 2025, Tulip Trading, and AA v Persons Unknown vindicates the theory’s grain of truth while refuting its overreach. The work of the mind is not parasitic on whoever owns the substrate; it is, far more often than not, the source of the value the substrate carries. The argument protects the warehouse and forgets the architect.

Keywords

Property theory · Intellectual property · Conflictability · Digital assets · Rivalrousness · Jurisprudence

A correspondent recently put the case against intellectual property about as sharply as it can be put. It is worth setting it out in full, because it is not a careless position. It is the distilled form of a serious and internally disciplined theory, and it deserves to be answered on its merits rather than caricatured:

“Ideas or information cannot be owned since they never exist as independent things but are just arrangements of features of material substrates which are already owned by someone by homesteading or contract. IP rights are in rem rights and thus they violate existing property rights. IP rights cannot be generated by contract. ‘Scarcity’ is not what is relevant, it is the existence of rivalrous or material objects over which there can be conflict — for such things, property rights rooted in original appropriation and contractual transfer fully determine who the owner is. … You cannot own someone else’s already owned material resource by running around and claiming ‘hey I came up with digital scarcity.’” @NSKinsella

I want to take this argument seriously, which means three things. First, conceding what is true in it, and a good deal is true. Second, isolating the precise points at which it fails — not by appeal to intuition, but by appeal to what property actually is and how property systems actually work. Third, and this is the part that distinguishes the present essay from the usual exchange of slogans, testing the argument against a body of law that has, in the last three years, adjudicated its central empirical claim. The argument says, in effect, that you cannot build a new ownable thing out of the immaterial — that “digital scarcity” is a contradiction in terms, a confused attempt to own an idea. The law of England and Wales has now decided otherwise, in considered terms, after the most thorough institutional review the subject has ever received. That decision is the most important evidence in the case, and the anti-IP theory, properly understood, does not survive contact with it.

The thesis of this essay is therefore narrow and, I hope, exact. The conflictability theory of property is a powerful and largely correct account of the physical baseline of ownership. It becomes false when it is promoted from a partial account into a complete one — when the true claim that pure, non-rivalrous information cannot be possessed is silently expanded into the false claim that only material objects can be owned, and when the genuine difficulty of owning an abstract idea is used to deny that intellectual production can be a primary source of value and of rights. The argument, in the end, protects the warehouse and forgets the mind. And on the one frontier where its strongest empirical premise could be tested — the digital — it has been refuted.

I. What the argument gets right

Begin with the concessions, because a critique that does not make them is not worth reading.

The argument’s foundation is the concept it variously calls scarcity, rivalrousness, or — in its most careful recent formulation — conflictability. The thought is this. Property rules are not arbitrary social furniture. They exist because human beings act in a world where some things cannot be used by two people at once, so that one person’s use excludes another’s, and conflict over those things is therefore possible and indeed inevitable. A condition of contestable control exists wherever a thing cannot be simultaneously held by rival claimants. Property is the institution that answers that problem: it assigns an owner, and by doing so converts a potential conflict into a determinate allocation, enabling cooperation, exchange, and the division of labour.

On this account, the relevant feature of a thing is not that it is rare but that it is rivalrous. The distinction is real and the theory is right to insist on it. A banana in a jungle where bananas hang from every branch is not scarce in the ordinary sense — there are plenty — yet it is conflictable in the technical sense, because if I pick this banana and you take it from me, I no longer have it. Rivalry, not abundance, is what calls property into being. And the corollary the theory draws is correct as far as it goes: information as such is not rivalrous in that sense. If I know a fact and tell it to you, we both now know it; my knowing is not diminished by your knowing; we are not in conflict over the fact. A melody, a theorem, a recipe, a method — considered purely as patterns that can sit in many minds at once — are not consumed by use and cannot be the subject of the kind of zero-sum contest that physical objects generate.

From this the theory draws several conclusions that are, in their narrowest form, simply true. You cannot own an abstract idea floating free of any determinate embodiment. The bare fact that you thought of something — that the pattern occurred in your mind before it occurred in another’s — does not, without more, generate a right enforceable against the rest of humanity. The slogan “good ideas are scarce, so of course they can be owned” really is an equivocation, trading on the colloquial sense of “scarce” (rare, valuable) to smuggle in a conclusion about the technical sense (rivalrous, conflictable) that does not follow. And the observation that much intellectual-property doctrine has overreached — patents of absurd breadth, copyright terms that have swollen to a century and more, litigation deployed to suppress rather than to protect — is not a fringe complaint but a mainstream one, shared by many who have no quarrel with property as such.

A critique that denied any of this would be dishonest. The conflictability theory is not a crank’s invention. It is a disciplined working-out of the insight that property is a response to conflict, and it is right that conflict, in the strict sense, arises only over rivalrous things. The error is not in the criterion. The error is in three moves that the theory makes after stating the criterion: a move of classification, a move of ontology, and a move of empirical assumption. Each is fatal, and the third has now been tested in court.

II. The classification error: “idea” is not the unit of analysis

The argument’s first move is to conduct the entire analysis at the highest possible level of abstraction — the level of “ideas or information” — and then to declare victory because nothing at that level can be owned. But “idea” is not the unit that any serious property claim attaches to. It is a deliberately frictionless abstraction chosen because it cannot bear ownership, and the choice does the work that the argument pretends its logic is doing.

Consider the gradient that the word “idea” flattens. There is a vague notion; a discovered fact; a method reduced to a concrete technical teaching; a manuscript; a body of source code; a circuit layout; a compiled database; a working machine; a protocol specification; a confidential design held in confidence; a sign used in trade to identify a maker. These are not the same kind of thing, and the legal systems of the world have never treated them as the same kind of thing. The protected object in a functioning intellectual-property regime is never “the idea.” It is an identified intellectual product — fixed, disclosed, implemented, or commercially deployed in a determinate form — and the right is never over another person’s brain or over the bare fact that someone has come to know something. It is over defined acts: copying, manufacture, publication, distribution, adaptation, or commercial exploitation of a delimited work or invention, under conditions the law specifies.

So “ideas cannot be owned” is not an answer to the question. It is an evasion of the prior question, which is one of classification: whether a particular intellectual output is sufficiently defined, original, useful, fixed, disclosed, or distinctive to warrant legal protection of specified uses. The argument refuses to ask that question. It collapses the manuscript, the machine, the protocol, and the trade mark back into “ideas,” and then announces that ideas are non-rivalrous. That is not analysis; it is category-flattening, and it works only because the flattened category was selected for its inability to be owned.

The point is not peculiar to intellectual property. The law everywhere distinguishes between a thing and the determinate rights, powers, and relations that may attach to it, and it does its real work at the level of the determinate, never at the level of the abstraction. A serious treatment of the question would have to descend from “ideas” to the actual subject matter and ask, of each, whether the conditions for protection are met. The argument’s strength is entirely a function of its refusal to descend.

III. The ontological error: reducing property to its material substrate proves too much

The argument’s second move is its deepest, and its most revealing. Information, it says, is “just arrangements of features of material substrates which are already owned by someone.” The implication is that because the pattern is not a free-standing lump of matter, it cannot be an object of property at all; the only owner in the picture is whoever owns the substrate.

This proves far too much, and the way it proves too much exposes the hidden premise driving the whole position.

By the same logic, a written contract is merely ink arranged on paper already owned by the stationer. A bank balance is merely a pattern of entries in a ledger owned by the bank. A share in a company is merely a record. A debt is merely an accounting relation. An easement is merely a legal relation concerning land owned by someone else. An option, a licence, a bond, a security interest, a covenant, a trust interest, a corporate right — every one of these would dissolve on the argument’s reasoning, because none of them is a free-standing physical object, and each is “just” an arrangement of features of substrates owned by others. No mature legal system operates this way, and none could. The commercial world runs on enforceable intangible claims, and an ontology that cannot accommodate a debt or a share is not a theory of property; it is a refusal of most of property.

This is the point at which the systemic, or relational, view of property — which is not a fringe doctrine but the considered consensus of modern legal theory — has to be stated plainly, because the argument depends on its being false. Property is not a thing. Property is a relationship between persons with respect to a thing. The Law Commission of England and Wales, in the most careful recent statement of the orthodoxy, puts it exactly so: the word “property” “describes a relationship between a person and a thing, and not the thing itself” (Law Com No 412, Digital Assets: Final Report (2023), and the companion Supplemental Report No 416 (2024), para 2.5). What the law recognises, under the heading of property, is a structured set of rights, powers, claims, priorities, and exclusions — the incidents of ownership, in the familiar analysis: the rights to use, to exclude, to the income, to the capital, to transfer, the immunity from expropriation, and the rest. Some of these relations concern tangible objects. A great many do not. Property is, as one influential formulation has it, less about the enjoyment of a thing than about control over access to it (a point the Law Commission adopts at Law Com No 412, para 5.1).

Once this is seen, the argument’s ontology is exposed as a smuggled absolute. The hidden premise is not the conflictability criterion at all; it is the dogma that only physical rivalry counts, only material objects can be owned, and therefore every intangible right is parasitic on whoever happens to own the matter beneath it. That dogma is never demonstrated in the argument. It is assumed, and then deployed to invalidate not merely intellectual property but — if its logic were honestly followed — the entire architecture of intangible commercial rights on which contract itself depends. Property is not physics. It is a normative and institutional order imposed upon the physical and social world, and the order routinely creates and protects valuable relations that are not chunks of matter. An argument that cannot say this without abolishing the chose in action has mistaken one province of property for the whole country.

IV. The error of inverting the mind and the matter

The third concession-and-correction goes to the moral foundation of the dispute, and it is here that the productive, or rights-of-the-mind, tradition makes its decisive contribution.

The argument pictures property as beginning with the grasping of material things: someone takes hold of an unowned stick or field or lump of metal, and ownership flows from that first physical seizure and from the contracts that subsequently move the matter around. On that picture, the originator of a design, a method, or an expression has done nothing that bears on ownership; the only acts that count are the physical appropriation of substrate and its contractual transfer.

But human wealth is not, in the main, produced by grasping. It is produced by thought — by reason applied to reality to reshape it for human purposes. This is not a sentimental flourish; it is the central fact about production that the materialist account omits. A machine is not merely metal; it is metal integrated by an act of design into something that does work it could not otherwise do. A medicine is not merely a quantity of chemicals; it is chemicals organised by an act of discovery into a compound that heals. A protocol is not merely a string of symbols; a book is not merely ink; a chip is not merely silicon. In each case the value resides overwhelmingly in the intellectual integration that renders the physical substrate useful, and that integration is the source of the value, not an afterthought to it. The mind is not ornamental to production. It is the engine of production.

From this the productive tradition draws a conclusion that the conflictability theory cannot accommodate: that the right to property is, at bottom, the right to keep and dispose of the product of one’s effort, and that since the distinctively human form of productive effort is the work of the mind, the products of intellectual effort have the strongest, not the weakest, claim to protection. To affirm property in physical production while denying it in intellectual production is to reward the last person to handle the matter and to strip the person whose design, method, or expression made the matter worth handling. It is to protect the warehouse and to ignore the mind that filled it.

The conflictability theorist has a reply, and it must be met rather than waved away. The reply is that the products of the mind, once expressed, are non-rivalrous: many people can use the design or the melody at once, so there is no conflict, so there is nothing for property to do. But this misidentifies what the productive theory asks the law to protect. It does not ask the law to assign ownership of an abstract pattern as if it were a field. It asks the law to secure to the creator the right to the specific product of his effort and to the defined commercial uses of it — to publication, manufacture, exploitation — against those who would appropriate the value of that effort without having borne its cost. The non-rivalry of the underlying pattern is precisely why a distinct body of doctrine is needed, with its own conditions and its own limits, rather than a simple extension of the law of chattels. Non-rivalry is the problem the doctrine is shaped around; it is not a proof that no doctrine can exist.

There is a genuine and unresolved tension here, and intellectual honesty requires that it be named rather than papered over. Reasonable people who accept the primacy of production in the creation of value still disagree about how far, and on what terms, the law should convert that primacy into exclusionary rights over reproducible patterns — about durations, about the line between idea and expression, about the proper scope of a claim. Those are hard questions, and the productive tradition does not settle them by fiat. But they are questions about the design of intellectual-property rights — their scope, term, standards, and remedies — not about whether intellectual creation can be a source of rights at all. The conflictability argument does not engage them. It forecloses the entire category by treating the work of the mind as legally parasitic on whoever owns the substrate, and that foreclosure is what the productive theory shows to be an inversion of the actual order of production.

V. The empirical error — and the verdict of the new law of digital assets

We come now to the argument’s strongest and most testable claim, and to the evidence that decides it.

Stripped to its core, the argument contains an empirical prediction masquerading as a conceptual truth. It says that you cannot bring a genuinely new ownable thing into existence out of the immaterial — that the digital realm contains only non-rivalrous information, that “digital scarcity” is therefore a contradiction, and that anyone who claims to have created a conflictable resource by software and rules is merely dressing up an attempt to own an idea. “You cannot own someone else’s already owned material resource,” the argument concludes, “by running around and claiming ‘hey I came up with digital scarcity.’”

This is the one place where the dispute is not a matter of competing intuitions but of fact, and the fact has now been found. Over the past several years the question whether a purely digital, ledger-recorded token can be an object of property rights — not by analogy, not as a courtesy, but as a matter of law — has been litigated, reviewed by a law-reform commission across two reports and a consultation running to many hundreds of pages, and settled by statute. The answer is yes. And the reasoning by which the answer was reached is, remarkably, a vindication of everything true in the conflictability theory and a refutation of the false extension built on top of it.

Consider what the law actually held. In AA v Persons Unknown [2019] EWHC 3556 (Comm), the Commercial Court held that a cryptoasset is property capable of supporting a proprietary claim, applying the orthodox criteria of property (definability, identifiability, capacity for assumption by third parties, and permanence) at [55]–[61]. The Court of Appeal confirmed the analysis in Tulip Trading Ltd v van der Laan [2023] EWCA Civ 83, holding at [24] that a cryptoasset “is property,” and doing so on a ground that the conflictability theorist will recognise as his own: that such a token is rivalrous, because “the holding of it by one person necessarily prevents another from holding that very thing at the same time.” Parliament then put the matter beyond doubt. The Property (Digital Assets etc) Act 2025, s 1, provides that a thing “is not prevented from being the object of personal property rights merely because it is neither a thing in possession nor a thing in action” — a deliberate statutory recognition that an immaterial thing, neither a chattel nor a mere claim, can be owned.

Now observe how carefully the Law Commission, whose draft this enacted, drew the line — and notice that it drew the same line the conflictability theory draws, only to the opposite practical effect. The Commission agreed, emphatically, that pure information is not property and cannot be possessed. It accepted the orthodox holding that intangible information is not susceptible to possession (the database case, Your Response Ltd v Datateam Business Media Ltd [2014] EWCA Civ 281, applying OBG Ltd v Allan [2007] UKHL 21). It went further and held, in terms the anti-IP theorist could have written, that a private key — the credential by which a token is controlled — “comprise[s] only information,” can be copied without limit, and is not the asset (Law Com No 416, paras 2.25–2.26). On the question of non-rivalrous information, in other words, the law and the conflictability theory are in complete agreement: information as such is non-rivalrous, non-excludable, and not ownable.

But then the Commission identified the very thing the anti-IP argument insists cannot exist: rivalrousness that is engineered. A crypto-token, it found, is “rivalrous by design,” its single-valid-state quality “contingent on the existence of robust technical authentication and validation mechanisms” that prevent the same unit from being spent twice (Law Com No 412, para 4.33). Rivalrousness, the Commission held, “acts as a useful separator between pure information and cryptoassets” (para 4.29): the token is rivalrous in exactly the respect in which a copied file is not, because the system’s rules make one person’s control of it exclude everyone else’s. The token “is not a record or representation of something; it is a distinct and independent thing” (para 4.19). And because it is rivalrous, excludable, and conflictable, it is — on the orthodox criteria, and now by statute — property.

The significance of this for the argument under examination is difficult to overstate. The conflictability theory says that property is for conflictable resources: things over which a condition of contestable control exists, where my control excludes yours. That is a precise description of an engineered-scarce digital token. Run the theory’s own test on such a token and it passes: only one person can validly control a given unit at a given time; one person’s control excludes another’s; the resource is rivalrous, conflictable, and therefore — by the theory’s own criterion — a proper object of property. The theory, applied honestly to the facts, does not yield “nobody can own a digital token.” It yields the opposite. The token is precisely the kind of thing the theory says property exists to allocate.

The error, then, was never in the conflictability criterion. It was in the empirical assumption silently bolted to it — the assumption that the digital world contains only non-rivalrous information, so that “digital scarcity” must be a confused metaphor. Engineering falsifies that assumption. Scarcity, in the technical sense the theory cares about, can be constructed: a system of cryptographic commitments, validation rules, and consensus over state can manufacture a resource that genuinely cannot be held by two parties at once. To insist, in the face of this, that “you cannot own someone else’s already owned material resource by claiming you came up with digital scarcity” is to answer a claim no one with a clear head is making. The claim is not that one owns the concept of digital scarcity, nor that one owns the computers of the world by having devised a rule-set. The claim is that a particular, engineered, rivalrous, controllable digital state is a conflictable resource and therefore an object of property — and that claim the law has now accepted, on reasoning that is the conflictability theory’s own.

Two cautions are owed here, in the interest of the rigour this essay has tried to maintain, and they cut in the argument’s favour as far as they honestly can.

First, the digital-token result does not, by itself, establish copyright or patent over non-rivalrous expressions and inventions. What it establishes is narrower and, for the argument, fatal in a different way: that the blanket proposition “only material objects can be owned” is simply false. An immaterial thing — neither a chattel nor a claim — can be, and now in law is, owned. That demolishes the ontological move of Part III directly and on authority. Whether the law should also protect non-rivalrous patterns, through copyright and patent, remains a distinct and genuinely harder question, on which the productive theory of Part IV does the argumentative work and on which honest people continue to disagree about scope and limit. The token case and the IP case should not be conflated; they are answered by different considerations, and only the first is settled by the new law. But the first is enough to dispose of the argument’s most sweeping premise.

Second, the result says nothing about who created any particular system, and nothing in this essay should be read as a claim about authorship or about ownership of any specific network. Indeed the law’s own logic points away from such claims: ownership of a token attaches to control of it, so the units are owned by those who control them, not by whoever first described the method; and a protocol, as an abstract method, is no more ownable as such than any other discovery, though a particular written specification or body of code may attract copyright in the ordinary way. The lesson of the new law is not that some person owns a network. It is that engineered digital scarcity is real, that the conflictable resources it produces are property, and that the argument’s prediction to the contrary has been tested and has failed.

VI. The remaining errors, briefly

Three further moves in the argument can be dealt with more shortly, because each is a variation on the errors already identified.

“IP rights are in rem and therefore violate existing property rights.” An in rem right is simply a right good against persons generally rather than against a single counterparty, and the legal system is replete with such rights that no one regards as violations of property: a landowner’s right excludes all the world; an easement binds successors; a registered charge follows the asset into third hands; a trust binds the conscience of those who receive trust property with notice; and a third-category digital asset, now, is good against the world save a person with a superior title (Law Com No 412, para 5.84). The question is never whether a right is in rem — pervasively, legitimate property rights are — but whether the legal order has good grounds for recognising the exclusionary structure, grounds typically furnished by public rules of registration, notice, statutory definition, fixation, or grant. The argument treats “in rem“ as a disqualifying incantation. It is nothing of the kind. To assume that only first possession of matter can generate a right good against the world is to assume precisely the conclusion in dispute.

“IP rights cannot be generated by contract.” In its narrow form this is true and unremarkable: a private agreement between two parties cannot, by itself, impose duties on strangers. But the argument converts a limitation into an abolition. That a contract cannot bind the world does not entail that intellectual property is impossible or unrelated to contract; it entails only that the universal element of an intellectual-property right comes from public law, while a vast portion of the doctrine’s actual operation runs through contract — licences, assignments, confidentiality undertakings, software terms, employment-invention clauses, research and publishing agreements, standards licences. And legal systems routinely attach proprietary or quasi-proprietary consequences to arrangements once there is notice, registration, or statutory authority. The argument’s own enthusiasm for contract is, moreover, quietly parasitic on the intangible infrastructure it disdains: contracts depend on enforceable claims, assignability, damages, injunctions, identity, and recognised legal objects, and a world in which only rivalrous matter counted would render most of contract unstable. The argument cannot consistently keep contract while discarding the legal abstractions that make contract reliable.

Appropriation by copying. The argument leans heavily on the observation that a copyist does not take the original object — the author keeps her manuscript, the inventor keeps his prototype. True, and beside the point. A property system concerned only with physical subtraction cannot explain a large class of injuries the law has long recognised: the destruction of an exclusivity, the collapse of a market, the misleading of consumers, the defeat of a disclosure bargain, the breach of a confidence, the conversion of another’s productive achievement into uncompensated advantage. The copyist does not remove the author’s paper, but he may appropriate the commercial value of the text; he does not seize the inventor’s prototype, but he may take the invention after its disclosure; he does not steal the merchant’s signboard, but he may use the mark to divert the trade. That these are real harms, and that a property theory which can see only the taking of lumps of matter cannot account for them, is itself a demonstration that property protects more than physical possession.

VII. What survives, and what does not

Let me state the balance as fairly as I can, because the point is not to win a debate by exaggeration.

The conflictability theory survives as what it always was: a rigorous and largely correct account of the physical baseline of property, and a genuinely useful corrective against lazy arguments for intellectual property that equivocate on the word “scarce.” It is right that property answers a problem of conflict, that conflict in the strict sense arises only over rivalrous things, that pure non-rivalrous information cannot be possessed, that one cannot own an abstract idea by being the first to think it, and that much intellectual-property doctrine has grown overbroad and abusive. Each of those propositions is true, and a critic who denies them forfeits the right to be taken seriously.

What does not survive is the promotion of that partial account into a total theory of property. The argument is right that non-rivalrous information is not ownable, and wrong that only material objects can be owned — a proposition now falsified, on authority, by the recognition of immaterial rivalrous tokens as property. It is right that conflict requires rivalrousness, and wrong that rivalrousness cannot be engineered — a proposition falsified by the deliberate construction of conflictable digital states. It is right that a contract cannot bind the world, and wrong that intellectual property is therefore incoherent. It is right that in rem rights require justification, and wrong that they are inherently suspect. It is right that “I thought of it” creates no right, and wrong that creation is therefore irrelevant to property, when production by the mind is the very source of most of the value that property protects.

On the specific provocation that prompted all this — the claim that one cannot own anything “by claiming I came up with digital scarcity” — the honest verdict cuts cleanly in two directions, and refuses both absolutisms. The abstract concept of digital scarcity is a discovery, and no more ownable as such than any other idea: on that the critic is right. But the rivalrous tokens that an engineered-scarce system produces are owned — by those who control them — which is why the contrary slogan, that “nobody owns” such things, is equally mistaken; the law is now explicit that they are property and that ownership tracks control. And a particular implementation — a specification, a body of code — may attract copyright in the ordinary way, on grounds the productive theory supports and the conflictability theory has no good answer to. The truth lies between the two flags planted at the extremes: neither “you own the idea and therefore the network” nor “nothing here can be owned,” but the disciplined middle that the new law has in fact adopted — control-based ownership of conflictable digital things, an open and contestable question about exclusive rights in non-rivalrous patterns, and no ownership of abstractions at all.

VIII. Conclusion: property is built, not found

Underneath the whole dispute lies a picture of property as something discovered — already complete once physical objects have been seized and traded, requiring of the law only that it ratify the brute facts of possession. That picture is false to the most elementary features of property law. Property systems are designed. They are institutional technologies that specify what counts as possession, transfer, title, priority, notice, fraud, abandonment, accession, agency, trust, security, registration, limitation, and remedy. There is no property without rule-making, and the rules are not read off the surface of matter; they are constructed by a legal order to secure produced values, allocate control, prevent wrongful appropriation, support investment, and make cooperation possible across time and distance.

Once that is admitted, the real question is no longer whether intellectual property appears in some imagined pre-legal state of nature — it does not, but neither does the chose in action, the registered company, the floating charge, or the digital token. The real question is whether a rational legal order has good reason to recognise defined rights in intellectual production. And the answer cannot be supplied by the materialist reduction the argument relies on, because that reduction, followed honestly, would abolish most of the intangible architecture of modern commerce along with intellectual property, and because it has now been refuted at its most testable point by a body of law that recognised engineered, immaterial scarcity as property on reasoning that is the conflictability theory’s own.

The argument fails, in the end, not because conflictability is the wrong criterion but because it is not the whole of property; not because non-rivalrous information can be possessed but because rivalrousness can be made; not because contract binds the world but because property is larger than contract; and most fundamentally because it treats the work of the mind as a parasite on whoever owns the substrate, when in truth the mind is, far more often than not, the source of the value the substrate carries. It is a theory that protects the warehouse and forgets the architect, the engineer, the author, and the inventor who made the warehouse worth protecting. A fuller account — older than the materialist one, and now better supported by the law — holds that property is a normative and institutional system for securing produced values and enabling rational human action across time, and that on such an account the products of the mind are not the weakest claimants to protection but, very often, the strongest.

Keywords

property theory; intellectual property; anti-IP; conflictability; rivalrousness; non-rivalrous information; engineered scarcity; digital scarcity; crypto-assets as property; third-category property; Property (Digital Assets etc) Act 2025; Law Commission digital assets; Tulip Trading; AA v Persons Unknown; in rem rights; property as a system of rights; productive theory of value; choses in action; possession and control

Keywords: Power laws, backfitting, overfitting, forecasting, philosophy of science, Bitcoin

I. The Seduction of the Perfect Fit

Man is the animal that cannot leave a pattern alone. Give him three points and he will draw a line; give him a line and he will call it a law; give him a law and he will sell you the future. The faculty that lets us find the lion in the grass and the harvest in the season is the same faculty that finds faces in clouds, conspiracies in coincidence, and prophecy in the random walk of a price. It is a magnificent instrument, and like most magnificent instruments it is most dangerous precisely where it is most confident.

Look at what we have done with the night sky. The stars are scattered by accident across distances so vast that no two in a constellation have anything to do with one another, and yet every culture that ever looked up drew the same kind of pictures: hunters, bears, scales, kings. We did not discover the Plough; we imposed it. The constellation is not in the heavens. It is in the eye. And the eye, having drawn the figure, insists that the figure was always there.

Move from the sky to the screen and the habit is unchanged. The chartist sees a head and shoulders where a statistician sees noise. The numerologist finds significance in a birth date. The forecaster rules a confident line through a decade of data and pronounces it destiny. In each case the procedure is identical: a pattern is detected, the detection is mistaken for a discovery, and the discovery is mistaken for a mechanism.

Here is the trouble. The mind evolved to detect patterns, and it was not penalised for detecting too many. The ancestor who mistook the wind for a predator wasted a sprint; the ancestor who mistook a predator for the wind was removed from the gene pool. We are the descendants of the nervous, and nervousness has a signature: it cries wolf. We see pattern where there is only chance, because in the long arithmetic of survival a false alarm is cheaper than a missed threat. The cost of that bargain is that we are constitutionally credulous about regularities.

Now place before such a creature a graph. Let a smooth line pass cleanly through the scattered points of the past, each observation sitting obediently near the curve. The effect is hypnotic. The line seems not merely to describe the points but to command them, as if the data had been trying all along to obey a rule it had not yet been told. The observer draws the only conclusion his wiring permits: if it explained the past so beautifully, it must govern the future.

That conclusion is false. It is the central falsehood this essay exists to dismantle. A curve that fits the past has demonstrated exactly one thing — that it fits the past — and the past is the one stretch of time about which prediction is impossible, because it has already happened. The fit is a record of obedience already rendered, not a promise of obedience to come.

The whole question of whether a model is science or theatre turns on a single distinction, and almost everyone gets it backwards. The distinction is not how well the model fits the data. It is whether the model predicted the data or merely accommodated it. A theory that announced the number before the number was known has staked something and won. A theory that was shaped, after the fact, until it matched the number already on the table has staked nothing and proved nothing. The fitted line is the most flattering of mirrors. It shows the past exactly as the past was, and asks to be congratulated for the resemblance.

II. Prediction and Retrodiction

A prediction is a statement about an observation that has not yet been made. A retrodiction is a statement about an observation that has already been made. The words are nearly twins and the things they name are nearly opposites, and the entire health of a science can be read in which of the two it lives by.

The difference is not pedantry; it is the difference between a wager and a memoir. When you predict, the world has not yet voted, and it may vote against you. The prediction can be wrong, and because it can be wrong, its coming true means something. When you retrodict, the world has already voted, the result is on the record, and you are free to compose a theory that the record cannot embarrass. A retrodiction cannot lose the bet, because the bet was placed after the race was run. This is why the past is so cheap and the future so dear. Anyone can be a prophet of yesterday.

Consider the case that did more than any other to fix this standard in the modern mind. In 1915 Einstein completed the general theory of relativity, and the theory said something specific and frightening: that light passing close to the Sun would be bent by a definite amount, nearly twice the deflection that Newton’s physics allowed. No one had measured it. There was no result to fit. The number existed only in the theory, daring the universe to contradict it. In 1919 an eclipse permitted the measurement, the starlight was seen to bend, and the deflection matched. That is prediction in its purest and most dangerous form: a claim made before the fact, falsifiable in principle, confirmed in practice. Einstein had wagered his theory against the sky, and the sky had paid out.

But the same episode contains a subtlety that almost every retelling flattens, and the subtlety is the real lesson. General relativity also accounted for an old anomaly in the orbit of Mercury, whose perihelion advanced by a tiny excess that Newtonian mechanics could not explain and that astronomers had known about, and puzzled over, since Le Verrier described it in the eighteen-fifties. The excess was not predicted in the temporal sense. It was already on the books. Strictly, relativity retrodicted it. And yet physicists rightly counted Mercury as powerful confirmation, every bit as serious as the eclipse. Why? Because the theory accounted for Mercury with no adjustable parameters. Einstein did not reach into the equations and tune a dial until the orbit came out right. The figure fell out of a structure that had been built for entirely independent reasons, with nothing left free to fudge. The fit was not purchased. It was forced.

There is the whole matter in a single contrast. Temporal order — before or after — is not the deepest thing. The deepest thing is whether the agreement was bought with free parameters. A retrodiction from a rigid theory that could not have been bent to fit is worth more than a prediction from a theory so limp it could have fitted anything. What damns the curve-fitter is not that he speaks after the event. It is that he speaks with a fistful of dials, and turns them until the past obeys, and then presents the obedience he manufactured as a discovery he made. Mercury was a retrodiction and it was science, because nothing was adjusted. The typical market “law” is a retrodiction and it is not science, because everything was.

Gregor Mendel’s peas make the same point from the other side. He proposed a mechanism — discrete inherited factors, combining by simple rules — and from the mechanism flowed ratios that one could go and count: three to one, nine to three to three to one. The numbers were consequences of the theory, not inputs to it. He did not observe the ratios and then invent factors to explain them; he posited the factors and the ratios came due. That is what it looks like when explanation runs in the honest direction, from cause to consequence, and then submits the consequence to the world for checking.

Set beside these the characteristic productions of the forecasting trades. The economic cycle theory that appears, with impeccable timing, after the crash, and shows that the crash was written in the stars all along. The chartists’ patterns, named and catalogued, every one of them discovered in hindsight and none of them able to tell you on Tuesday what Wednesday holds. The power-law overlay drawn across the price history of a speculative asset, fitted to the very observations it claims to explain, advertised as a glimpse of an iron necessity governing the years to come. Each of these is a retrodiction in costume. Each was shaped to fit a past already known, with parameters free to be chosen for exactly that purpose. Each explains everything and forecloses nothing.

Three voices stand behind this distinction, and it is worth naming them, because they disagreed about much and converged on this. Karl Popper made the asymmetry the centre of his philosophy: a theory earns its standing not by the confirmations it collects but by the refutations it survives, and a theory that forbids nothing — that is compatible with every possible observation — tells us nothing about the world. Milton Friedman, from a different tradition and with a different temperament, insisted that the only test of a theory worth the name is the accuracy of its predictions. There is an irony here we shall return to: Friedman’s famous indifference to whether a theory’s assumptions are realistic, pressed to its limit, licenses the very overfitting he would have despised, for if all that matters is the fit, the cynic will simply manufacture the fit. A man who cares only that the answer comes out right, and not at all about why, has built the back door through which the charlatan walks. And Richard Feynman, in his lecture on what he called cargo-cult science, drew the line in plain language: the first principle is that you must not fool yourself, and you are the easiest person to fool. The fitted curve is the most efficient instrument of self-deception ever drawn, because it wears the dispassionate face of mathematics while flattering the one hope the modeller cannot give up.

The proposition that organises everything that follows is therefore very simple, and it is merciless. Explaining what has already happened is easy; any sufficiently flexible scheme can do it, and the more flexible the scheme the more easily it is done. Predicting what has not yet happened is hard, and stays hard, and cannot be faked by anyone forced to commit before the result is in. Science is the enterprise that submits to the second standard. Everything that ducks it and lives on the first is, whatever its mathematics, a kind of storytelling — and we shall see that the mathematics makes the story more dangerous, not less.

III. The Mathematics of Backfitting

Why is the past so easy to fit? The answer is not psychological but mathematical, and once it is grasped the spell of the well-fitted curve is broken for good. The answer is parameters.

A parameter is a dial: a number in a model that you are free to choose. Give a model one dial and it can shift up and down. Give it two and it can shift and tilt. Give it enough and there is almost no shape it cannot be bent into. A straight line has two parameters and can pass exactly through any two points you like. A parabola has three and can pass exactly through any three. A polynomial of degree n has n plus one, and can be driven exactly through n plus one arbitrary points — through the closing prices, if you wish, of n plus one consecutive days, no matter how those prices jump and stagger. The fit will be perfect. It will also be worthless, because a curve threaded exactly through every past point has learned the noise along with the signal, and will lurch off into nonsense the moment it is asked about a point it has not already seen.

The most famous sentence in the literature of overfitting makes the case with more wit than a textbook can, and its provenance is itself a small parable. Freeman Dyson, late in his life, recounted a meeting with Enrico Fermi in 1953 at which Fermi demolished in a few minutes a research programme Dyson and his students had pursued for years. Fermi asked how many free parameters Dyson had used. Dyson said four. Fermi replied: “I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk.” With that, Dyson wrote, the conversation was over. Notice the chain of custody. The remark is von Neumann’s, reported by Fermi, recalled by Dyson, and printed in a journal half a century later. It is an aphorism about the manufacture of false authority that has itself acquired authority by being passed from one great name to the next — and, lest anyone think it mere wit, mathematicians have since shown that four complex parameters really do suffice to draw a passable elephant. The joke is a theorem. With enough freedom you can fit the beast; with a little more you can animate it; and you will have learned precisely nothing about elephants.

The discipline that studies this formally calls it the bias–variance trade-off, and the dry name conceals an iron law. A model that is too rigid has high bias: it cannot bend to the real structure in the data, and it is wrong in a stable, honest way. A model that is too flexible has high variance: it bends to everything, including the accidents, and so it is wrong in an unstable, flattering way — superb on the data it was fitted to, hopeless on the next sample, because the next sample carries different accidents. Every additional parameter buys a closer fit to the past at the cost of a looser grip on the future. Beyond a point — and the point comes quickly — you are no longer modelling the world. You are memorising your sample and mistaking recall for understanding. A model with as many parameters as data points does not explain the data; it merely repeats it back, the way a parrot that has learned a sentence has not learned the language.

And this is before we reach the subtler corruptions, the ones that do their work without a single dishonest intention. The first is data snooping, sometimes called data dredging: the practice of trying relationship after relationship against the same body of data until one of them, by the ordinary luck of large numbers, comes out looking significant. If you test twenty independent hypotheses at the usual threshold, you should expect one to clear the bar by chance alone, having nothing whatever to do with the world. Test a thousand, and you will harvest a field of impressive coincidences. The literature on calendar effects in stock returns — the January effect, the Monday effect, the turn-of-the-month effect — is in large part a monument to this: torture a long enough price series with enough candidate patterns and it will confess to anything, and the confession will not replicate.

The second is multiple testing’s near relation, the practice the field has lately learned to call p-hacking, and which Andrew Gelman has aptly named the garden of forking paths. Here the analyst does not even run a thousand explicit tests. He makes, instead, a long sequence of small choices — which observations to include, where to start the sample, how to bin the data, which outliers to discard as anomalous, which transformation to apply — and at each fork he takes, often in perfect good faith, the path that makes the picture a little cleaner. No single step is fraud. The cumulative effect is a result selected for its beauty rather than its truth, sculpted out of the degrees of freedom hidden in the analyst’s own discretion. The fitted curve that emerges from such a process is not a finding. It is the residue of a thousand quiet decisions, each made because it helped.

Put the pieces together and the moral is stark, and it runs exactly opposite to common intuition. The closer a model fits the history, the more suspicious we should be, not the less — for the closeness may be the symptom of disease rather than the sign of health. The more freedom a model is given, the more easily it conforms to the past; and the more easily it conforms to the past, the less its conformity can possibly tell us, because a model that could have fitted anything has told us nothing by fitting this. Flexibility and evidential value are not merely different. They are at war. Every dial you add to capture the past is a dial subtracted from your claim on the future. This is the arithmetic the curve-seller never shows you, and it is the whole of the matter.

IV. Power Laws: What They Are and Why People Misuse Them

Of all the curves fitted to history and sold as destiny, the power law has lately become the favourite, and it deserves a section of its own, because the abuse of it is so widespread and the genuine article so misunderstood.

A power law is a relationship of the form y equals a times x raised to a fixed power b — in symbols, y = ax^b. Its defining property is scale invariance: stretch the input by any factor and the output stretches by a constant factor of its own, regardless of where on the scale you began. This is the mathematical signature of a system with no characteristic size, no natural unit at which things change their behaviour. And it has a seductive visual fingerprint. Take logarithms of both axes and the power law becomes a straight line, its slope equal to the exponent. It is this last fact that has caused most of the trouble, for it means that anyone who can plot data on log-log axes and lay a ruler against it can persuade himself, and a credulous audience, that he has found a law of nature.

Power laws are real, and where they are real they are genuinely remarkable. The frequency of earthquakes falls off as a power of their magnitude — the Gutenberg–Richter relation — so that great quakes are rare and small ones common in a fixed mathematical proportion. The sizes of cities, ranked from largest to smallest, trace something close to a power law, the regularity Zipf made famous. Certain networks have degree distributions that follow power laws over a range. Some quantities in biology scale as powers of body mass. These are not numerological curiosities. They are clues to mechanism.

And that is the first thing the abusers forget: a power law is not a mystical universal that descends upon any sufficiently grand process. It is the fingerprint of a particular kind of generating mechanism, and only of that kind. Power laws arise from specific causes — from multiplicative growth checked by constraints; from preferential attachment, the rich-get-richer dynamic in which the already-large grow fastest; from systems poised at a critical point, balanced on the knife-edge between order and disorder; from certain optimisations under cost. Where one of these mechanisms operates, a power law is the expected and intelligible result. Where none of them has been shown to operate, a straight-ish line on a log-log plot is not evidence of a hidden law. It is a shape, and shapes are cheap.

This is the heart of the misuse. The market mystic reasons: the chart looks straight in log-log; therefore a power law governs the system; therefore the future is constrained to follow the line. Every step of that inference is unearned. The first error is empirical. A straight line on doubly logarithmic axes is one of the weakest patterns in all of statistics, because logarithms are merciful: they compress vast ranges into small spaces and flatter all manner of curves into apparent straightness. Over a limited span — and the price history of any modern speculative asset is a very limited span — a great many distributions masquerade convincingly as power laws. The log-normal distribution in particular, the natural result of any process built from many multiplied random factors, can imitate a power law over orders of magnitude so faithfully that the eye, and even a careless regression, cannot tell them apart. To distinguish a true power law from its impostors requires not a ruler but a great deal of data and a great deal of care.

It was precisely this care that Aaron Clauset, Cosma Shalizi, and Mark Newman brought to the question in their 2009 paper in SIAM Review, the most important corrective in the modern literature and one that ought to be read by everyone tempted to wave a log-log plot. Their argument was twofold and devastating. First, the method by which power laws are usually “found” — fitting a straight line to log-log data by least squares — is statistically unsound: it produces biased estimates of the exponent and, worse, gives no indication whatever of whether the data obey a power law at all, since it never asks the question. It will return a slope for data that are not remotely power-law distributed, and the slope will look just as authoritative as a real one. Second, when one does the work properly — estimating the exponent by maximum likelihood, measuring goodness of fit with a proper statistic, and, crucially, testing the power law against serious alternatives such as the log-normal and the stretched exponential — a great many of the power laws confidently claimed in the scientific literature simply evaporate. For dataset after dataset, the data are fit at least as well, and often better, by an alternative that no one had bothered to rule out. The lesson is not that power laws do not exist. It is that almost no one who claims one has earned the claim, because almost no one has done the test that could have refuted it.

The point has been made again and again by those who study these distributions for a living. Stumpf and Porter, writing in Science in 2012 under the deliberately blunt title “Critical Truths about Power Laws,” observed that a convincing demonstration of a power law requires far more than a plausible-looking straight line: it requires a statistically sound fit over a sufficient range and, ideally, an account of the mechanism that would produce it. Absent the mechanism, even a clean fit is merely descriptive — and the mechanisms behind several of the textbook examples remain contested to this day. Whether the famous “scale-free” networks are as common as a generation of enthusiasts believed is itself now seriously doubted. Even the canonical cases, in other words, are harder to establish than the popular accounts pretend. So when a forecaster takes a decade or so of prices, plots them on log-log axes, lays his ruler against the smear, and announces a law of the asset’s nature, he has not done the hard part. He has skipped it. He has performed the one procedure that the people who actually understand power laws identified, years ago, as the procedure that proves nothing.

V. The Bubble Problem

History has run the experiment we need, over and over, for four centuries, and it always returns the same verdict. During the inflation of a speculative bubble, prices rise with a smoothness and an apparent inevitability that invite exactly the mathematical regularities the modellers love; and after the bubble bursts, the regularity vanishes as though it had never been, because it never was a law — only the trajectory of a particular madness, fitted in flight.

Begin with the tulips, though we must begin honestly, for even here the story is cleaner than the facts. In the popular telling, the Dutch Republic of the sixteen-thirties lost its collective mind over flower bulbs, drove their prices to the cost of houses, and was ruined in the crash of 1637. The careful history — Anne Goldgar’s archival work is the modern correction — shows a mania more confined and a collapse less catastrophic than the legend insists; the ruin was largely a ruin of reputations and contracts, not of the economy. The qualification matters, and it matters in a way that serves our theme exactly: the smooth ascending curve of tulip prices is itself partly a retrofitted artefact, a tidy line drawn through a messier reality by people who wanted a tidy lesson. The bubble narrative, like the bubble model, is more orderly in the retelling than it ever was in the living.

The pattern, with that caution entered, recurs with grim reliability. In the railway mania of the eighteen-forties, British investors poured their savings into railway shares on the unanswerable logic that railways were the future — which they were — and that the shares must therefore rise without end — which they did not. During the boom the rise looked structural, almost a law of progress; in the bust it was revealed as an ordinary speculative overshoot. The South Sea Bubble of 1720 inflated the stock of a company whose actual trade never remotely justified its price, on a tide of official endorsement and contagious greed, and burst spectacularly. Isaac Newton is supposed to have lost a fortune in it and to have remarked that he could calculate the motions of the heavenly bodies but not the madness of men. The line is almost certainly apocryphal — which is itself a fitting detail, a manufactured quotation about manufactured value, passed down because it is too apt to discard. The Mississippi scheme convulsed France in the same years on the same principle. The dot-com boom at the turn of our own century bid the shares of companies with no earnings, and sometimes no revenue, to heights justified entirely by extrapolation of the recent past — until the extrapolation stopped. The housing bubble of the mid-2000s rested on the proposition, encoded in models of impressive sophistication, that house prices in aggregate did not fall: a proposition that was true in the sample to which the models had been fitted and false the moment it mattered.

Lay these episodes side by side and the common anatomy is unmistakable. During the expansion, the trend appears stable, the growth appears inevitable, and clean mathematical relationships emerge from the data and beg to be fitted. Analysts oblige. Lines are drawn, exponents estimated, channels and bands constructed, and each fits the run of prices to that point with gratifying precision. After the collapse, the relationship breaks, the line is overshot or undershot by a margin that no confidence band had contemplated, and the fitted law is quietly abandoned or, more often, quietly recalibrated to the wreckage so that it may be fitted afresh to the next ascent.

Here is the argument the whole section exists to establish, and it is short. The existence of a trend during a bubble does not establish a law. It describes the bubble. A bubble is, almost by definition, a stretch of faster-than-justified growth, and faster-than-justified growth will of course trace out a curve, and the curve can of course be fitted, and the fit will of course look impressive for as long as the bubble lasts. None of this is evidence of an underlying necessity. It is the mathematical shadow of a crowd inflating a price, and when the crowd reverses, the shadow goes with it. To fit a curve to the inflating phase and call it a law is to mistake the symptom for the cause, the trajectory for the engine, the weather of a particular folly for the climate of the world.

In fairness — and the fairness sharpens rather than blunts the point — there exists an honest way to bring power laws to bubbles, and it throws the dishonest way into relief. Didier Sornette and his collaborators have for years modelled bubbles as faster-than-exponential growth decorated with accelerating oscillations and tending toward a critical point at which the bubble becomes unstable. Whatever one makes of its track record, which is contested, the Sornette programme does the thing the mystics will not: it commits, in advance, to dated and falsifiable statements about when a regime is likely to end, and it accepts the scoreboard when the date arrives. That is the difference between a scientist using a power law and a salesman wielding one. The scientist sticks his neck out and lets the future cut it off if it will. The salesman draws his line through the past and, when the future declines to cooperate, draws a new one.

VI. Moving Goalposts and the Illusion of Validation

We come now to the mechanism by which a worthless model achieves immortality, and it is worth dwelling on, because it is the engine that keeps the whole enterprise running long after it should have died. A model that makes a fixed claim can be killed by a single contrary fact. A model whose claim is allowed to move can never be killed at all. Most market “laws” survive not because they are right but because their custodians retain, and freely exercise, the right to redefine what counts as success.

The methods of rescue are a small and predictable repertoire, and once you have seen them named you will see them everywhere. There is boundary adjustment: the price strays above the line, so the line acquires an upper band, and above the band a further band, until the “law” is no longer a line but a corridor wide enough to contain whatever the price does. There is confidence-interval inflation: the original prediction was a value, then it was a range, then it was a range so generous that no outcome could fall outside it, at which point the prediction predicts nothing while appearing to predict everything. There is parameter recalibration: the exponent that fitted the data through last year no longer fits, so the exponent is revised, and the revised model is presented as the same triumphant law rather than as the confession of failure that every recalibration actually is. There is start-date migration: begin the fit in 2011 and it works; when it stops working, begin it in 2012, or 2015, and announce that the “true” law was always the one anchored to the new origin. And there is the simple excision of the inconvenient: the observations that violate the model are reclassified as anomalies, special cases, exogenous shocks, this-time-is-different exceptions, and removed from the reckoning, so that the law is judged only on the data that flatter it.

Each of these moves has the same structure and the same cost. Each preserves the appearance of success, and each, in exactly the same motion, destroys whatever predictive content the model had. For the content of a prediction lies entirely in what it forbids. A claim that rules out nothing carries no information; a claim that rules out everything except what happened carries no information either, since it was fitted to what happened. Every widening of the band, every revision of the exponent, every shift of the origin, every quiet deletion of an embarrassing point, is a transfer of the model from the column of things that could have been wrong into the column of things that cannot be wrong — and a thing that cannot be wrong cannot be right, because in matters of fact rightness is just the property of having survived the chance of being wrong.

This is the insight that Popper made the cornerstone of his philosophy, and he came to it, by his own account, by watching exactly this kind of intellectual machinery at work. What struck him about the grand explanatory systems of his Vienna — he named Adler’s psychology, Marx’s theory of history, Freud’s psychoanalysis — was not that they explained too little but that they explained too much. Whatever a man did, the theory could account for it; the same act and its opposite were equally confirming; no conceivable observation was inconsistent with the doctrine. To its adherents this universal explanatory reach felt like overwhelming strength. Popper saw it for what it was: the absence of any empirical content at all. A theory compatible with every possible state of the world says nothing about which state obtains. The apparent strength was the disease. The power to explain anything is identical to the inability to predict something, and the two are not opposites but the same fact seen from two sides.

Imre Lakatos refined the diagnosis into a distinction we badly need here, between research programmes that are progressive and those that are degenerating. The test is not whether a programme ever meets an anomaly — all of them do — but what it does in response. A progressive programme answers an anomaly by proposing a modification that goes beyond the anomaly, that predicts some new and independent fact which then turns out to be true, so that the patch buys you more than it cost. A degenerating programme answers an anomaly by proposing a modification whose only function is to absorb that anomaly, that predicts nothing new, that exists solely to stop the bleeding. The serial recalibration of a fitted curve is the very type of the degenerating programme. Each adjustment is purely defensive. None of them ever yields a fresh, checkable prediction that pays its way. The model is not learning. It is retreating, in good order, forever.

State the principle in the bluntest possible form and nail it over the door of every forecasting shop: a model that cannot fail cannot succeed. The capacity to be refuted is not a weakness a theory must regrettably tolerate. It is the precise thing that makes a theory worth holding, the membership fee for the company of serious ideas. A model whose advocates have arranged matters so that no outcome can ever count against it has not been strengthened by their ingenuity. It has been quietly removed from the realm of knowledge and re-housed in the realm of faith, where it may comfort its believers but can no longer inform anyone.

VII. Why Backfitted Models Convince

If backfitted models are so empty, why are intelligent people so reliably taken in by them? Not because those people are stupid, but because the human apparatus of judgement is fitted with a series of biases that the fitted curve exploits with almost diabolical precision. The persuasiveness of these models is not an accident of presentation. It is the predictable output of well-documented flaws in how we weigh evidence.

Begin with confirmation bias, the deepest of them, the tendency to seek, notice, and remember what agrees with a belief we already hold and to overlook, discount, and forget what does not. Show a believer a year in which the curve held and you have given him a vivid confirmation he will treasure; show him a year in which it failed and he will find a reason the year does not count, and the reason will feel, to him, like rigour rather than rescue. The model is never tested in his mind, only celebrated, because the mind that holds it is built to gather evidence for the defence and to lose the evidence for the prosecution.

Compounding it is selection bias in the evidence itself, of which survivorship bias is the most instructive variety, and it carries a famous illustration. During the Second World War the statistician Abraham Wald was asked to advise on where to add armour to bombers, given a survey of the bullet holes in the planes that came back. The instinct was to armour where the holes clustered. Wald saw the trap: the survey covered only the survivors. The planes hit in those places had returned to be counted; the planes hit elsewhere — the engines, the cockpit — had not come back at all, and so left no holes in the data. The armour belonged precisely where the returning planes were unscathed. The lesson generalises to every forecasting record we are ever shown. We see the models and the modellers that survived, the calls that happened to come good, the curves still standing because they have not yet been overshot. The failures have crashed silently out of view. A presented track record is a record of survivors, and to reason from it as though it were the whole population is to armour the wings and lose the plane.

Authority bias does its part. A claim advanced by someone eminent, confident, and credentialed is granted a weight its content has not earned, and the curve-seller is typically all three. Availability does more: we judge the probability of a thing by how easily examples come to mind, and the vivid, oft-repeated success of a famous forecast is far more available than the diffuse, unremembered failures, so that the mind’s estimate of the model’s reliability is corrupted by the mere memorability of its hits — the bias Tversky and Kahneman identified and that Gigerenzer has spent a career teaching people to correct. We remember the prediction that landed and forget the hundred that did not, and from this asymmetric memory we construct a wholly false impression of a method’s power.

Over all of this presides what Taleb named the narrative fallacy: the irresistible human compulsion to wrap a sequence of events in a story, to convert the random into the inevitable by the simple act of telling it backward. After the fact, every crash has its clear cause, every boom its obvious logic, every price its reason — not because the causes were knowable in advance, but because the mind cannot tolerate a sequence without a plot and will supply one whether the world provides it or not. The fitted curve is the narrative fallacy rendered in mathematics. It tells the story of the past as a smooth necessity, and the smoothness is supplied by the teller.

And the mathematics is the final and most powerful seduction, for numbers wear the mask of objectivity. A story told in words invites argument; the same story told in equations, exponents, and goodness-of-fit statistics seems to have descended from some realm above human motive, impartial and exact. But a fitted curve is not an oracle. It is a rhetorical device with a lab coat on. The precision of its decimals says nothing about the truth of its claim; a meaningless quantity can be computed to ten places. When mathematics is used not to derive a consequence and submit it to test but to dress a backward-looking story in the costume of rigour, it has ceased to be a tool of inquiry and become an instrument of persuasion — and it is the more dangerous for being so widely mistaken for the opposite.

That none of this is idle theory was shown, at length and at cost, by Philip Tetlock’s long study of expert prediction, which followed the forecasts of hundreds of specialists across many years and many domains. The findings are humbling and ought to be famous. The average expert predicted the future about as well as a reasonable guess, and sometimes worse; and — the detail that bears directly on our subject — the experts who were most confident, most famous, and most committed to a single grand explanatory framework were among the least accurate of all, while doing the best business in attention. The market for forecasts, Tetlock’s work makes plain, does not reward being right. It rewards being vivid, confident, and memorable, which are exactly the qualities a well-fitted curve and a self-assured curve-seller possess in abundance, and exactly the qualities that have nothing to do with whether tomorrow will obey.

VIII. The Economics of Selling a Model

We must now be impolite and follow the money, because the persistence of empty models is not finally explained by cognitive bias alone. Bias explains why the audience is willing to be fooled. Incentive explains why there is always someone willing to do the fooling, and to keep doing it. A model is not only an instrument of analysis. It is a product, and behind every durable product stands an interest in its survival.

Consider what a successful model returns to the man who sells it. Followers, first — an audience that grows with each retelling. Attention, the currency of the age. Influence over the decisions of others, which is to say power. And beneath these, the harder rewards: consulting fees, subscription revenue, the capital that flows toward a confident forecaster, the speaking engagements, the standing of the seer. A model that has acquired a following has acquired, for its proprietor, an income and an identity. And no man with an income and an identity bound up in a thing is a disinterested judge of whether the thing is true.

Upton Sinclair put the principle once and for all: it is difficult to get a man to understand something when his salary depends on his not understanding it. The custodian of a popular model is in precisely that position with respect to the model’s falsity. Every instinct of self-interest pulls him toward preservation and away from refutation. When the data turn against the model, he is not moved, as a scientist is supposed to be moved, to ask whether the model is wrong; he is moved to find the adjustment that will save it, because the model is not merely his theory but his livelihood and his name. The recalibrations we catalogued earlier are not, at bottom, intellectual errors. They are the rational responses of a man defending an asset.

The structure is a textbook principal–agent problem. The forecaster, the agent, is rewarded for the appearance of insight, not for the substance of it, and the two come apart precisely when it matters. His audience, the principals, cannot easily tell a genuine edge from a lucky run or a flattering fit, and by the time the truth arrives the fees have been collected and the attention spent. The agent’s optimal strategy, given all this, is not to be right. It is to seem right for as long as possible and to have an exit ready when seeming fails — and a model whose success criteria can be quietly moved provides exactly that: a way to seem right indefinitely.

Reputational lock-in seals it. A forecaster who has staked his public identity on a model cannot abandon it without abandoning himself. The more loudly and the longer he has championed it, the more the model becomes load-bearing for his standing, and the more unthinkable its repudiation grows. He is chained to his own past confidence. To recant would be to confess that the followers were misled and the fees unearned, and few men will pay that price while any recalibration remains available to postpone it. So the model is defended past all reason, not in spite of the seller’s interests but in faithful service to them.

Keynes saw the general law behind this and stated it with his usual cool: worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally. The forecasting industry runs on that wisdom. The analyst who hedges, who issues the vague and unfalsifiable call, who stays with the herd, suffers nothing when he is wrong, because everyone was wrong together and no particular blame attaches. The analyst who commits to a sharp, checkable prediction and misses is exposed and remembered. The payoffs are therefore brutally asymmetric, and they select, with perfect efficiency, against precisely the behaviour that science requires. A bold, falsifiable forecast is a career risk with little private upside and large private downside; a fitted curve with movable goalposts is a career asset that pays whether or not the world cooperates. The rational forecaster, responding to these incentives, will make exactly the unfalsifiable, endlessly adjustable, retrospectively fitted claims we have been anatomising — not because he is a fool, but because he is not one. And because there is, in this industry, almost no penalty for being wrong — no licence revoked, no fee returned, no record kept by anyone but the obscure and the obsessive — the supply of confident curves is, and will remain, effectively unlimited. The wonder is not that so many worthless models are sold. The wonder is that anyone expected otherwise.

IX. A Bayesian Reckoning

It is time to make the argument precise, and the right instrument is Bayes’s theorem, because it forces into the open the question the curve-seller most wants to keep hidden: not “does the model fit?” but “how surprised should I be by this fit under each of the explanations available to me?” That is the only question that bears on belief, and the fit, taken alone, cannot answer it.

Bayesian reasoning, stripped to its bones, says that the credibility a piece of evidence lends a hypothesis depends on two quantities and their comparison across rivals. The first is the likelihood: how probable is this evidence if the hypothesis is true? The second is the prior: how probable was the hypothesis before this evidence arrived? And the decisive move is comparative — we do not ask how well the evidence fits our favoured hypothesis in isolation, which is the chronic error, but how much better or worse it fits ours than it fits the alternatives. Evidence supports a hypothesis only to the extent that it is more expected under that hypothesis than under its competitors. A fact equally expected under every explanation discriminates among none of them, however dramatic it looks.

Set the two relevant hypotheses against each other. Theory A holds that the asset’s price is governed by a genuine law, a stable mechanism that constrains its path and will go on constraining it. Theory B holds that the price is the output of a speculative process — a crowd of human beings buying because the price is rising and selling because it is falling, generating, as such crowds do, runs and trends and temporary regularities that persist for a while and then dissolve. Now ask the Bayesian question of the evidence on offer, which is a clean curve fitted to the price history. Under Theory A, such a fit is expected. But under Theory B it is also expected — entirely, unremarkably expected — because speculative processes routinely throw off clean-looking trends over their runs; producing a stretch of data that a curve fits well is one of the most ordinary things a bubble does. The likelihood of the observed fit is high under both hypotheses. The evidence is therefore almost worthless for telling them apart. The Bayes factor sits near one; the fit that the modeller presents as decisive is, in the only calculation that matters, nearly silent. If anything it tilts the wrong way, since over a short and recent window the mundane process has the easier time manufacturing a persuasive curve.

Then weigh the priors, and the case against the law steepens. Theory A is an extraordinary claim. It asserts that the price of a speculative asset — a quantity set, moment to moment, by the shifting beliefs and appetites of millions of people — is in the grip of a fixed mathematical necessity of the kind we otherwise find only in the deep regularities of physics. That is a very strong claim about the world, and it should carry a correspondingly low prior. Laplace stated the standard two centuries ago and Sagan made it a slogan: the weight of evidence required for an extraordinary claim must be proportioned to its strangeness; extraordinary claims demand extraordinary evidence. Historical fit is not extraordinary evidence. It is the most ordinary evidence there is — exactly what the unremarkable hypothesis predicts anyway. To overturn a low prior one needs evidence that the mundane alternative would be very unlikely to produce, and the fitted curve is the opposite of that: evidence the mundane alternative produces as a matter of course.

This is the same truth that Deborah Mayo has pressed under the banner of severe testing, and her formulation is the most useful single criterion in this whole essay. A hypothesis is corroborated by passing a test only if the test was severe — only if, were the hypothesis false, it would very probably have failed that test. A test the hypothesis would pass comfortably whether it were true or false is no test at all; clearing it tells us nothing. Fitting a curve to the data the curve was chosen to fit is the least severe procedure imaginable. The fit could not have failed, because the parameters were selected, after the fact, to guarantee it. A claim that has survived only such a non-test has survived nothing, and deserves not one degree of added belief. The proper response to “but look how well it fits” is therefore not admiration but a single question: under what competing account of this data would the fit have been surprising? If the answer is none — if the speculative, law-free hypothesis predicts the very same fit — then the fit, however lovely, is not evidence of a law. It is evidence that a curve can be drawn, which we never doubted.

X. What a Real Law Looks Like

It will sharpen everything if we set beside the counterfeit a portrait of the genuine article, so that the contrast can do its work. What does a real law look like — the kind of regularity that has earned the name, as against the kind that has merely borrowed it? A genuine law has, at least, the following marks.

1. It is predictive. It tells you something about cases you have not yet seen, and it told you before you saw them.

2. It is falsifiable. There is a clearly specifiable observation that, if it occurred, would show the law to be false — and the believers can tell you what it is.

3. It is stable. Its content does not have to be revised every time a new observation arrives; the same law, unaltered, keeps accounting for fresh cases.

4. It is mechanistic. It is connected to an understanding of why the regularity holds, not merely that it has held, so that the regularity is intelligible and not just observed.

5. It is robust to new data. It goes on working outside the range and the era in which it was first found, on data that played no part in its construction.

6. It is independent of parameter manipulation. Its success does not depend on tuning dials to each new dataset; it holds with its parameters fixed, or with no free parameters at all.

Measure the great laws against this list and they pass without strain. Newtonian gravitation did not merely fit the planetary motions known when it was framed; it predicted motions no one had yet seen. When the orbit of Uranus misbehaved, the law was not patched — it was trusted, and the trust was specific: Le Verrier and Adams used it to compute where an unseen planet must be to cause the discrepancy, and when the telescopes were pointed at the calculated spot, Neptune was there. The same law had already told Halley that his comet would return, decades ahead, and it did. These are predictions in the strong sense: claims about the unseen, made in advance from a fixed theory, that the world then confirmed. No dials were turned. The law stuck out its neck a century at a time and kept its head.

The laws of thermodynamics are sterner still, and they show another mark of the genuine: they forbid. The second law does not fit a curve to anything; it prohibits a class of outcomes outright — no engine shall convert heat wholly into work, no machine shall run forever on nothing. Every attempt to build a perpetual-motion machine has failed, and each failure is not an awkward anomaly to be explained away but a fresh confirmation of the prohibition. This is the signature that inverts the behaviour of the fitted curve. A real law grows stronger every time reality is given the chance to break it and declines to. A backfitted model grows weaker every time reality is consulted, which is why its custodians work so hard to keep reality from being consulted cleanly.

Two of the examples often grouped with these deserve a more careful word, because intellectual honesty forbids lumping different things under one triumphant heading. Information theory — Shannon’s — is frequently and rightly cited as a body of deep, predictive law, and so in a sense it is: his coding theorems fix exact limits on how much information a channel can carry and how far a message can be compressed, and those limits bind every communication system ever built, predicting with certainty what no engineering cleverness can exceed. But they are, strictly, mathematical theorems — propositions proved from definitions — rather than empirical laws wrung from observation in the manner of gravitation or thermodynamics. Their predictive force is real, but it is the force of proof, not of induction, and the distinction is worth keeping rather than blurring. Game-theoretic equilibria deserve a sharper caution still. As mathematics, the existence of equilibrium in broad classes of games is a theorem, secure. As a description of what actual human beings do, equilibrium is a model of variable and often disappointing accuracy; whether real agents play the predicted strategies is an empirical question whose answer is frequently no. To wave at “game-theory equilibria” as though they were laws of nature on the footing of the second law of thermodynamics would be to commit, in miniature, the very confusion this essay condemns — to mistake a framework that can be fitted to behaviour for a necessity that governs it. The honest taxonomy keeps three things apart: empirical laws that forbid and predict, like gravitation and thermodynamics; mathematical theorems that bind with the certainty of proof, like Shannon’s limits; and modelling frameworks whose empirical standing must be earned case by case, like equilibrium in economics. Only the first two are “laws” in the sense the curve-seller is trying to borrow, and he has earned neither.

The deepest contrast is the one already glimpsed and now made general. Genuine laws become stronger when challenged. General relativity was a rigid theory that risked everything on advance predictions, and a century of attempts to break it has instead extended it: the bending of light, the slowing of clocks in gravity, and in our own time the direct detection of gravitational waves by LIGO in 2015 and the imaging of a black hole’s shadow in 2019 — each a place the theory could have been wrong and was not, each a tightening rather than a loosening of our confidence. That is what corroboration looks like: a fixed claim, repeatedly exposed to the chance of refutation, repeatedly surviving. The fitted market model exhibits the exact opposite biography. It does not risk and survive; it accommodates and retreats. Each new observation, instead of testing it, occasions another adjustment of it. Where the real law is strengthened by confrontation, the counterfeit must be shielded from it. That asymmetry is not a detail. It is the whole difference between a law and a story, written in the way each behaves when reality is finally allowed into the room.

XI. The Test of Scientific Honesty

All of this can be compressed into a handful of questions — a short interrogation to which any model claiming the dignity of a law should be subjected, and which it should be made to answer before it is believed, funded, or followed. The questions are simple. The discomfort they produce in the wrong sort of model is the point.

1. Were the parameters fixed in advance, or chosen after the data were seen? A model whose dials were set before the fit risked something; a model whose dials were set to produce the fit risked nothing.

2. Was the hypothesis stated before the test, or assembled afterward to match what happened? Prediction registered ahead of the fact is evidence; explanation composed after it is decoration.

3. Can the model be falsified, and will its advocates say how? Ask them, plainly: what observation would prove this wrong? If no such observation exists, or none can be named, the model is not a claim about the world but an article of faith.

4. How many times have the parameters been changed? Each recalibration is a failure the model already suffered and survived only by being altered. A long history of revisions is a long confession.

5. How many failed versions preceded the one now on display? The graveyard is never shown. Ask to see it. A model is only as impressive as the count of its discarded predecessors allows it to be.

6. Has the model been tested out of sample — on genuinely held-back data, from a different period or domain, that played no part in fitting it? In-sample success is not evidence; it is the thing to be explained. Only out-of-sample success counts.

7. Can independent researchers, with no stake in the outcome, reproduce the result from the data and the stated method? A result that survives only in the hands of its proprietor is not a finding. It is a performance.

Where these questions can be answered well, a model has begun — only begun — to earn its standing. Where they cannot, the proper attitude is not hostility but a settled, unembarrassed scepticism. The burden lies on the claim, not on the doubter. A model that will not submit to this interrogation has told you, by its refusal, exactly what it is.

XII. A Curve Is Not a Law

The history of science is not, whatever the textbooks’ tidy diagrams suggest, the history of drawing lines through data. It is the history of discovering mechanisms — accounts of why the world behaves as it does — and then sending those accounts out, defenceless and specific, to be confronted by a reality with no obligation to be kind. The honour belongs not to the men who fitted the past most snugly but to the men who told the future most riskily and were not destroyed.

So when a fitted power law is laid across thirteen years of an asset’s prices and pronounced a law of its nature, an inevitability, a glimpse of the decades to come, the right response is to measure the claim against everything we have said. A power law fitted to thirteen years of observations is not evidence of inevitability. At the very most it is evidence that a mathematical curve can be drawn through thirteen years of observations — which was never in doubt, and which a great many false laws and a great many real bubbles would produce in exactly the same way. The fit establishes the drawability of the curve. It establishes nothing about the necessity of the path.

The decisive question is never whether a model explains yesterday. Every model explains yesterday; that is the one thing every model can do, and the worse the model the more effortlessly it does it, because flexibility and hindsight together can account for anything that has already occurred. The question is whether the model predicts tomorrow — and predicts it without revision, without the band quietly widening, the exponent quietly shifting, the start date quietly migrating, the inconvenient observation quietly removed. The moment a model survives only by adjusting its boundaries, recalibrating its parameters, redefining its criterion of success, or moving its point of origin, it has crossed the line that separates the two kinds of thing this essay has tried to keep apart. It has ceased to be a law and become a narrative — a story told backward, in the language of mathematics, about why what happened had to happen.

There lies the whole of it, and it can be put in a sentence. Science advances through prediction. Pseudoscience survives through explanation. The one stakes itself on the unknown and grows stronger when reality confirms it; the other feeds on the known and grows more elaborate each time reality refuses. They can look identical on the page — the same axes, the same exponents, the same confident decimals — and they are nonetheless opposites, divided by a question that has nothing to do with how the curve looks and everything to do with what it dared. Did it tell you the number before the number was known? Or did it wait, as the charlatan always waits, until the answer was safely on the table, and then draw its elegant line through the past, and ask you to mistake a memory for a prophecy?

A curve is not a law, any more than a portrait is a prophecy or an obituary is a life. It is a record of where the points have already fallen. Whether the next point will fall there too is a question the curve cannot answer and was never built to ask — and the surest sign that you are in the presence of a fraud rather than a finding is that the curve is offered as the answer all the same.

Sources and Further Reading

Box, G. E. P. (1976). “Science and Statistics.” Journal of the American Statistical Association, 71(356), 791–799. https://doi.org/10.1080/01621459.1976.10480949

Box, G. E. P., & Draper, N. R. (1987). Empirical Model-Building and Response Surfaces. New York: Wiley. (Source of “Essentially, all models are wrong, but some are useful,” p. 424; ISBN 0-471-81033-9.)

Broido, A. D., & Clauset, A. (2019). “Scale-free networks are rare.” Nature Communications, 10, 1017. https://doi.org/10.1038/s41467-019-08746-5

Clauset, A., Shalizi, C. R., & Newman, M. E. J. (2009). “Power-Law Distributions in Empirical Data.” SIAM Review, 51(4), 661–703. https://doi.org/10.1137/070710111

Dyson, F. J. (2004). “A Meeting with Enrico Fermi.” Nature, 427, 297. https://doi.org/10.1038/427297a (The source of von Neumann’s elephant remark, by way of Fermi.)

Feynman, R. P. (1974). “Cargo Cult Science.” Commencement address, California Institute of Technology; reprinted in Surely You’re Joking, Mr. Feynman! New York: W. W. Norton (1985).

Friedman, M. (1953). “The Methodology of Positive Economics.” In Essays in Positive Economics. Chicago: University of Chicago Press.

Gelman, A., & Loken, E. (2014). “The Statistical Crisis in Science.” American Scientist, 102(6), 460–465. https://doi.org/10.1511/2014.111.460

Gigerenzer, G. (2002). Reckoning with Risk: Learning to Live with Uncertainty. London: Penguin. (US edition: Calculated Risks: How to Know When Numbers Deceive You. New York: Simon & Schuster.)

Goldgar, A. (2007). Tulipmania: Money, Honor, and Knowledge in the Dutch Golden Age. Chicago: University of Chicago Press.

Keynes, J. M. (1936). The General Theory of Employment, Interest and Money. London: Macmillan. (The remark on failing conventionally is in ch. 12.)

Lakatos, I. (1970). “Falsification and the Methodology of Scientific Research Programmes.” In I. Lakatos & A. Musgrave (eds.), Criticism and the Growth of Knowledge (pp. 91–196). Cambridge: Cambridge University Press.

Mangel, M., & Samaniego, F. J. (1984). “Abraham Wald’s Work on Aircraft Survivability.” Journal of the American Statistical Association, 79(386), 259–267. https://doi.org/10.1080/01621459.1984.10478038

Mayer, J., Khairy, K., & Howard, J. (2010). “Drawing an Elephant with Four Complex Parameters.” American Journal of Physics, 78(6), 648–649. https://doi.org/10.1119/1.3254017

Mayo, D. G. (2018). Statistical Inference as Severe Testing: How to Get Beyond the Statistics Wars. Cambridge: Cambridge University Press.

Popper, K. R. (1959). The Logic of Scientific Discovery. London: Hutchinson. (German original Logik der Forschung, 1934.) See also Conjectures and Refutations. London: Routledge & Kegan Paul (1963).

Sagan, C. (1980). Cosmos. New York: Random House. (The “extraordinary claims require extraordinary evidence” formulation.)

Silver, N. (2012). The Signal and the Noise: Why So Many Predictions Fail — but Some Don’t. New York: Penguin Press.

Sinclair, U. (1935). I, Candidate for Governor: And How I Got Licked. (The remark on salary and understanding.)

Sornette, D. (2003). Why Stock Markets Crash: Critical Events in Complex Financial Systems. Princeton: Princeton University Press.

Stumpf, M. P. H., & Porter, M. A. (2012). “Critical Truths About Power Laws.” Science, 335(6069), 665–666. https://doi.org/10.1126/science.1216142

Taleb, N. N. (2001). Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets. New York: Texere.

Taleb, N. N. (2007). The Black Swan: The Impact of the Highly Improbable. New York: Random House.

Tetlock, P. E. (2005). Expert Political Judgment: How Good Is It? How Can We Know? Princeton: Princeton University Press.

Tversky, A., & Kahneman, D. (1974). “Judgment under Uncertainty: Heuristics and Biases.” Science, 185(4157), 1124–1131. https://doi.org/10.1126/science.185.4157.1124

Tags: backfitting; retrodiction; power laws; statistical inference; data mining; overfitting; predictive models; scientific method; financial bubbles; market forecasting; falsifiability; Bayesian inference; econometrics; model risk; speculative assets; scientific realism; forecasting.

A single hash trial forgets everything. The Nakamoto protocol forgets nothing — and a forthcoming paper proves that the most influential economic account of blockchain security has been resting on the difference.

Based on “Why Hash-Trial Memorylessness Does Not Extend to the Nakamoto Protocol,” by Craig S. Wright, forthcoming in the International Journal of Cryptocurrency Research, Vol. 6, Issue 1 (June 2026).

Every security model is a confession. It tells you, by what it chooses to leave out, which facts its author found inconvenient. The economics of blockchain trust is no exception, and its most consequential confession is a single word: memoryless. When a model forgets what the protocol actually remembers, it overstates the cost of trust and understates how long that trust can last. My forthcoming paper is about exactly that forgetting — where it came from, what it conceals, and what it costs to put right.

The claim is narrow and precise, and I want to state it without ornament before dressing it up. A single hash trial — one roll of the SHA-256 dice — is memoryless. The Nakamoto protocol is not. The first is a theorem. The second is also a theorem, and it is the one the field has been quietly ignoring.

Memoryless is a theorem, not an adjective

Begin with what the word means, because the entire argument turns on taking it literally.

A random waiting time is memoryless if its future is independent of its past. Put plainly: the probability that you wait at least a little longer, given that you have already waited a while, is exactly the same as the probability of waiting at least that little from a cold start. The clock has no recollection of how long it has already been running. A process that has kept you waiting an hour is, if it is memoryless, no more and no less likely to deliver in the next minute than one you just switched on.

This is not a figure of speech, and it is not a vague gesture at “randomness.” It is one of the most tightly characterised facts in all of probability theory. Among every continuous distribution that exists, exactly one is memoryless: the exponential. Among every distribution on the whole numbers, exactly one: the geometric. Feller proved it; every probability textbook since has repeated it. Memorylessness is a property with a single solution, not a label you may pin on any process that happens to involve chance. If a process is memoryless, it is exponential or geometric. If it is not exponential or geometric, it is not memoryless. There is no third option, and no room for “approximately, if you don’t look too hard.”

That precision is what makes the word so dangerous when it is used loosely — and in the economics of proof of work, it has been used very loosely indeed.

The assumption the whole field is built on

The most influential recent economic treatment of Bitcoin’s security is Eric Budish’s. He states, flatly, that Nakamoto trust is memoryless — that the system is as secure at any given instant as the quantity of mining effort supporting it at that instant, no more and no less — and he leans on precisely this property to draw a sharp line between proof of work and proof of stake, arguing that computational work is more memoryless than stake. The instantaneous flow of mining expenditure becomes, in his hands, a single sufficient statistic for security. One number tells you everything.

He is far from alone. The selfish-mining literature that began with Eyal and Sirer models block creation as a Poisson process and says outright that all block-creation events are driven by memoryless processes. That assumption is not decorative; it is load-bearing. It is the only reason a problem unfolding in continuous time can be collapsed into a tractable discrete Markov decision process, where the only thing that can change the attacker’s calculation is the arrival of a new block, and the mere passage of time changes nothing. The queueing models of confirmation time, the double-spend random walks, the Markov chains catalogued in the field’s own surveys — all of them rest on the same Poisson foundation.

And here is the awkward part: it had already been shown, empirically, not to hold. Bowden and colleagues demonstrated that a homogeneous Poisson process simply does not fit the observed pattern of block arrivals once difficulty is allowed to move, with a statistical test that rejects the assumption at better than one chance in a thousand. The reason they identified is the reason this paper formalises. The field has been building on sand and calling it bedrock.

There is nothing wrong with using an exponential model for the gaps between blocks within a single stretch of fixed difficulty. That is a local approximation, and for many purposes it is an excellent one. The error is in elevating that local convenience — without qualification — into a characterisation of the entire Nakamoto protocol as a probabilistic and economic system. The protocol is not the hash trial. It is the hash trial wrapped in feedback machinery, and the machinery has a memory.

Four levels, and where memory creeps in

The paper’s first contribution is to stop the equivocation by naming the thing precisely. “Mining is memoryless” is a sentence with no clear subject. Memoryless about what? The paper distinguishes four levels, from narrowest to broadest.

The first level is the individual hash trial: how many tries until you find a winning hash, at fixed difficulty. This is genuinely memoryless — a geometric random variable, the discrete twin of the exponential. The second level is the gap between blocks within a single difficulty epoch, with the rate held constant. This is memoryless too, but trivially so: it holds because difficulty is fixed by definition for the duration of the epoch. It is a tautology, not a discovery — and, crucially, it holds only because of the very assumption that the next level destroys.

The third level is the protocol’s full state as it moves across epochs — difficulty, block height, chain structure, the contents of the mempool. The fourth is the miner’s realised economic payoff over time — whether the blocks he found stay canonical, whether his rewards have matured, what his fees are worth. These last two are where the money lives, and these last two are where memorylessness dies.

The decisive mechanism: difficulty adjustment

Every 2,016 blocks — roughly a fortnight — the Bitcoin protocol stops and looks backward. It measures how long those blocks actually took to mine, and it resets the difficulty to drag the average block time back toward ten minutes. A fortnight that ran fast raises the difficulty; a fortnight that ran slow lowers it. The new difficulty is a deterministic function of the realised block times of the epoch that just ended.

That single feedback loop is fatal to the memoryless claim at the level that matters. The rate at which blocks arrive after a retarget depends, by construction and with certainty, on the history of arrivals before it. A memoryless process requires that the distribution of what comes next be independent of everything that came before. Difficulty adjustment guarantees the precise opposite. The protocol writes the recent past into the present, on a fixed schedule, by design.

So the cross-epoch process is not exponential at all. It is what the time-series literature, after Hamilton, calls a regime-switching process: a sequence of exponential regimes whose rates are reset, at known boundaries, as a function of what has already happened. The protocol does not forget the last fortnight. It remembers it exactly, and it prices it in at the next retarget. This is the paper’s central theorem, and once it is stated it is almost obvious — which is the usual fate of the things a field has trained itself not to see.

There is a complementary result here from recent work by Noda and colleagues, and the paper folds it in. Bitcoin’s difficulty algorithm is stable only when hash rate does not respond too elastically to changes in reward. Past a threshold, the same feedback loop that destroys memorylessness produces the oscillating boom-and-bust pattern economists call a cobweb. Difficulty and non-memorylessness, instability and history-dependence, all come from the same retargeting mechanism.

Three more kinds of memory

Difficulty adjustment is the decisive mechanism, but it is not the only one, and the paper is careful to show that any one of three further features would break the memoryless claim on its own.

The first is coinbase maturity. The reward in a freshly mined block cannot be spent for 100 blocks. Whether that reward is ever actually realised depends on whether the block survives the next sixteen-odd hours of competition — which depends entirely on what happens after it is found. The payoff carries a memory because it serves a probation.

The second is fees. A single hash is worth the same no matter when it lands, but a block is not. The longer it has been since the previous block, the more fee-paying transactions have accumulated in the mempool waiting to be included, and the more a new block is worth. Block value is an increasing function of elapsed time: the prize the miner is competing for literally grows while he waits. That is the exact opposite of memorylessness, and it becomes more important every year as the subsidy shrinks and fees become the main event.

The third is the heaviest-chain rule. Whether a block counts as part of the real history depends on every block found after it. A transaction buried under fifty confirmations is far safer than one buried under five, and that depth — the accumulated weight of work stacked on top, which is nothing but the protocol’s memory of effort — is precisely what determines the probability that anyone could still rewrite it.

Difficulty adjustment, coinbase maturity, fee accumulation, the heaviest chain: the Nakamoto protocol is a machine built almost entirely out of memory. Calling it memoryless was never a description of the protocol. It was a description of the one piece of it simple enough to be convenient.

The scalar becomes a vector

What does this do to the economics? It changes the object you have to track.

Budish’s model compresses the security of the entire system into a single number — the flow of mining expenditure at a given instant. Security equals current trust support, full stop. The correction this paper forces is that the honest scalar must become a vector. To know whether a double-spend attack will succeed, you cannot watch one quantity; you must watch several at once: the work gap between the attacker’s hidden chain and the public one, the current difficulty on each chain, how far each chain sits from its next retarget, how many confirmations the target transaction has accumulated, and how many fees are sitting on the table. Set difficulty to a constant, delete the retargets, and zero out the fees, and this vector collapses cleanly back to Budish’s single number — which is exactly why his result is not wrong, only local.

The mining decision stops being a static, one-shot, zero-profit comparison and becomes a dynamic investment problem of the kind Ericson and Pakes formalised three decades ago: a question about the entire future path of difficulty, rewards, and competition, not a snapshot of this instant. Budish wrote down the snapshot. This paper supplies the film.

The free attack was always local

Budish’s most arresting result — his Theorem 2 — is that the net cost of a majority attack is zero. The attacker simply mines, collects the same block rewards an honest miner would have, pays the same costs, and so the attack is, in expectation, free. It is a genuinely uncomfortable claim, and within its proper domain it is correct. That domain, this paper shows, is a single fee-free epoch. Step one inch outside it and the zero evaporates.

It evaporates for two reasons, and the second is the more elegant.

The first reason is fees. An attacker, by definition, mines faster than the honest network — that is how he overtakes it. Mining faster means shorter gaps between his blocks, which means fewer accumulated fees per block, which means he collects strictly less than the steady-state reward against which his costs were calibrated. He pays for a full reward and receives a diminished one. The shortfall is modest per block, but it is positive, and it compounds across the attack.

The second reason is the difficulty adjustment turning on him. The attacker’s extra hash power makes the epoch run fast. At the next retarget — automatic, deterministic, and unstoppable unless he abandons the attack entirely — the protocol punishes that speed by raising the difficulty. Now every hash he throws earns less than it costs him. He is mining at a guaranteed per-hash loss, and he cannot opt out without surrendering the chain he has spent so much to build. This is the crucial difference from the external costs Budish does acknowledge — depreciating hardware, a collapsing coin price, legal exposure. Those are contingent, and an attacker might dodge them. This one is written directly into the consensus rules. It is the protocol’s own memory, weaponised against the attacker, and it cannot be bargained with.

And here is the finding that ought to give the lawless-attack literature genuine pause: honest miners leaving does not help the attacker. The naive intuition says that if everyone else quits, the attacker simply inherits the network. The arithmetic says the reverse. The fewer honest miners remain, the longer the attacker’s epoch runs, the higher the protocol sets the next difficulty, and the deeper his per-hash loss becomes. The paper proves the attack cost rises monotonically as honest participation falls. Every honest miner who walks away makes the attack more expensive, not less. Capitulation is a weapon that fires backward.

There is a further, sharper twist. Because a large attack depresses honest miners’ revenue and pushes the marginal ones toward shutting down, it can itself tip the difficulty algorithm past the elasticity threshold into Noda’s oscillating cobweb regime — making the difficulty environment swing violently and raising the attacker’s costs further still. The attack destabilises the protocol, and the destabilised protocol punishes the attack. It is a self-reinforcing trap of the attacker’s own making.

Zero dollars, or a billion

The paper does not leave this as algebra. It runs the two models head to head — ten thousand simulated attack paths each, on identical random draws, for an attacker controlling a bare majority of the network and sustaining the attack across five fortnightly epochs.

Under Budish’s constant-difficulty assumption, the simulation reproduces his theorem exactly: the mean net cost of the attack is essentially zero — minus four hundred thousand dollars, statistically indistinguishable from free — with 49.1 percent of paths showing a profit and the rest a loss. A coin flip.

Under the corrected model, with the retargets simply switched on and nothing else changed, the mean net cost is $1.275 billion, and every single one of the ten thousand paths costs the attacker more than $1.2 billion. The difficulty adjustment installs a deterministic floor beneath the cost that no run of luck can dig under. The same randomness, the same attacker, one missing mechanism — and the difference between free and a billion dollars.

The size of the gap scales with how long the attack runs and how fee-heavy the system is. For an attack that completes inside a single epoch at today’s fee levels, the understatement is modest — about $1.6 million — which is precisely why Budish’s short-horizon arithmetic remains a perfectly good local approximation, and the paper says so without grudging. But for the sustained attacks Budish himself identifies as the real threat to a payment system, the gap runs from hundreds of millions to, in a fee-dominated future, tens of billions of dollars a year. The benchmark is not demolished. It is confined to the short run and the fee-free corner, and told plainly where its writ ends.

The Poisson approximation has a shelf life

All of that concerns the protocol and the payoff. The paper’s final movement returns to the hash trial itself — the one level where memorylessness genuinely holds — and shows that even that is living on borrowed time.

The reason is a detail almost every model waves away: the search space is finite. A miner varies a few fields in the block header — a four-byte nonce and an eight-byte field in the coinbase — which yields roughly two-to-the-ninety-sixth distinct inputs to try. As long as the network samples a vanishing fraction of that space before someone finds a block, drawing without replacement is indistinguishable from drawing with replacement, and the geometric, memoryless model is exact for all practical purposes. Today the global network explores about six ten-thousandths of one percent of that space per block. The approximation is pristine.

But hash rate grows, and the explored fraction grows with it. Parra-Moyano and colleagues first pointed this out; this paper puts dates on it. On a long-run historical growth rate — hash rate doubling every eighteen months — the network explores about one percent of the space per block by around 2041, at which point the departure from memorylessness becomes detectable inside a single block’s contest. By about 2045 it reaches the five-percent mark, the standard statistical threshold past which sampling without replacement can no longer be passed off as sampling with replacement. By about 2050 it is exploring more than half the space per block, and the memoryless model fails outright — at the hash-trial level, the one floor everyone assumed was solid. The whole building gives way at once.

Throughput is a design choice, and it sets the expiry date

Here the paper turns a probability result into an economic argument, and the argument is the most important thing in it.

That two-to-the-ninety-sixth search space is the space for a fixed block template. But the template is not fixed. Every time the set of transactions in the candidate block changes, the Merkle root changes, the header changes, and the miner is suddenly searching a completely fresh region of the hash function — a brand-new urn of two-to-the-ninety-sixth balls. And what changes the transaction set? Incoming transactions. Every transaction that arrives refreshes the template.

So the effective search space is not the fixed domain. It is that domain multiplied by the number of distinct templates a miner cycles through while mining a single block — and that number is governed by transaction throughput. A protocol that processes four transactions per second refreshes the template a few thousand times per block. A protocol that processes a billion transactions per second refreshes it six hundred billion times per block: the urn is swapped out faster than any miner could sample a measurable sliver of it.

This is the hinge of the whole paper. Throughput is not a market outcome handed to the protocol by the world. It is a design parameter, fixed by the block-size policy and the engineering of the node software. And it determines when the entire stochastic foundation of blockchain security economics expires.

The numbers are stark. A throughput-constrained protocol running at four transactions per second — and the paper labels this case BTC — reaches the memoryless breakdown around 2061. A high-throughput protocol running at a billion transactions per second pushes that horizon past 2103; at ten billion, past 2108. The relationship is logarithmic, but the effect is enormous: more than four decades of additional model validity, bought purely by the decision to let the protocol scale. The constrained protocol and the unconstrained one face exactly the same physics of hardware growth. They differ only in whether their designers allowed throughput to rise. One design ages out within the operational lifetime of mining hardware bought today. The other does not.

And these are not thought-experiment throughputs. The paper notes that a UTXO transaction processor of this kind decomposes into independent services that scale horizontally — add machines, add throughput — and reports that independent security verification has confirmed sustained throughput above seventy thousand transactions per second on a single production node, with internal engineering measurements running above a billion per second per node and into the tens of billions across a modest cluster. The demand to fill that capacity is the demand of machines paying machines: micropayments, sensor networks, supply-chain settlement, high-frequency data integrity — oceans of individually trivial transactions. High throughput, on this account, is not a luxury or an ideological preference. It is the condition under which proof-of-work security remains mathematically coherent into the next century.

The security budget after the subsidy dies

Throughput does a second job, and this one bites long before 2061.

Bitcoin’s block subsidy halves every four years, marching steadily toward zero. By 2050 it will be worth a few thousand dollars a block, and the security of the system will have to be paid for almost entirely out of transaction fees. Budish’s equilibrium condition — that the reward per block must stay large relative to the value an attacker could steal — has to hold continuously, subsidy or no subsidy. The question is whether it can.

The answer, once again, is set by throughput. To keep today’s level of security funded in 2050, a four-transaction-per-second protocol would need to charge roughly $130 per transaction. That is not a payment network; it is a toll booth that prices out everyone it was built to serve, drives users away, and in driving them away strips out the very fees and hash rate that were holding the system up. A protocol processing a million transactions per second funds the identical security budget at five hundredths of a cent per transaction; at a billion per second, a rounding error. Same security, same equilibrium condition — but one design can pay for it through sheer volume, and the other cannot pay for it at all without strangling itself.

The paper calls the constrained protocol’s predicament a compounding trap, and the description is exact. Low throughput limits template refresh, which hastens the memoryless breakdown. Low throughput caps fee volume, which starves the security budget as the subsidy fades. And the high per-transaction fees that low throughput forces drive users out, which lowers hash rate, which hastens the breakdown again. Each problem feeds the others. A high-throughput protocol escapes the entire spiral: volume refreshes the templates, volume funds the security budget, and low per-transaction cost keeps users — and therefore the hash rate that secures them — inside the system.

What survives, and what does not

It is worth being exact about what this paper does and does not do to Budish, because precision is the whole point, and an argument that overreaches deserves to be dismissed.

It does not refute him. His central equilibrium logic — that permissionless, anonymous, free-entry proof of work requires a flow of honest expenditure large relative to the value of an attack — does not depend on any Poisson machinery at all, and it survives the correction completely intact. His short-horizon attack formulas remain good local approximations. What the paper removes is narrow and surgical: the unqualified claim that the protocol-level environment is memoryless, and the specific corollary that a majority attack is free. The first is replaced by a regime-switching process. The second is demoted from a universal truth to a local one — valid in a single fee-free epoch and nowhere else. The right summary statistic for the security of the system was never a scalar. It was always a state vector.

Why it matters

The deepest claim in the paper is the one in its final line: the stochastic foundations of proof-of-work security are themselves functions of protocol design. This is not a small thing to say. It means the probability theory underneath every economic model of Bitcoin — the selfish-mining analyses, the confirmation-time queues, the double-spend random walks, Budish’s own equilibrium — is not a fixed law of nature handed down by the mathematics. It has a shelf life, and the length of that shelf life is a choice.

A protocol that refuses to scale is not merely making a throughput decision. It is dating the expiry of its own security model, and, as the subsidy fades, signing up for a security budget it cannot fund. A protocol that scales is buying decades of model validity and a fee base that actually works. The economics and the engineering are not separable questions, and the paper’s contribution is to prove, rather than assert, that they are the same question.

The individual hash trial forgets everything. That much is true, and it is a theorem. But the Nakamoto protocol was never the hash trial. It is the hash trial wrapped in difficulty adjustment, coinbase maturity, fee accumulation, and the heaviest-chain rule — a machine assembled almost entirely out of memory. The models that called it memoryless were not describing the protocol. They were describing the one part of it simple enough to be convenient, and then mistaking the convenience for the thing. The protocol remembers the last fortnight, remembers every confirmation stacked upon a block, remembers every fee waiting in the mempool, and prices all of it. Good economics will have to remember too.

Based on “Legal Deterrence in ‘Permissionless’ Consensus,” by Craig Wright, forthcoming in the International Journal of Cryptocurrency Research, Vol. 6, Issue 1 (June 2026).

There is a story the cryptocurrency industry likes to tell about itself, and like most flattering stories it has survived chiefly because nobody with serious money has yet had a reason to test it. The story is that a blockchain is a law unto itself: a self-contained machine that secures value through mathematics and electricity alone, serenely indifferent to courts, prosecutors, exchanges, and the grubby apparatus of human enforcement. It is a romantic picture. It is also, for any transaction large enough to be worth stealing, false.

My new paper — Legal Deterrence in ‘Permissionless’ Consensus, forthcoming in the International Journal of Cryptocurrency Research (Volume 6, Issue 1, June 2026) — sets out to say precisely where that story stops being true, and why. It does so not by waving the usual flags, but by taking the most rigorous statement of the lawless view, granting it everything it asks for, and then showing how narrow its territory actually is. The subject is blockchain security, and the argument is that we have been modelling it with one hand tied behind our backs.

The benchmark worth taking seriously

The rigorous statement belongs to Eric Budish. In Trust at Scale, published in the Quarterly Journal of Economics, Budish derived a deterrence condition for proof-of-work consensus and proved something genuinely uncomfortable: in a purely permissionless system, trust at scale is expensive. If you want a Nakamoto-style ledger to deter an attacker from rewriting history — from spending the same coin twice, which is the only attack that actually matters here — then the honest network must keep spending. Not once, but continuously. The flow of resources burned to secure the chain has to be large enough that no attacker can profitably assemble enough hash power to overwhelm it.

Budish’s condition is simple to state in plain words. The cost of mounting an attack — the share of network power the attacker must control, multiplied by the ongoing cost of sustaining the network’s level of trust support, multiplied by the length of time the attack must be held — has to exceed the value the attacker stands to capture from a double-spend. If that inequality holds, the attack is deterred. If it does not, the attack pays.

I want to be unambiguous about this, because the temptation in my corner of the world is to treat Budish as an adversary. He is not. His theorem is correct. Within the domain he specifies, it is precise, and it is important. My paper does not lay a finger on it. What the paper does is read the fine print — and the fine print is a single phrase that does an enormous amount of work.

Budish derives his result under what he calls a no-rule-of-law scope condition. In plain terms: he assumes the attacker operates in a world without legal consequence. No prosecution. No civil liability. No seizure. No exchange freezing the proceeds. No counterparty refusing to deal. The attacker is a ghost, reachable by nothing but the protocol itself.

That assumption is not an error. It is a modelling choice, and a useful one, because it isolates the cost of securing a ledger in pure form. But an assumption that is useful is not the same as an assumption that is true, and the entire argument of my paper turns on that difference.

One narrow move, made carefully

The paper makes a deliberately narrow move. It extends Budish’s framework along two dimensions and no more.

The first extension is organizational. Budish’s attacker is a monolith — a single anonymous entity that can acquire whatever hash power it likes and hold it for as long as it pleases. That is not how mining works, and it has not been how mining works for well over a decade. The second extension is institutional, and it is the heart of the paper: it admits the rule of law into the attacker’s payoff.

Here is the part that matters for anyone inclined to accuse me of special pleading. Set both extensions back to zero — strip out the law, strip out the pool structure — and the model collapses exactly onto Budish’s original condition. Not approximately. Exactly. The paper contains a formal proposition to this effect, with a two-step proof. Budish’s theorem is recovered as the no-law, no-pool special case of a more general model. I do not refute him. I contain him. The disagreement is not about whether his result is right; it is about how much of the world it describes.

Mining is not a crowd of hermits

Start with the organizational extension, because the institutional one cannot stand without it.

The folk image of blockchain security is a vast field of solitary miners, each grinding away independently, anonymous and interchangeable. That image is dead. The mining literature buried it years ago. Bitcoin production is organised through mining pools — coordinated operations run by managers who assemble the block, choose which transactions go into it, decide which chain to extend, and distribute the rewards to contributors according to a payout rule. Cong, He and Li showed that pools became the dominant organizational form early in Bitcoin’s history. Chatzigiannis and colleagues showed that contributors shift their hash power between pools in response to expected returns and risk, much as an investor rebalances a portfolio. Arnosti and Weinberg went further still and proved that mining is a natural oligopoly: economies of scale drive the market toward concentration even when nobody is colluding.

This matters for one reason above all others. A contributor who points his machine at a pool is not the entity deciding to attack. The pool operator is. The operator builds the block; the operator chooses the chain; the operator coordinates the double-spend. The contributor merely supplies horsepower — necessary for the attack, but not sufficient, and certainly not the decision-maker. So when a deliberate double-spend runs through a pool, the first-order actor to whom attribution runs is the coordinating operator, not the thousands of machine owners feeding it.

The analogy the paper leans on here is, I confess, an indecent one, which is precisely why it works. Henry Thompson’s study of the industrial organization of the mafia — published, with no apparent sense of irony, in the Journal of Law and Economics — describes how criminal enterprises adopt a hierarchy: a boss who monopolises the right to authorise extreme action, an internal mechanism for settling disputes quietly, and a structure of incentives that keeps everyone aligned. The hierarchy exists precisely to limit the disputes that would raise the organization’s public profile. Replace “boss” with “pool operator” and “soldier” with “contributing miner” and you have described pooled mining with uncomfortable accuracy. The boss decides. The soldiers can walk. And the moment the boss’s conduct threatens the soldiers’ livelihood or liberty, they do.

The legal term, in plain English

Now the institutional extension. The paper introduces a single new quantity into Budish’s payoff: the expected cost of legal enforcement, net of the cost of invoking it.

In words, that legal cost is the probability that the attack is detected, multiplied by the probability that it is correctly attributed to a responsible party, multiplied by the probability that a sanction actually lands — a freeze, an injunction, a settlement, a prosecution, a recovery — multiplied by the size of that sanction. From this gross figure you subtract the cost the enforcing party must bear to invoke the law at all. And then — this is the crucial design choice — you take whichever is larger: that net figure, or zero.

That last operation is not a mathematical nicety. It is the engine of the entire argument. It encodes a participation constraint on the enforcer. A law that exists on the books is worth nothing if no rational party will pay to use it. If the cost of pursuing the attacker exceeds the expected recovery, no victim sues, no prosecutor charges, no exchange lifts a finger — and the attacker faces no legal risk whatsoever. The legal term is zero, and we are back in Budish’s world.

So the paper draws a sharp line between two things the lawless story quietly conflates: the existence of law, and the economic value of using it.

Why a five-dollar theft and a five-million-dollar theft are different worlds

Consider a double-spend worth five dollars. Who calls a lawyer? Who opens an investigation? Who convenes a prosecutor? Nobody, because the cost of doing any of it dwarfs the loss. At that value the enforcer participation constraint binds, the legal term is exactly zero, and the system relies on protocol-side deterrence alone. Budish is not merely approximately right here. He is exactly right. The paper concedes this without reservation and proves it as a proposition: for sufficiently small transaction values, the pure protocol-cost benchmark applies, full stop.

Now consider a double-spend worth five million dollars, executed against an identified pool operator with seized assets and balances sitting on exchanges. The calculus inverts. The expected sanction is enormous; the cost of enforcement, while real, is dwarfed by the prize being clawed back. Victims sue. Prosecutors charge. Exchanges freeze. The legal term is large and positive.

The paper proves that the legal cost rises with the size of the prize over the economically relevant range. The intuition is almost embarrassingly direct. Disgorgement alone returns at least the stolen amount, so the sanction grows at least dollar-for-dollar with the theft. Fines, forfeiture, and the threat of imprisonment pile on top. Meanwhile, the bulk of enforcement cost is fixed overhead — filing, investigation setup, institutional capacity — so it grows far more slowly than the sanction does. Bigger thefts are therefore more legally dangerous, not less.

This does not require the law to switch on at some magic number. The paper is careful to insist there is no hard threshold, only a region where the no-law approximation steadily weakens as the stake, and the visibility of the actor, rise. The honest answer is a gradient, and the paper gives a gradient.

The coalition is not a fortress

Here the two extensions fuse, and the result is the most economically interesting claim in the paper: pooled attacking capacity is not a fixed asset you buy once. It is a coalition you must hold together under deteriorating conditions — and there are two distinct ways it falls apart.

The first channel operates before anyone knows an attack is under way. When a pool secretly diverts effort to a hidden chain, its ordinary performance degrades. Payouts slip. Variance rises. Stale shares climb. Contributors need not understand the cause; they need only notice that this pool now pays worse than the one next door. And so they leave, for the most boring economic reason imaginable — better money elsewhere. This is the mining analogue of a result by Bloch and Kranton on corporate cover-ups: concealment can be sustained while outsiders’ beliefs remain favourable, but the very act of concealment generates the observable distortions that eventually erode those beliefs. A covert attack, in other words, leaks.

The second channel opens once the misconduct is suspected or named. Now continued association is no longer a neutral commercial decision. A contributor who keeps feeding hash power to a pool credibly accused of an attack risks legal scrutiny, contract termination, exclusion from the exchanges he needs in order to sell his coins, and — most painfully — the impairment of his hardware. So he exits to protect himself.

The arithmetic of this is brutal for the attacker. Effective attacking hash power is not the nominal capacity the pool advertises; it is what remains after contributors flee through both channels. A pool that nominally controls a third of the network can watch that share evaporate during the attack itself — not over weeks, but over hours.

Four more ways the prize bleeds out

Beyond the law and beyond contributor flight, the paper identifies four further losses that the lawless benchmark simply omits, each of them growing more severe as the prize grows.

Reputational sanctions. Here the paper refuses to hand-wave, and the refusal is instructive. It would be lazy to assert that reputation matters and move on. The evidence is more particular than that. Karpoff and Lott studied 132 firms caught committing criminal fraud and found that the reputational penalty — the collapse in market value over and above the legal fines — accounted for more than ninety percent of the total punishment. But Karpoff and colleagues later studied 478 firms caught committing environmental violations and found the reputational penalty there to be negligible: the market docked them roughly the value of the fines, and no more. Why the difference? Because fraud has identifiable victims who can withdraw their future business, whereas environmental harm is diffuse, with no single counterparty bearing a concentrated loss. A double-spend is structurally fraud against specific transaction recipients who accepted payments later invalidated. It therefore sits squarely in the category where reputational sanctions bite hardest. This is the disciplined way to invoke reputation, and it is the way the paper invokes it.

ASIC-capital loss. Mining hardware is not a general-purpose asset you can repurpose into a server farm. It is exquisitely specialised silicon with essentially no productive use outside Bitcoin and its close relatives. That specificity is ordinarily a mere inconvenience; in an attack it becomes a noose. A pool that is excluded, sanctioned, or rendered commercially toxic does not lose a few days of revenue — it risks the value of its entire capital base, stranded with nowhere to go. The paper converts part of Budish’s flow-cost problem into a stock-loss problem, which is a far heavier deterrent.

Honest-longest-chain exclusion. This one is elegant because it asks nothing of the honest network at all. Honest miners do not need to organise, coordinate, or change a single rule. They simply keep mining on the longest valid chain, exactly as the protocol instructs them to. If the attacker’s effective hash power falls below the honest remainder — through contributor flight, through degraded connectivity, or both — the attacker’s hidden chain is orphaned automatically. The protocol does the work. The attacker’s chain dies of natural causes.

Network-friction loss. A pool depends on infrastructure it does not own: domain naming, internet routing, interconnection, propagation. Root-zone and identifier coordination remain externally governed, and the security literature — Apostolaki and colleagues on routing attacks against Bitcoin, and the routing-integrity guidance from the National Institute of Standards and Technology — shows that interference at this layer can isolate mining capacity, slow propagation, and raise the rate at which a pool’s blocks are stranded. The paper is careful here: it makes no claim that any single state can simply switch off another country’s miners. The narrower, more defensible point is that a publicly identified, dishonest pool is operationally entangled with infrastructure it cannot unilaterally control, and that entanglement is one more margin on which an attack can be made to bleed.

Stacked together, these terms produce the paper’s central object: a net-gain function for the attacker that subtracts, from the realised value of the theft, the protocol-resource cost and all five of these losses. The attack pays only if that net figure comes out positive. The lawless benchmark keeps only the first cost and throws the rest away — which is exactly why it overstates how profitable a large attack can be.

The numbers are not gentle

A theory of this kind invites the obvious objection: fine in principle, but do these terms amount to anything in practice? The paper answers with three calibration exercises, and it is scrupulous about their status — they are illustrative, drawn from public filings and disclosed data, not econometric estimates. They are meant to establish orders of magnitude, and on that modest ambition they succeed comprehensively.

First, who actually runs the pools. Using thirty days of block data ending in March 2026 — 4,149 blocks, identified by the coinbase tags that pools stamp into the blocks they mine — the coordination layer of Bitcoin turns out to be startlingly legible. United-States-linked pools account for just over 42 percent of pool share, with Foundry USA alone at 31.48 percent. China-linked pools account for roughly 41.5 percent. F2Pool, registered in Singapore, accounts for 11.33 percent. More than 98 percent of hash power is attributable to pools with a verifiable, public operator identity. Less than two percent is attributable to pools you cannot name. This is not a swarm of anonymous ghosts. It is a small set of identifiable companies in a handful of jurisdictions — several of them, like MARA, listed on public stock exchanges and filing accounts with regulators.

Second, where the enforcement threshold actually sits. By bounding the probability of detection and attribution from the pool data, the magnitude of sanction from real enforcement settlements, and the cost of enforcement from litigation-cost surveys, the paper estimates the transaction value at which the legal term first turns positive. Across a range of parameter assumptions it lands between roughly one million and four-and-a-quarter million dollars. Even under the assumptions most generous to the attacker, the threshold sits below five million. Below it, Budish’s world. Above it, the law is in the room. And these are upper bounds, because they exclude criminal penalties, imprisonment, and reputational loss — any one of which would push the threshold lower.

Third, how much capital an attacker is actually risking. Take a single listed miner, MARA Holdings, which disclosed in its 2024 annual report roughly 400,000 mining rigs, energised hash power of 53.2 exahashes per second, and a treasury of 44,893 bitcoin. Set its total capital exposure against various attack targets and the ratios are absurd in the attacker’s disfavour. For a ten-million-dollar prize, capital exposure exceeds the prize by something on the order of 540 to 620 times. For a hundred-million-dollar prize, by 54 to 62 times. And that is one firm. A pool the size of Foundry aggregates the capital of many such operators. The plain statement is this: a pool operator contemplating a double-spend faces capital destruction measured in billions against a prize measured in millions.

Then the dynamics. The paper simulates a hidden attack by a pool starting with 31.5 percent of an 800-exahash network — roughly the Bitcoin network’s scale in March 2026 — diverting a rising fraction of its effort to a hidden chain. Across a thousand Monte Carlo runs, and tested under three different assumptions about how contributors decide to leave, the pool’s effective capacity after twenty-four hours falls to somewhere between 10.6 percent and 26.8 percent of where it began. Under the central specification it retains 26.8 percent. Every single specification leaves it far below the 50 percent it would need to sustain a majority attack. The capacity does not erode over days. It erodes over hours. Budish’s assumption of a static attacking capacity, the simulation suggests, badly understates how fragile a pooled attack really is.

What the mafia, the EPA, and DUI checkpoints have to do with Bitcoin

One of the quiet pleasures of writing this paper was the company it forced me to keep. To turn a protocol-only deterrence condition into one that includes the rule of law, you need tools — and the right tools already existed, not in the cryptocurrency literature, which is curiously incurious about enforcement, but in the law-and-economics literature, where deterrence has been studied seriously since Gary Becker formalised it in 1968.

So the paper borrows. From Thompson’s anatomy of the mafia it takes the structure of a hierarchical coordinator surrounded by members who can defect. From Bloch and Kranton’s work on cover-ups it takes the dynamics of concealment and the result that penalties must escalate to offset the temptation to hide. From Gulen and Myers it takes hard evidence that enforcement is selective and politically contingent — their finding that Clean Water Act violation rates run nearly 40 percent lower in electorally important battleground states is a gift to anyone modelling an enforcer who must decide whether to act. From Matsuzawa it takes deterrence through salience: his finding that banning DUI checkpoints raised alcohol-related traffic deaths by over twelve percent shows that visible, targeted enforcement deters a far wider population than the few it directly touches — which is exactly why sanctioning one identified pool operator changes the risk calculus for every other pool watching. And from Belnap and colleagues it takes the economics of transparency regimes, drawn from automatic information-sharing between tax authorities, where the great majority of institutions commit to sharing but the quality of what they actually share varies — a useful caution that detection is never perfect, even where the obligation plainly exists.

Not one of these papers is about Bitcoin. That is the point. The vocabulary of modern enforcement economics was sitting on the shelf, fully developed, waiting to be pointed at proof-of-work double-spending. Nobody had done it. That gap is what the paper fills.

Two regimes, one system

I have saved the most counterintuitive consequence for last, because it is the one that matters most to anyone who actually wants to build a payment system rather than merely admire one.

If the legal term only turns positive above a threshold of a few million dollars, then a system that processes an enormous volume of small payments — each one individually trivial, each one below the value at which any enforcer would bother — never leaves Budish’s protocol-only regime at all. Every one of those payments is secured by protocol cost alone, and because each payment is small, that cost is small. The per-transaction security cost falls as volume rises. Aggregate turnover can be vast while every individual transaction stays safely inside the lawless regime where Budish reigns.

So the system does not have one security model. It has two, running at once. Protocol-only deterrence for the ocean of small payments. Protocol-plus-law deterrence for the comparatively rare large transfers that pass through identifiable institutions and become worth attacking. Budish’s theorem describes the first regime with complete accuracy. My paper describes the second. Neither displaces the other, and the same chain runs both simultaneously.

This is not a rhetorical flourish. It is the resolution of an apparent paradox that has haunted the economics of these systems. Budish’s result has been read by some as a death sentence — proof that blockchains are uneconomic at scale. It is no such thing. It is a precise statement about the cost of securing a single high-value transfer in a world without law. The moment you populate the world with the law that actually exists, and the moment you recognise that most economic activity consists of small payments rather than nine-figure transfers, the supposed death sentence becomes a sensible division of labour.

What this does, and does not, claim

Let me be exact about scope, because precision is the whole point, and an argument that overclaims deserves to lose.

The paper does not claim that the law makes Bitcoin safe. It does not claim that every attack is detectable, or that every miner is easily attributable. It does not claim that enforcement is perfect, that courts reliably restore every victim, or that permissionless consensus intrinsically requires law for all its uses. A high-volume, small-payment system can and does operate happily inside Budish’s protocol-only regime with no legal recourse at all.

What the paper claims is narrower and, I think, much harder to dismiss. Once economically significant value moves through legally legible pooled organizations, the no-rule-of-law benchmark stops being the correct positive description of how profitable an attack is. The relevant deterrence condition becomes value-dependent and organizationally specific. It is jointly determined by protocol costs, by legal sanctions net of enforcement cost, by reputational loss, by the mobility of pool members, by the destruction of specialised capital, and by the protocol’s own quiet habit of orphaning a chain that has lost its majority.

Why it matters

The broadest contribution is methodological, and it is the one I would ask a reader to carry away. The interesting contrast in this field was never between a lawless blockchain and a law-governed outside world. That framing flatters the technology and corrupts the analysis. The real contrast is between different bundles — different combinations of protocol incentive, organizational structure, legal exposure, and infrastructural dependence. Once you frame the problem that way, the economics of double-spending stop fitting inside a single flow-cost inequality. They become a question about whether an attacker can preserve the legal, commercial, and operational environment he needs in order to convert a clever rewrite of the ledger into durable wealth he can actually keep.

For a five-dollar theft, he can. The enforcer shrugs, the legal term is zero, and the gain is his to keep. For a five-million-dollar theft against a named operator with a fleet of stranded ASICs, a treasury sitting on an exchange, and a contributor base already heading for the exits, he very probably cannot — and the protocol he tried to subvert orphans his chain while the lawyers are still filing their first motion.

Budish gave us the lower bound: the irreducible cost of trust where no law reaches. That bound is real, and it is his. The work that remains — the work this paper begins — is to map the far larger territory where law does reach, where mining is a concentrated and identifiable industry rather than a swarm, and where the rule of law was never actually absent. It was simply waiting, as it usually does, for the numbers to grow large enough to be worth its attention.

The lawless blockchain, in the end, is a story we tell for small change. It is a perfectly good story for small change. It is only when we mistake it for the whole economy that we begin to believe our own romance.

KEYWORDS / TAGS: cross-border payments, trade finance, working capital, days sales outstanding, payment finality, CBDC, BRICS, receivables finance, settlement, monetary plumbing

There is a particular kind of error that is hard to see because everyone is making it at once. The cross-border payments debate is in the grip of one. We have decided that the problem with international trade is that money moves too slowly, and we have built an enormous reform agenda on top of that premise: central bank digital currencies, multi-CBDC settlement platforms, instant domestic rails bolted to cross-border corridors, China’s CIPS as an alternative to the incumbent messaging layer, tokenised deposits, stablecoins pitched as the plumbing of a faster world. The shared assumption is that settlement latency — the interval between a buyer releasing funds and a seller holding final, usable money — is a meaningful cost, and that compressing it from days to seconds will unlock something large.

I went and measured it. The premise is wrong for the trade that actually carries the value. Not slightly wrong, not wrong at the margins. Wrong about which margin is binding. And once you see the arithmetic, you cannot unsee it.

Let me state the conclusion before I defend it, because I dislike essays that make you wait. In the large-value, trade-credit-financed trade conducted by listed firms across the BRICS and BRICS-relevant economies, the seller’s money is not tied up in the payment rail. It is tied up in the receivables cycle — the months a sale sits on the balance sheet as an unpaid claim before cash arrives. Payment-rail latency is one to five days. The receivables cycle is forty-five to a hundred-and-ten. Faster finality acts only on the days; it cannot touch the months. The people selling milliseconds are solving a problem these firms do not have.

What the money is actually doing

Start with the object, because the whole argument lives in a definition that most of the payments conversation never pauses over.

When a firm sells on terms — and serious cross-border trade is almost always on terms — it ships the goods, recognises the revenue, and books a receivable. That receivable is a claim, not cash. It sits on the asset side of the balance sheet until the buyer pays, which by contract might be thirty, sixty, or ninety days later, and in practice often longer once collection friction and documentary reconciliation are added. During that interval the seller has performed but has not been paid. It has, in effect, lent the buyer the value of the shipment. That loan has to be financed, whether out of the firm’s own equity, a revolving credit line, a trade-finance facility, or a discounted invoice. The financing has a cost. That cost — the carry on working capital tied up across the receivables cycle — is the real price of moving goods on credit.

The standard balance-sheet measure of this duration is days sales outstanding. You take receivables, divide by sales, multiply by 365, and you get the average number of days a unit of sales remains an unpaid claim. It is not a precise settlement clock — it is an annual stock-over-flow approximation, affected by seasonality, revenue-recognition policy, the customer mix, and year-end timing — but it is the right order of magnitude for the holding period, and it is the figure every treasurer and credit analyst already lives by.

Now hold that next to settlement latency. Settlement latency is the last step of the timeline: the gap between the buyer instructing payment and the seller having final funds. A faster rail compresses that gap. What a faster rail cannot do is move the sixty days the claim sat as a booked receivable before payment was even released. Those days are the negotiated credit term plus the collection and documentary friction that precede the transfer. They are not a payments problem. They are a financing-and-contracting problem wearing a payments costume.

This is the whole essay in one sentence: the receivables cycle, not the rail, is the thing the seller’s capital is locked inside, and a payment rail can only ever reach the sliver at the very end.

The one number that matters

Reduce it to a ratio and the point becomes unanswerable. Call the holding period T — the full time the money is tied up — and call the rail latency L — the slice a faster rail can remove. The fraction of the cost that settlement speed can address is L divided by T. I call that λ. It is the share of the cycle that finality speed can reach, and everything turns on its size.

If T is ninety days and L is two, λ is a little over two percent. If you drove latency to zero — instant, final, free settlement, the maximum the reform agenda could ever deliver — you would remove a little over two percent of the working-capital carry and leave the other ninety-eight percent exactly where it was. That is not a rounding error you can argue your way out of. It is the structure of the problem.

And here is the property that closes the last escape route. λ is rate-invariant. The carry on the whole cycle scales with the financing spread; so does the carry on the latency slice; the spread is in the numerator and the denominator, and it cancels. This matters because the natural objection from someone defending the speed agenda is to reach for an interest-rate assumption — “but at high rates the latency cost is large.” No. At any rate, the latency share of the carry is L over T. You cannot inflate the importance of settlement speed by assuming a higher cost of capital, because a higher cost of capital raises the cost of the months by exactly as much as it raises the cost of the days. The ratio does not move. This is the single most useful thing in the analysis, and it is the reason the conclusion is robust to the parameter people would otherwise fight about.

So the policy-relevant statistic is not a dollar figure that depends on a dozen assumptions. It is a pure ratio, λ = L/T, and it is small whenever T is months. The only question left is empirical: how long is T, actually, for the firms doing the trade?

The data

I did not want to argue this from stylised facts, so I built it from firm-level accounts.

The receivables side comes from Compustat Global — standardised, consolidated annual statements — for the operating firms domiciled in eight economies central to the China–BRICS settlement debate: China, India, Indonesia, Brazil, Saudi Arabia, the United Arab Emirates, South Africa, and the Russian Federation. That is 12,241 firms and 73,533 firm-years over FY2019 to 2025. I computed days sales outstanding for every firm-year, screened to operating firms with genuine sales and a cost of production, and applied each ratio’s sane validity range so the statistic is computed on the firms that identify it rather than on accounting artefacts. The financing side comes from Dealscan — 13,031 syndicated loan facilities for borrowers in the same economies — collapsed deterministically to one all-in-drawn spread per facility, which gives the price at which these firms actually fund themselves.

The medians are not subtle. Days sales outstanding runs from about forty-five days in the Russian Federation to about a hundred-and-eight in China, with the others — India in the low seventies, Indonesia in the mid-fifties, Brazil around eighty, Saudi Arabia in the mid-eighties, the UAE near a hundred, South Africa in the low fifties — strung between. Financing spreads run from roughly 145 basis points over base in Saudi Arabia to 441 in China. Put the duration and the spread together and the annual working-capital carry on a unit of trade value lands between about twenty-nine and a hundred-and-thirty basis points, depending on the economy.

Against all of that, set a latency of two to five days. The latency share λ comes out at a median of about 2.6 percent at a two-day finality and about 6.5 percent at five days. The largest value anywhere in the panel — in the Russian Federation, which has the shortest cycle and therefore the least room — is eleven percent. Eleven percent is the ceiling, in the one economy most favourable to the speed story, at the most generous latency assumption. Everywhere else, the number a faster rail can touch is low single digits.

This is the empirical heart of it. The cross-border payments reform agenda is optimising a variable whose maximum achievable effect, in the trade that matters, is in the low single digits of a cost that is itself only one of several the agenda does not address.

It is not just the median

A median can hide a tail, and a careful reader will immediately ask whether the distribution has a meaningful mass of firms with very short cycles — firms for whom settlement speed really would dominate. So I looked at the whole distribution, not the midpoint, and reported the exact share of firm-years below each threshold that the theory makes relevant.

The thresholds come straight from the ratio. A two-day latency only reaches even ten-percent relevance once the holding period drops below twenty days; a five-day latency needs the period below fifty. So the right question is: how many firm-years actually sit below those lines?

Below twenty days, the share ranges from 2.6 percent in Brazil to 17.2 percent in Indonesia. Below fifty days — the five-day-latency threshold — it ranges from about nineteen percent in the UAE to about fifty-five percent in the Russian Federation. And below ten days, the region where finality speed would genuinely dominate the cost, the share is at most 9.1 percent in any economy. The overwhelming mass of firm-years sits well to the right of the latency band, in every single economy. The distribution does not rescue the speed thesis. It buries it.

I find it useful to picture it. Lay each economy’s distribution of receivables periods along a horizontal axis measured in days. Shade the one-to-five-day band where rail latency lives. In every economy, the bulk of the distribution — the interquartile box, the median, most of the tails — sits dozens to well over a hundred days to the right of that shaded sliver. The gap between where the money is and where a faster rail can reach is not a matter of interpretation. It is visible at a glance: months on one side, days on the other.

“But the cycle is mostly voluntary”

The sharpest objection to all of this is not about the data. It is conceptual, and it deserves a real answer rather than a brush-off. It goes: you are comparing latency against the entire receivables cycle, but most of that cycle is a voluntary commercial decision. The seller chose to offer ninety-day terms. That credit period is a product feature — a financing service bundled into the sale — not a friction. Comparing a settlement cost against a voluntary credit term is comparing a cost against a choice.

This is a good argument and it is partly right. Some of the cycle is genuinely chosen. But it does not get the speed agenda where it needs to go, for two reasons.

First, even the voluntary part is a cost the seller bears and finances, and the question on the table is which instrument reduces it. A faster rail does not shorten a negotiated credit term by one minute. So whether the term is voluntary or not is beside the point when you are asking what payment-rail speed can accomplish: the answer is nothing, because rails do not touch terms.

Second, and more carefully, I separated out the part of the cycle that is plausibly involuntary — the collection lag and the documentary and reconciliation friction that sit on top of the contractual term — and computed the latency share against that frictional residual alone, bounding it conservatively two ways because the residual is not directly observed. Even measured against only the involuntary friction, the latency share is in the mid-single digits. The conclusion survives the strongest form of the objection. Settlement speed is a small lever even against the part of the cycle nobody chose.

Does the composition trick you?

The other serious challenge is statistical rather than conceptual. Days sales outstanding varies a great deal across industries — a heavy-equipment maker selling on long terms looks nothing like a supermarket — and the mix of listed industries differs across economies. So perhaps the long cycles are an artefact of which sectors happen to be listed where, and the cross-economy story is really a sector-composition story in disguise.

This is the kind of objection you cannot wave away; you have to do the work, so I did. I pulled the industry classification and recomputed the receivables cycle four ways: manufacturing only, with construction and real estate excluded, with financial firms excluded, and with the industry effect removed altogether by adjusting each firm-year against its own industry’s global median. Across every one of those cuts, every economy’s median receivables cycle stays between roughly forty-five and a hundred-and-thirteen days. The lowest median under any restriction — the Russian Federation, again — is still more than twenty times a two-day latency. The order-of-magnitude result does not turn on sector composition. Months stay months whichever way you slice the listed sector.

I will be honest about what sector composition does affect, because the discipline of the exercise is to say exactly what survives and what does not. The cross-economy ranking — whether China’s cycle is longer than the UAE’s, and by how much — does shift under industry adjustment. So I treat the ranking and the within-economy size gradient as descriptive colour, not as identified findings. What I claim is the robust thing: the gap between a months-long cycle and a days-long latency, in every economy and under every reasonable cut. Not the league table. The league table is not the point; the order of magnitude is.

One more robustness check, because it pre-empts a methodological complaint: days sales outstanding strictly needs only sales and receivables, not the cost-of-goods screen I used to define operating firms. So I recomputed it the other way — sales only, financial firms removed by industry code instead — and every economy’s median landed within about a day of the original. The screen is not driving the result. The result is in the data, not in my filters.

The dollars

Ratios persuade analysts; magnitudes persuade everyone else. So I scaled the comparison to trade flows, with the explicit warning that this is an illustration of order of magnitude and not a welfare estimate or an observed bilateral settlement cost.

Take China’s goods trade with seven of these partners in 2024 and apply the economy-level working-capital terms. The annual carry on the receivables cycle implied by that exercise is on the order of three-and-a-half to four billion dollars. The slice of it attributable to a two-day settlement latency — the part a perfect, instant, free rail could remove — is about a hundred-and-ten million. Three-point-seven billion against zero-point-one-one billion. The cycle carry is real money, and the point of the paper is emphatically not that it is small; it is that the instrument the speed agenda offers reaches the wrong end of it. Compressing finality removes the hundred-and-ten million at most. The other three-and-a-half billion is the financing cost of the credit term and the non-payment friction, and a faster monetary rail does not reach it.

If you wanted to act on the three-and-a-half billion, you would not build a faster rail. You would do something to the holding period or to the spread — shorten the documentary and collection friction, or deepen the market that discounts the receivable. That is a different toolkit entirely, and it is the toolkit nobody is funding with the enthusiasm reserved for settlement speed.

Where the speed agenda is actually right

Now I want to do the thing that separates an argument from a polemic, which is to say precisely where the opposing view is correct — because it is correct, in a regime this trade does not occupy, and the boundary between the two regimes is the most interesting part of the whole problem.

The entire result is conditional. λ is small because T is large. T is large for a specific, identifiable reason: this trade is relationship-based and credit-financed. The buyer’s promise to pay in sixty or ninety days is itself the instrument; the receivable is an asset precisely because there is a counterparty with a balance sheet standing behind it and a contract worth enforcing. Strip those features away and the whole structure inverts.

Consider a transaction with a counterparty that has no balance sheet to extend credit against, a value too small to negotiate terms over, and a delivery metered in seconds rather than quarters. There is no receivables cycle there, because there is nothing to finance and no one to finance it. The holding period collapses toward the settlement interval itself. T falls toward L, and λ — latency over holding period — rises toward one. In that regime, settlement speed is not a small margin. It is the entire margin. The cost of delay stops being a financing carry on a float and becomes the foregone throughput of an operation that cannot proceed until finality clears.

The limiting case is machine-to-machine settlement: autonomous systems paying each other for compute, bandwidth, energy, data, or physical priority at the moment of use, in fractions of a cent, with no credit relationship and no negotiated term. There, finality speed is the binding constraint, because finality is the holding period. The same arithmetic that makes latency negligible for a ninety-day, million-dollar receivable makes it decisive for a microsecond, sub-cent machine payment.

So the people building for instant, final, micro-value settlement are not wrong. They are building for a different point in the same parameter space — a regime defined by the joint collapse of value and duration. The error is not that they value speed. The error is importing a settlement architecture optimised for the machine regime into the analysis of large-value trade credit, where the parameters are the opposite and the optimisation target is therefore the opposite. An architecture tuned for one is not thereby tuned for the other. Today’s cross-border trade between firms with balance sheets, long terms, and large invoices is squarely in the regime where latency is the smallest lever. Tomorrow’s machine economy may not be. Both can be true, because they are different coordinates, and conflating them is exactly the mistake the current debate makes.

The policy that would actually move the number

If settlement speed is the wrong lever, the constructive question is which lever is right, and the decomposition answers it directly. Rank the instruments by the margin each one acts on, and the ordering falls out of the cost structure rather than out of fashion.

First, and largest, is the financeability of the trade claim. Anything that lets a verified receivable be discounted cheaply — invoice verification, receivables registries, enforceable assignment of the claim, credit insurance — acts on the financing spread, and the spread is paid on every single day of the cycle. Lower the cost of turning a receivable into cash and you lower the carry across the entire holding period at once. This is the instrument with the most leverage, because it attacks the price of the whole thing rather than a few days at the end.

Second is documentary and legal transferability. A great deal of the involuntary friction in the cycle is documentary state that cannot be discharged or financed without manual reconciliation — bills of lading, warehouse receipts, the paper apparatus of trade. Legal recognition of controllable electronic trade documents, of the kind the UNCITRAL Model Law on Electronic Transferable Records contemplates, compresses that frictional residual. This is plumbing worth building, but notice it is the plumbing of documents and title, not the plumbing of payment finality.

Third is counterparty-risk compression — export-credit guarantees, buyer payment undertakings, standardised confirmation — which acts on the contractual term by making the credit relationship safer to extend and therefore shorter and cheaper to carry.

And fourth, last, is payment-rail speed. Not zero — it is not nothing — but last, because it acts only on the latency slice, which is the smallest direct component of the cost this trade actually bears.

I want to be clear that “last” is not “useless.” Faster, more resilient settlement can matter for reasons this analysis does not measure: resilience against the weaponisation of payment infrastructure, sanctions exposure, liquidity management, financial inclusion, reach into populations the banking system does not serve. Those can be excellent reasons to modernise rails. They are simply not the same as reducing the working-capital cost of large-value trade, and the case for rail modernisation should be made on its real merits rather than on a working-capital benefit the arithmetic does not support. Match the instrument to the margin, and the margin says: finance the receivable first, digitise the documents second, de-risk the counterparty third, and speed the rail fourth.

What I am not claiming

Rigour is mostly a matter of refusing to claim more than you have shown, so let me draw the lines clearly, because they matter as much as the result.

This is descriptive measurement, not causal identification. I measured the lock-up and contrasted it with the latency. I did not estimate the effect of adopting any particular settlement architecture, and I would not, because the data cannot support it. The most tempting natural experiment — the 2022 exclusion of Russian banks from the conventional rails — bundles the settlement-access shock with sanctions, asset freezes, and counterparty exit so thoroughly that no honest researcher could isolate a clean settlement effect from it. The Russian receivables cycle did shift over that period; that is consistent with a settlement effect and is not evidence of one. I report it and decline to over-read it.

The sample is listed firms, by construction. The segment most exposed to settlement friction — small and unlisted exporters — does not appear in Compustat at all. I corroborated the mechanism in that segment using the World Bank Enterprise Surveys, which show working capital financed heavily through inter-firm credit and access to finance among the most frequently cited obstacles, more so for smaller firms. But survey evidence on financing structure is consistent with the mechanism; it does not measure settlement cost and it does not prove that latency is immaterial for an SME exporter. I will not pretend it does.

And the model is an accounting identity, not a structural estimate of behaviour. It decomposes the cost and shows which component each instrument touches. It does not predict how firms would re-optimise under a new settlement regime. That is a different paper, and it would need transaction-level settlement data nobody has handed me.

What I am claiming is narrow and, I think, solid: in the observed large-value, trade-credit-financed, listed-firm regime, the receivables cycle is months, the rail latency is days, the share a faster rail can reach is low single digits, and that share is rate-invariant and robust to sector composition and to the screens I used. The redirection follows from the measurement. The treatment effect does not, and I do not assert it.

The discipline

The reason this matters beyond one corner of trade finance is that it is an instance of a general failure: choosing an instrument before measuring the margin it is supposed to move. The cross-border payments agenda picked settlement speed — visible, technically exciting, fundable, easy to demonstrate on a stage — and then went looking for the benefit. When you do it in that order, you end up optimising the variable that is easiest to move rather than the one that is binding, and you can spend a decade and a great deal of capital making a number smaller that was never the problem.

Do it in the other order. Decompose the cost. Find the component that actually carries the weight. Then match the instrument to that component. For the trade that moves real value across these economies today, the binding cost is the financing of a months-long receivable, and the instruments that reach it are the ones that make the claim cheaper to finance and faster to clear of its documentary and contractual friction. Settlement speed will have its decade, and it will be the decade of the machine economy, where value and duration collapse together and finality becomes the whole game. That decade is coming. It is not this trade, and it is not now.

Measure the margin before you choose the instrument. Everything else is engineering in search of a problem.

There is always a priest at the end of a failed speculation, and he is always telling the congregation to believe harder.

He does not say, “The machinery is broken.” He says, “The faithful are weak.”

He does not say, “The asset produces nothing, settles little, carries no meaningful commerce, and survives by selling future hope to later entrants.” He says, “You lack conviction.”

He does not say, “We need another buyer.” He says, “Store of value.”

That phrase has become the velvet curtain thrown over an ugly little table where the same card trick is performed again and again. The cards are not even marked well. The dealer is drunk. The room smells of old confidence and cheap aftershave. Yet the crowd keeps applauding because everyone in the room has a ticket, and everyone knows the ticket only matters if someone even more credulous walks in after them.

BTC has become the perfect monument to this ritual. Not Bitcoin as digital cash. Not peer-to-peer electronic cash. Not a system for small casual payments. Not a machine for global commerce, micropayments, EDI, logistics, supply-chain data, settlement, and actual productive exchange. No. BTC is now defended as a holy relic whose chief function is to be bought and then sold to someone later at a higher price.

That is not economics. That is a queue.

The entire “store of value first” mythology rests on a concealed premise: that value can be stored without first being created. This is the financial equivalent of bottling smoke and calling it wine. A productive asset may store value because it participates in value creation. A farm produces crops. A factory produces goods. A bond promises cash flows. A share represents residual claim over enterprise. A monetary system, if it is to be more than a speculative token, facilitates exchange. It reduces friction. It lowers transaction costs. It enables commerce that otherwise would not occur.

But a token that does not scale, does not serve commerce, does not support ordinary transactions, and cannot be used at volume is not storing value in any meaningful economic sense. It is storing belief. Belief is portable. Belief is intoxicating. Belief is also flammable.

The problem with every Ponzi-like structure is not merely fraud. Fraud is the legal category. The deeper economic category is dependency on continual external inflow. A system becomes Ponzi-like when present holders depend primarily on future buyers rather than present utility, productive yield, or transactional necessity. The old holder is paid by the new holder. The new holder becomes old and waits for a newer holder. The ceremony continues until the room runs out of fools or the exits jam.

That is the part the moon boys do not like discussing. They prefer the bright side of the curve. They like the early slope, the dopamine, the little green candles rising like church spires over a village of idiots. They do not like the second half of the mathematics. Exponential narratives die because the world is finite. There are only so many buyers, only so much capital, only so much attention, only so many people willing to trade productive assets, cash-flowing instruments, business equity, land, labour, and real commerce for a sterile token that promises salvation later.

Later is the god of every speculative mania.

Later the adoption will come.

Later the fees will make sense.

Later the scaling will happen.

Later the world will arrive.

Later the last buyer will not be the last buyer.

But later has a habit of becoming now, and now has a habit of asking for receipts.

BTC’s defenders have a particularly comic version of the problem. They claim it is destined for global monetary settlement while defending a system that cannot process the transaction volume of a modest commercial network. At roughly a handful of transactions per second, the thing is not a global economy. It is a turnstile with a marketing department. It is a broom closet calling itself a cathedral.

And this matters because exits matter.

A speculative asset can appear liquid while confidence rises. Everyone congratulates himself because the price is high and the order books look alive. But liquidity in a mania is not the same thing as liquidity under stress. Liquidity is not proven when everyone wants in. Liquidity is proven when everyone wants out.

At five transactions a second, the exit is not a door. It is a straw.

When the crowd wants to reach the gateway, when they need to move, redeem, settle, hedge, transfer, or flee, the network itself becomes the bottleneck. Then the beautiful mythology of sovereign escape becomes a queue of trapped believers paying higher and higher fees for the privilege of discovering arithmetic.

That is not freedom.

That is congestion with a logo.

A genuine monetary system must be useful before it is sacred. It must move. It must clear. It must carry ordinary economic life. It must handle small casual payments and industrial settlement. It must make commerce cheaper, faster, broader, and more auditable. It must allow businesses to build without being told that their transactions are “spam” by some priesthood of hobbyists defending artificial scarcity as though starvation were nutrition.

This is where the BTC doctrine becomes grotesque. The people claiming to defend Bitcoin have converted ordinary commercial use into heresy. Logistics? Spam. EDI? Spam. Data? Spam. Micropayments? Spam. Enterprise records? Spam. High-volume business processes? Spam. Anything that resembles the actual design of a digital cash system becomes an offence against the sacred bottleneck.

So businesses do the rational thing. They test where testing is cheap. They test where a few dollars can produce large volumes of real network activity. They test where failure does not cost a fortune and where experimentation is not strangled by fees, congestion, and ideological scolding. A company can run meaningful trials on BSV for the price of lunch. It can test workflows, record data, stress processes, integrate systems, and learn something. On BTC, it can perform a ritual sacrifice to the fee market and then be told that its use case should not exist.

Then the BTC crowd looks around, baffled, and asks why nobody builds.

Because you told them not to.

You called their business spam.

You turned the network into a gated museum and then wondered why no factories appeared inside.

The “store of value” narrative is not a triumph. It is an admission of failure dressed as philosophy. It says: we cannot process global commerce, therefore commerce is unnecessary. We cannot support small payments, therefore small payments are beneath us. We cannot scale, therefore scaling is vulgar. We cannot compete as digital cash, therefore digital cash was never the point.

This is not adaptation. It is retreat with better typography.

And the retreat cannot continue forever because holding BTC gets the holder nothing except the hope that another person will later want the same nothing more intensely. There is no yield intrinsic to holding it. There is no productive engine beneath it. There is no universal transactional necessity forcing adoption. There is only the promise that belief itself will compound.

But belief without utility decays.

It may decay slowly. It may decay after another bubble. It may decay under a mountain of podcasts, slogans, institutional products, and solemn men saying “monetisation” as though they have discovered physics. But decay it must, because eventually buyers ask what they own.

Not what it might be worth to the next buyer.

What it does.

That question is fatal.

If BTC is digital cash, then it must function as cash. It must support payments. It must scale. It must allow commerce. It must make economic exchange more efficient. If it is settlement infrastructure, then it must settle meaningful volumes. If it is a global monetary network, then it must be capable of global network activity. If it is none of these, then it is merely a speculative object whose price depends on persuading later entrants to confuse scarcity with value.

Scarcity alone is not value. A dead rat in a locked box is scarce. Bad poetry written on a napkin is scarce. A unique disease is scarce. Scarcity becomes economically meaningful only when joined to demand grounded in utility, beauty, status, necessity, productivity, or power. Scarcity without use is not money. It is clutter.

BTC’s defenders try to avoid this by treating price as proof. The price is high, therefore the thesis is true. This is the oldest stupidity in markets. Tulips had prices. Railway shares had prices. South Sea dreams had prices. Dot-com vapour had prices. Housing derivatives had prices. Price proves only that someone paid it. It does not prove that the payment was wise, sustainable, or grounded in productive reality.

A bubble is not refuted by pointing at the bubble.

A Ponzi-like structure is not redeemed by saying it has not collapsed yet.

All such systems end because their implied growth requirement exceeds the available world. Early entrants need later entrants. Later entrants need still later entrants. As the base grows, the required inflow grows larger. The asset must attract not merely buyers, but increasingly large amounts of capital relative to its prior size. The story must become grander, the promises larger, the enemies more sinister, the doubts more forbidden.

Eventually the necessary buyer does not arrive.

Or arrives too late.

Or arrives with insufficient money.

Or arrives, looks at the machine, and asks why the thing cannot handle a corner shop.

That is when slogans become accounting.

The holder who believed he owned destiny discovers he owns exposure. The priest tells him to hold. The exchange tells him withdrawals are delayed. The fee market tells him the exit is crowded. The influencers tell him this is healthy. The chart tells him something else.

There is a cruelty in arithmetic, but also a mercy. It does not flatter. It does not care who has the blue tick, who shouted loudest, who called whom stupid, who sold the conference ticket, who coined the phrase, who wore the laser eyes. It simply adds the columns. If the system cannot scale, it cannot serve. If it cannot serve, it cannot anchor value. If it cannot anchor value, it depends on belief. If belief depends on price, and price depends on new buyers, then the machine has already confessed what it is.

The end may not be immediate. Manias can persist longer than dignity. They can acquire institutions, jargon, custody products, committees, lobbyists, analysts, professors, and all the other respectable upholstery of bad ideas. But respectability does not create utility. It merely gives failure better shoes.

The decisive question remains: why hold?

Not why speculate.

Not why trade.

Not why gamble.

Why hold, permanently, as an economic actor?

If the answer is “because someone else will pay more,” then the system is buyer-dependent. If the answer is “because it enables commerce,” then it must demonstrate commerce. If the answer is “because it settles value,” then it must settle at scale. If the answer is “because it is scarce,” then the answer is incomplete to the point of comedy.

BTC asks the world to accept the conclusion while exempting it from the test. It wants the status of money without the burden of monetary function. It wants the valuation of infrastructure without the throughput of infrastructure. It wants the reverence given to settlement systems while rejecting the messy, vulgar, necessary work of settlement.

And when someone points this out, the faithful retreat into insults. They call use “spam.” They call scale “centralisation.” They call commerce “attack.” They call throughput “fake.” They call enterprise testing “irrelevant.” They call digital cash “obsolete.” Then they wonder why the economic world does not kneel.

The world does not kneel because the world has invoices.

It has shipping manifests, purchase orders, stock movements, customs records, payments, refunds, audits, receipts, sensor data, contracts, settlements, remittances, and a thousand other dreary beautiful facts of commerce. Real economies are not memes. They are pipes, ledgers, records, obligations, deliveries, signatures, and deadlines. A system that cannot carry them is not a monetary revolution. It is a speculative ornament.

The final irony is that the BTC narrative now depends on non-use. It survives by claiming that use is unnecessary or undesirable. That is the logic of a restaurant that refuses to serve meals because food would damage the purity of the kitchen. The fewer transactions, the cleaner the myth. The more useless the network, the more sacred the scarcity. The less it resembles peer-to-peer electronic cash, the more fiercely they insist it has transcended cash.

No. It has not transcended cash.

It has abandoned function.

And abandonment has consequences.

When a monetary system is not used, its value must be explained by future use or future buyers. If future use is structurally prevented by throughput limits and cultural hostility to ordinary transactions, then only future buyers remain. That is the naked mechanism under the robes. Buy now because later someone else will buy higher. Hold now because later institutions will arrive. Believe now because later the world will understand.

Later, later, later.

The gospel of the last fool.

Every Ponzi-like system has an ending because no chain of buyers is infinite. The only way to escape that end is to become useful enough that holders are not merely waiting for exit liquidity. They hold because the system performs a function. They use it because it reduces cost. They integrate it because it solves a problem. They transact because the network is better than the alternative.

BTC chose the opposite path. It chose artificial constraint and called it virtue. It chose speculative hoarding and called it monetisation. It chose fee pressure and called it security. It chose non-use and called it purity. It chose the museum over the marketplace.

Museums are quiet.

Markets are not.

The future belongs to systems that move economic reality, not systems that sneer at it. A network that treats business transactions as pollution will not become the foundation of business. A network that cannot clear ordinary volume will not become global settlement. A network whose holders depend on perpetual appreciation will eventually meet the finite world.

And when it does, the question will not be whether the believers were sincere.

Of course they were sincere. Sincerity is cheap. Every doomed scheme has sincere men in it. Some of them cry beautifully on the way down.

The question will be whether the thing worked.

Whether it carried commerce.

Whether it scaled.

Whether it produced utility.

Whether holding it gave anyone anything beyond the hope of selling it to the next man.

That is where the romance ends. Not in a dramatic collapse necessarily. Sometimes these things end as farce, sometimes as stagnation, sometimes as slow irrelevance covered in slogans. The priests keep talking. The faithful keep nodding. The price twitches. The conferences continue. The word “early” is still used long after the grey hairs arrive.

But beneath it all, the last buyer waits.

He is always there.

He is the necessary saint of every speculative religion. Everyone before him needed him. He arrives carrying the final bag, though he does not know it yet. He has heard the sermons. He has seen the charts. He has been told that this is prudent, inevitable, revolutionary, mathematically ordained.

Then he looks for the utility.

He looks for the commerce.

He looks for the exit.

And at five transactions a second, he finds the queue.

Keywords

Austrian economics; money; monetary theory; economic calculation; Carl Menger; Ludwig von Mises; Friedrich Hayek; distributed knowledge; purchasing power; inflation; price signals; unit of account; capital theory; digital cash; Bitcoin; monetary competition; rule-based systems; decentralized money; information economics

Abstract

Modern monetary discourse rests upon a category error so ordinary that it now passes for common sense: the confusion of money with wealth, numerical magnitude with value, and monetary measurement with economic reality. This article develops a sustained Austrian and information-theoretic argument that money is not wealth but the calculative instrument through which wealth is compared, exchanged, and socially coordinated. Beginning with the paradox of Mr Darcy’s £10,000 per annum in Jane Austen’s Pride and Prejudice, the article shows how modern readers misread monetary history because they evaluate nominal figures rather than command over resources. Drawing on Menger’s account of the market emergence of money, Mises’s theory of economic calculation, Hayek’s analysis of distributed knowledge, and the constitutional logic of rule-based distributed digital cash, the article argues that money is best understood as an informational institution rather than a political artefact. Inflation, protocol mutability, and discretionary monetary governance are therefore not merely technical defects but corruptions of the measuring system through which civilization apprehends economic reality. The article concludes that a monetary system cannot be called decentralized in any meaningful economic sense if its rules can be altered by administrators, committees, or developers, because such mutability converts money from law into policy and from measurement into power.

Introduction: Darcy, Delusion, and the Monetary Eye

The modern reader’s reaction to Mr Darcy’s income is a diagnostic test for monetary understanding, and it is a test that modernity usually fails before it realizes that a question has been asked. In Pride and Prejudice, Austen (1813/2008) informs the reader that Darcy enjoys £10,000 per annum, a sum immediately intelligible to her original audience as the mark of extraordinary wealth, entrenched social standing, and immense command over productive resources; yet the modern reader, trained by two centuries of monetary expansion and numerical inflation, often encounters the same figure and instinctively treats it as modest. The number has not changed, but the interpretive faculty has. Austen’s contemporaries did not worship the arithmetic symbol; they understood the economic substance behind it. The modern mind, by contrast, sees the digits and mistakes them for the reality. That error is not literary trivia. It is a small window into the vast intellectual decay by which money has ceased to be understood as a measure of wealth and has instead been mistaken for wealth itself.

The Darcy paradox matters because it reveals the precise point at which monetary cognition collapses. A person who asks whether £10,000 was “a lot of money” is not truly asking about ten thousand units of sterling; he is asking what claims over land, labour, food, servants, horses, rents, clothing, books, travel, legal power, and social rank were represented by that monetary expression. The nominal figure is only the shell. The purchasing power is the organism. Once the two are confused, every subsequent act of economic reasoning becomes suspect, because the mind has begun treating the measuring instrument as the thing measured. The error is identical to an engineer mistaking the ruler for the bridge or a physicist mistaking the thermometer for heat. Such a scientist would be dismissed as incompetent, yet in monetary affairs entire governments, central banks, financial markets, and journalistic classes commit the same error daily with serene confidence and institutional applause.

The thesis of this article is therefore direct. Money is not wealth, value, capital, production, prosperity, or economic life. Money is the symbolic, institutional, and informational apparatus through which wealth is compared, exchanged, calculated, and transmitted across time and society. The distinction is not semantic. It is the difference between civilization and delusion. Wealth consists in productive capacity, capital structure, land, tools, machines, energy, knowledge, enforceable legal order, technical skill, entrepreneurial judgment, and goods capable of satisfying human wants. Money does not create these things. Money measures their exchange relations and enables calculation concerning their alternative uses. When a civilization forgets this, it begins to celebrate numerical expansion while consuming real capital; it applauds asset inflation while production withers; it praises valuations while mistaking them for value; it watches the scoreboard rise and congratulates itself, though the game on the field has already been lost.

The Austrian tradition remains indispensable because it begins where serious monetary theory must begin: not with the state, not with policy, not with banking aggregates, and not with the conceit of macroeconomic manipulation, but with human action under scarcity. Menger (1892) explained that money emerged through market processes as the most saleable good, not as a legislative gift descending from the throne. Mises (1953) explained that money’s purchasing power and its calculative role cannot be understood apart from exchange and prior market valuation. Hayek (1945) explained that prices coordinate dispersed knowledge more effectively than any planner could, precisely because no mind or committee can possess the contextual information held by millions of acting individuals. These are not museum pieces from an antique school of economics. They are foundations. They explain why a system of money is not merely a medium of exchange but a civilizational epistemology, a method by which individuals discover, compare, and act upon facts they could never centrally possess.

The argument becomes sharper in the age of distributed digital cash, because digital systems force monetary theory to confront governance in its naked form. A monetary system whose rules can be altered by a committee, foundation, cartel of developers, political agency, or managerial priesthood is not decentralized merely because its database is replicated across many machines. Replication is not autonomy. Distribution is not immunity from governance. A system is not rendered economically decentralized by spreading copies of a ledger while concentrating rule-making authority in a small class of administrators who can redefine the monetary unit, censor transactions, freeze claims, or change the conditions of exchange. Hayek’s problem was not merely the monopoly issue of who issues currency; it was the deeper institutional issue of whether money is governed by stable rules or by discretionary will (Hayek, 1976). Mises’s problem was not merely whether exchange occurs; it was whether economic calculation can remain rational when the unit of account is politicised, distorted, or made unstable (Mises, 1949). Distributed digital cash matters only insofar as it converts monetary rules from policy into law, and it ceases to matter the moment those rules become alterable by those who pretend that technical authority exempts them from economic reality.

I. The Ontology of Wealth and the Error of Monetary Reification

The first error to remove is the primitive superstition that money is wealth. This error survives because money is the universal form in which wealth is priced, recorded, inherited, taxed, borrowed against, insured, and litigated. The accountant sees pounds, dollars, francs, or satoshis. The investor sees balances. The bureaucrat sees aggregates. The journalist sees numbers. Yet none of these symbols is the productive reality underneath them, and no act of notation can transform a sign into the substance it denotes. A farm remains wealth because it can produce food. A factory remains wealth because it can transform inputs into valuable outputs. A trained surgeon, engineer, or mathematician possesses wealth in the form of rare human capital. A monetary balance is only a claim, and the economic importance of that claim depends upon the structure of production against which it may be exercised.

The ontological distinction between money and wealth is indispensable because it prevents the mind from collapsing measurement into existence. Smith (1776/1981) distinguished the wealth of nations from the mere possession of precious metals, arguing against mercantilist confusion that treated bullion as the essence of national prosperity. The same error now reappears in more sophisticated clothing. Modern societies do not necessarily worship bullion; they worship central-bank reserves, rising indexes, nominal GDP, housing valuations, fiscal stimulus, and balance-sheet magnitudes. The costume has changed. The metaphysical error has not. It remains the view that if a number grows, reality must have improved. Against this stands the brutal fact that real wealth is material, institutional, and intellectual before it is monetary. Printing claims upon bread does not bake bread. Expanding credit against factories does not maintain the machines. Raising the nominal price of houses does not create additional shelter. Arithmetic can describe production, but it cannot substitute for it.

This is where the Darcy example acquires its force. Austen’s £10,000 matters because it expresses a command over resources, not because the digits possess charm. To treat Darcy as less wealthy because the nominal figure appears small to modern eyes is to evaluate a monetary sign outside its historical exchange structure. The act is irrational in precisely the same way that judging an ancient mile by the emotional impression created by the word “mile” would be irrational. The question is not how the number feels. The question is what the number measured. A sophisticated reader must ask what rent rolls, land yields, service labour, transport costs, food prices, clothing, and social obligations could be commanded by such an income in the relevant institutional setting. Without that inquiry, the modern reader is not interpreting Austen but projecting inflationary ignorance onto the past.

Menger’s account of money’s origin begins by destroying the reified view of money. In his explanation, money emerges not because a sovereign labels an object “money,” but because individuals engaged in exchange discover that some commodities are more saleable than others (Menger, 1892). The less marketable good is exchanged for the more marketable good, not necessarily because the more marketable good is desired for direct consumption, but because it can more readily be exchanged again. Money therefore begins as an emergent solution to indirect exchange. It is not a metaphysical substance called “value.” It is a social institution produced by action, expectation, marketability, and repeated acceptance. This account matters because it roots money in human choice and exchange rather than in political magic. Money is not born when a ruler stamps a face on metal or when a legislature issues a statute. Such acts may shape monetary systems, but they do not explain why a monetary good becomes accepted in the first place.

The marketability principle also explains why money must be judged by its function rather than by its iconography. A thing used as money must carry purchasing power across transactions because participants expect others to accept it. This does not make money wealth in itself. It makes money the most exchangeable claim upon wealth. The difference is the whole science. A person holding money holds an option to acquire goods, discharge obligations, invest, lend, consume, or wait. The money is powerful because others will exchange real resources for it, but the wealth resides ultimately in those real resources and in the productive system that replenishes them. A dead economy can possess notes, tokens, coins, or ledgers; what it cannot possess is prosperity merely because those representations persist.

Mises sharpened this point through his analysis of money’s purchasing power and the regression theorem. In The Theory of Money and Credit, Mises (1953) argued that the purchasing power of money cannot be explained by circular appeal to its present monetary function alone; it must be traced back through prior valuations and market exchange. Whatever one makes of later debates over the theorem’s precise implications, its central lesson remains decisive: money cannot be understood as a pure abstraction floating above exchange. It derives its economic meaning from a history of valuation, acceptance, and calculative use. The state can impose legal tender rules, and institutions can impose settlement practices, but neither can abolish the underlying necessity that market participants believe the monetary unit will command goods in future exchange. Political force may compel nominal acceptance, but it cannot conjure real purchasing power from metaphysical emptiness.

The modern cult of numerical wealth ignores this history. It treats monetary expression as self-validating, as though the appearance of a larger number were a revelation of greater economic being. Yet every inflationary age teaches the contrary. The banknote with more zeros is not richer than the note with fewer zeros. The salary rising slower than food, rent, fuel, education, and capital goods is not a rising standard of living. The house whose nominal price doubles while the surrounding economy becomes less productive is not necessarily an instance of wealth creation. In each case, the mind that worships the number has departed from reality. It has accepted the superstition that signs create what they signify, which is merely primitive animism translated into the language of finance.

A Randian formulation is unavoidable here because the issue is epistemological before it is economic. Reality is not improved by altering the labels attached to it. To evade the distinction between the thing and the symbol of the thing is to place desire above fact and accounting above existence. The producer’s world is made of steel, wheat, code, energy, capital, labour, risk, time, and judgment. The bureaucrat’s world is made of categories, aggregates, directives, and claims against what others produce. A monetary system becomes corrupt when the latter begins to believe it has created the former by naming it. This is the inversion at the heart of modern money: the derivative claims authority over the primary reality, and the scoreboard declares itself the game.

II. Economic Calculation and the Necessity of a Stable Unit

The second stage of the argument concerns calculation. Wealth is heterogeneous. Land is not labour. Labour is not steel. Steel is not software. Software is not wheat. Wheat is not a ship. A ship is not a patent. A patent is not an electrical grid. The elementary fact of heterogeneity makes advanced economic life impossible without a method of comparison. A household may plan in physical terms at small scale, and a village may coordinate many activities through custom, obligation, and direct knowledge. An industrial civilization cannot. It must allocate scarce resources among millions of alternative uses across time, geography, uncertainty, and capital structure. That requires a common denominator, and money provides it.

Mises’s critique of socialism is inseparable from this problem because his argument was not merely political but calculative. In “Economic Calculation in the Socialist Commonwealth,” Mises (1920/1990) argued that without market prices for capital goods, rational economic calculation becomes impossible. The point is often caricatured as a slogan against planning, but the actual argument is deeper. Advanced production involves choosing among alternative uses of scarce capital goods, and such choices require monetary prices generated through exchange. Without those prices, the planner may possess engineering knowledge, physical inventories, and ideological enthusiasm, but he lacks the comparative framework necessary to decide which production structure economizes scarce resources. The calculation problem is therefore not that planners are stupid. It is that the necessary knowledge is not available in the form required for rational allocation.

Money is indispensable here because it compresses the comparison of heterogeneous goods into calculable form. An entrepreneur deciding whether to build a bridge, produce semiconductors, expand a railway, drill a well, or invest in a warehouse cannot compare all inputs directly in physical units. He requires monetary prices for labour, steel, land, machinery, capital, credit, transport, risk, and expected output. The money price does not make the choice automatically correct. It makes the choice intellectually possible. To abolish or corrupt the monetary frame is not to liberate production from capitalism’s vulgar arithmetic; it is to blind the producer while demanding greater precision.

This calculative function demonstrates why money must not be treated as arbitrary. If money is a unit of account, then its reliability determines the reliability of calculation conducted through it. A unit that changes unpredictably introduces noise into every economic judgment. The analogy to physical measurement is exact but insufficiently appreciated. A society would not permit engineers to build aircraft using rulers whose length changed according to parliamentary debate. It would not permit surgeons to administer medicine using dosage measures altered by committee expectations. Yet it routinely permits monetary authorities to alter the unit through which every investment, wage contract, pension promise, bond, lease, valuation, and long-term plan must be assessed. This is not merely policy discretion. It is epistemic vandalism.

Inflation is therefore misunderstood when treated merely as a rise in prices. Inflation is also a degradation of economic language. Prices are statements expressed in money. When the unit in which those statements are made loses stability, the meaning of the statements becomes harder to interpret. Some prices rise because goods have become scarcer. Some rise because demand has increased. Some rise because production costs have changed. Some rise because the monetary unit has been debased. Some rise because market participants expect further debasement. The entrepreneur must disentangle these causes while making decisions under uncertainty. Monetary instability makes the signal harder to read, and the consequence is not simply higher living costs but inferior calculation throughout the capital structure.

Mises’s broader theory of human action helps explain why this matters. Human beings act to replace a less satisfactory state of affairs with a more satisfactory one, and economic calculation permits them to compare alternative courses of action under scarcity (Mises, 1949). Money enables the actor to evaluate profit and loss, not as arbitrary capitalist rituals but as indispensable indicators of whether resources have been transformed into outputs more valuable than their alternative uses. Profit signals that consumers value the output more than the inputs in their competing employments. Loss signals that resources have been misallocated. Destroy or distort the monetary framework, and profit and loss become less reliable as guides. The moral consequence is not merely that investors lose money. The deeper consequence is that society wastes time, labour, capital, and human life on projects that should not have been undertaken.

This is why nominal growth can coexist with real deterioration. A government may expand expenditure, raise nominal GDP, inflate asset prices, and announce prosperity while the underlying capital structure decays. Roads may deteriorate, energy costs may rise, skills may decline, manufacturing capacity may weaken, and households may survive by consuming savings or borrowing against inflated assets. The monetary aggregates look alive. The productive organism is anaemic. Those who stare at the aggregates declare victory because they have mistaken the vital signs printed on a chart for the health of the patient. The Austrian lesson is colder and more demanding: economic life must be judged by real coordination, real production, real capital maintenance, and real satisfaction of wants, not by the theatrical expansion of nominal quantities.

The same point applies to personal wealth. A man whose salary doubles while the cost of shelter, food, taxation, transport, education, and energy triples has not become richer. A retiree whose portfolio rises nominally while the currency falls faster has not been rescued by markets. A worker whose pension promises are denominated in unstable units has been given numbers instead of security. The political class prefers nominal language because it conceals these losses. The citizen receives a larger figure and is invited to feel grateful. He is not meant to ask what the figure buys. He is not meant to ask whether the measuring stick has been shortened while he was applauding its decoration.

The Darcy paradox thus becomes more than literary evidence. It becomes a model of how inflationary societies lose historical consciousness. They cannot compare across time because the unit of comparison has been deformed. They cannot easily understand past wealth, past wages, past prices, or past capital, because the nominal language is deceptively familiar while the underlying purchasing power has changed radically. They read £10,000 and imagine £10,000 today, which is exactly the mistake. The sign resembles itself across centuries, but the reality it measures is different. A society that cannot understand this in literature will not understand it in policy. It will be fooled by every enlarged number offered as proof of prosperity.

III. Hayek, Prices, and Money as Distributed Knowledge

Hayek’s contribution enters at the point where calculation becomes knowledge. In “The Use of Knowledge in Society,” Hayek (1945) argued that the central economic problem is not merely allocation of given resources but the use of knowledge that is dispersed among individuals. This is the insight that should have ended the conceit of comprehensive economic control. No planner possesses the knowledge held in fragmented, local, tacit, and shifting form by millions of people. The farmer knows his soil. The engineer knows a material constraint. The shopkeeper knows changing demand in a street. The shipping manager knows a bottleneck. The consumer knows a preference not yet visible in official data. Economic order must therefore arise through mechanisms that permit coordination without requiring omniscience.

Prices are those mechanisms. A price is not a mere number hanging in space. It is a compressed packet of human knowledge, condensing facts about scarcity, demand, cost, substitution, expectation, risk, and opportunity into a form usable by actors who do not and cannot know the full circumstances that produced it. When the price of tin rises, Hayek’s famous point is that users of tin need not know whether the cause is increased demand or reduced supply; the price signal induces economizing behaviour without requiring full causal knowledge (Hayek, 1945). That is the genius of the price system. It economizes not only resources but knowledge itself. It allows partial minds to coordinate in a world too complex for total comprehension.

Money is the medium in which this informational system speaks. It is not merely one good among others once a developed monetary economy exists. It becomes the common language of relative scarcity and valuation. To corrupt money is therefore to corrupt the language of prices. The corruption may be gradual, technical, and hidden behind respectable institutional language, but its nature remains the same. If every price is a sentence in the language of economic coordination, monetary instability changes the grammar while pretending that communication remains unaffected. A people trained to accept this as normal has already surrendered the intellectual conditions of capitalism.

Hayek’s later work on monetary competition follows naturally from this epistemological framework. In Denationalisation of Money, Hayek (1976) argued that competition among privately issued monies could discipline issuers by allowing users to abandon inferior currencies. Whether one accepts every institutional detail of Hayek’s proposal is secondary to the central principle. Monopoly over money permits the issuer to degrade the unit while externalizing the costs across society. Competition forces monetary institutions to maintain credibility or lose users. Money, like every other institution, improves when the producer must answer to those who rely upon it rather than command them by privilege.

The modern state dislikes this conclusion because it exposes the essence of monetary monopoly. A monopoly issuer can convert monetary debasement into a hidden tax, fiscal relief, political discretion, and economic theatre. It can finance promises without immediate taxation. It can socialize losses through inflation. It can rescue favoured institutions while presenting the act as public necessity. It can transform prudence into punishment by eroding savings and reward leverage by inflating asset values. These are not accidental abuses. They are structural temptations. A monetary authority with discretionary control over the unit of account holds power over every contract denominated in that unit, and no vocabulary of stability, mandate, or expertise changes the nature of that power.

The informational view of money also exposes the poverty of many contemporary macroeconomic abstractions. Aggregates often conceal the heterogeneous structure of production that gives economic life its actual shape. An economy is not a bathtub filled with spending. It is an intertemporal, multi-stage, capital-intensive structure of plans, expectations, complementary goods, specialized labour, legal claims, and technological constraints. Treating it as a set of manipulable aggregates may be administratively convenient, but convenience is not truth. Hayek (1945) warned precisely against the pretense that knowledge available to no single mind can be replaced by statistical summaries. Aggregates may inform judgment, but they cannot substitute for the price-guided coordination of individual plans.

Austrian capital theory reinforces this point because capital is not a homogeneous blob. Böhm-Bawerk’s analysis of roundabout production and later Austrian capital theory emphasize the temporal structure of production, where goods at different stages contribute to future consumption through complex interdependence (Böhm-Bawerk, 1889/1959; Hayek, 1941). Monetary distortion affects this structure by falsifying the apparent terms on which long-term projects are evaluated. Artificially cheap credit, unstable monetary expectations, and manipulated interest rates can encourage investment projects that appear profitable only under distorted conditions. The eventual correction is then treated as a mysterious crisis rather than the revelation of prior miscalculation. The boom was not wealth. It was an accounting hallucination financed by corrupted signals.

This is where the rhetoric of prosperity becomes morally obscene. The productive individual is told that the system is healthy because indexes rise. He is told that inflation is manageable because official measures declare it within tolerable boundaries. He is told that his declining purchasing power is anecdotal, his housing costs are transitional, his energy costs are exogenous, and his savings losses are the price of macroeconomic management. Against this stands the concrete reality of his life. He works, produces, saves, and plans in units whose meaning is altered by authorities who do not bear the full cost of their decisions. The injustice is not merely distributional. It is cognitive. He is forced to live inside a language of value whose grammar is rewritten by others.

A Randian defence of money must therefore be a defence of reality against evasion. Money properly understood is not the root of greed but the instrument by which production is made exchangeable. It is the badge not of parasitism but of deferred choice, the claim earned by producing value for others and held until the producer elects what value to receive in return. The hatred of money often comes from those who resent the discipline it imposes: the demand that claims be measured, that losses be counted, that wishes confront costs, and that production precede consumption. Yet corrupt money deserves hatred of a different kind. Not because money is evil, but because corrupted money is the falsification of moral and economic accounting. It allows the unproductive to command resources through manipulation of the unit by which production is measured.

IV. Nominal Society and the Manufacture of Economic Illusion

The modern world’s obsession with nominal magnitude has created an entire civilization of scorekeepers who no longer understand the sport. Governments report growth, markets report capitalization, corporations report adjusted earnings, households report house prices, and commentators report billionaire rankings as though these figures were self-interpreting. They are not. Every number requires an ontology. Every valuation demands the question: value of what, measured against what, under what monetary conditions, supported by what productive facts, and redeemable in what real goods? A society that refuses these questions has not become sophisticated. It has merely learned to count without thinking.

GDP is a useful example because it is treated as a national vital sign despite being a monetary aggregate vulnerable to misinterpretation. Kuznets, who contributed to national income accounting, warned against conflating national income measures with welfare, and the later use of GDP as a proxy for prosperity often exceeds what such measures can responsibly support (Kuznets, 1934). The issue is not that GDP is useless. The issue is that monetary aggregates can grow through activity that does not necessarily represent improved human flourishing or sustainable capital formation. Reconstruction after destruction may increase measured output. Compliance costs may increase expenditure. Inflation may raise nominal values. Debt-financed consumption may mimic prosperity. The statistic records transactions within a framework; it does not automatically certify the moral or productive quality of what is occurring.

Asset markets exhibit the same confusion. Rising equity prices can reflect improved expected earnings, technological progress, superior productivity, or more efficient allocation of capital. They can also reflect monetary expansion, suppressed yields, speculative mania, accounting engineering, or the desperate search for stores of value in a depreciating monetary environment. The number alone does not answer the question. A market price is information, but information must be interpreted within institutional context. When the monetary environment itself is distorted, asset prices become harder to read, and financial wealth may detach from the underlying productive economy. At that point, the owners of assets may grow nominally richer while new entrants are priced out of capital ownership and productive investment becomes subordinated to financial arbitrage.

The housing market demonstrates the social cruelty of this illusion. A house is a consumption good, a capital asset, a collateral object, a political fetish, and in many countries the primary savings vehicle of the middle class. When house prices rise faster than wages because monetary conditions and credit structures inflate asset values, existing owners are told they have become wealthier. Yet the community has not necessarily gained more shelter, better construction, or greater productive capacity. It may simply have transformed future buyers into debt servants and converted land access into a speculative game. The nominal wealth of one class becomes the real exclusion of another. The scoreboard rises; the game becomes uglier.

This distinction also clarifies why inflation statistics so often fail to capture lived reality. Official measures may be methodologically coherent for defined purposes, but human beings do not live inside weighted abstractions. They live in households with specific consumption patterns, regional housing markets, career paths, medical needs, energy dependence, family structures, and savings goals. A young family facing housing inflation experiences monetary degradation differently from an asset-rich retiree. A producer dependent on fuel and inputs experiences cost pressures differently from a salaried bureaucrat. A saver experiences low interest and rising asset prices differently from a leveraged speculator. Aggregation can illuminate trends, but it can also conceal the distribution of monetary harm across different forms of life.

Nominal society is politically convenient because it allows rulers to offer symbols in place of substance. Wage increases can be celebrated while real wages stagnate. Benefit increases can be announced while purchasing power falls. Public spending can be expanded while service quality declines. Debt can finance consumption while being presented as investment. Monetary stimulus can inflate assets while being sold as broad prosperity. The citizen is meant to see the number and stop thinking. He is meant to confuse motion with progress and volume with value. The political class survives by exploiting precisely the cognitive weakness revealed by the Darcy paradox: the habit of reading the figure rather than the reality it commands.

A serious monetary theory must therefore be inseparable from a theory of truth. The issue is not merely that people misunderstand inflation. The issue is that monetary manipulation trains them to accept falsified signals as reality. It habituates the mind to nominalism, the doctrine that the name matters more than the thing. Once adopted in economics, the disease spreads. Credentials replace knowledge. Compliance replaces virtue. Public relations replace achievement. Valuation replaces production. The broader cultural pattern is one of symbolic substitution, and money is its most powerful instrument because money touches every exchange, every wage, every debt, every price, and every plan.

Against this decline, the Austrian insistence upon real choice, scarcity, and calculation appears almost severe. It refuses the narcotic of aggregates detached from action. It asks who produced what, at what cost, using which scarce means, in preference to what alternative, under what expectations, and with what demonstrated demand from others. It does not permit the planner to wave away opportunity cost. It does not permit the inflationist to call debasement prosperity. It does not permit the speculator to call every rising price production. This severity is not cruelty. It is respect for reality. Economics becomes humane only when it refuses to lie about the conditions under which human beings must act.

V. Distributed Digital Cash and the Constitutional Meaning of Monetary Rules

The question of distributed digital cash enters this argument not as a fashionable appendix but as the contemporary test of whether monetary theory has learned anything. A digital ledger by itself does not solve the problem of money. A replicated database does not abolish governance. A token does not become sound because it is surrounded by cryptographic vocabulary. The decisive issue is whether the monetary rules are stable, knowable, enforceable, and resistant to discretionary alteration. A system whose supply, transaction rules, validation conditions, or settlement principles can be changed by a privileged group is a governed monetary system, whatever mythology may surround its architecture.

This point is often obscured by careless invocations of decentralization. The term is used so loosely that it frequently conceals more than it reveals. A network may be topologically distributed while economically governed. It may contain many machines yet obey a small priesthood. It may possess open-source code yet depend upon a narrow class capable of defining which code counts as legitimate. It may describe itself as leaderless while following developers, foundations, exchanges, custodians, or mining interests whose coordination determines practical outcomes. The relevant question is not whether many people can run software. The relevant question is who can alter the rules that define the monetary unit and the validity of exchange.

A monetary system in which developers can change the protocol is not meaningfully outside governance. It has governance by developers. That governance may be informal, reputational, technical, or socially mediated, but it remains governance. To call such a system decentralized in the strong monetary sense is to confuse participation with sovereignty. The economic issue is not how many spectators hold copies of the rulebook. The issue is who can revise the rulebook while the game is being played. If the rules can be rewritten, then the monetary unit is not a stable measure but a managed artefact. It is policy wearing the mask of protocol.

Nakamoto’s formulation of Bitcoin as a peer-to-peer electronic cash system matters because the phrase identifies money as a transactional system, not a speculative idol (Nakamoto, 2008). The purpose of digital cash is exchange, settlement, and calculable use, not the creation of a floating theology of token appreciation. If the system becomes primarily an object of hoarding, narrative management, or protocol politics, it departs from the monetary function that made it significant. Digital cash must be cash. It must permit exchange under rules known in advance. It must function as a system of economic coordination rather than a committee-mediated casino whose monetary constitution is revised by those with influence over implementation.

The connection to Hayek is direct. Hayek’s monetary competition was competition among rule systems under market discipline (Hayek, 1976). A distributed digital cash system is valuable only if it gives users a monetary rule set superior to discretionary alternatives. If it reproduces discretionary governance in technical form, then it has not escaped the problem. It has merely changed the robes of the priesthood. The market does not need another committee-controlled token. It needs money whose properties are stable enough to support long-term calculation, contract, investment, and exchange. Stability here does not mean price immobility, which no market institution can guarantee. It means rule stability: the assurance that the conditions defining the monetary system will not be altered by political or technical convenience.

The connection to Mises is equally direct. Economic calculation depends upon reliable monetary prices generated within a framework of exchange (Mises, 1920/1990, 1949). If the unit, supply, or settlement rules are mutable, then calculation incorporates governance risk at the foundation. Participants must not only evaluate goods and services; they must evaluate the probability that the monetary system itself will be changed by insiders. This is not a minor technical concern. It is equivalent to asking every trader, saver, lender, and entrepreneur to price the possibility that the measuring instrument will be redesigned during use. Such a system may still be traded. It may still be profitable. It may still be fashionable. What it cannot honestly claim is the full moral and economic status of rule-based money.

The objection will be made that all systems require maintenance, and that immutability is impossible because software evolves. The objection fails because it substitutes engineering trivialities for monetary principles. A bridge may require maintenance without changing the definition of a metre. A court system may require clerks without granting clerks the right to rewrite property law. A monetary protocol may require implementation work without granting implementers authority to alter the monetary constitution. The relevant distinction is not between maintenance and abandonment. It is between preserving a rule and governing by revision. Those who blur that distinction do so because they want the prestige of law with the flexibility of power.

The deeper constitutional principle is that sound money must be boring to the ambitious. It must deny them the thrill of intervention. It must deprive administrators of the pleasure of relevance. It must convert monetary authority into rule-following rather than rule-making. This is why monetary systems governed by stable rules are morally superior to systems governed by discretion. They subordinate rulers, developers, and institutions to the same known constraints faced by users. They make calculation possible because they make the monetary environment less dependent upon personal will. They replace charisma with law and policy theatre with predictability.

A distributed monetary system that fails this test is not Hayekian in any serious sense. Hayek did not defend competition so that users could choose among rival committees of monetary managers. He defended competition because it disciplines issuers through user choice and penalizes debasement, unreliability, and abuse (Hayek, 1976). A digital system whose effective governance is captured by a small technical class has not realized that vision. It has created a new monetary bureaucracy without admitting that it is one. The fact that the bureaucracy speaks in code rather than statute does not change its nature.

This also clarifies why the phrase “not your keys, not your coins” is incomplete. Control over signing keys matters, but control over the rules defining what the signed transaction means matters more. A user may hold keys while the system’s governing class alters transaction policy, economic incentives, or settlement interpretation. Possession without stable rules is not sovereignty. It is tenancy under technical administration. The stronger monetary principle is that money belongs to the rule-bound order of exchange, not to the discretionary power of those who can change the terms after adoption. A system that violates this principle may be digital, scarce by convention, and widely traded, but it is not decentralized money in the sense demanded by Mengerian emergence, Misesian calculation, and Hayekian rule competition.

VI. Money as Moral Accounting

The moral importance of money is often misunderstood because critics treat money as though it were the cause of exploitation rather than the evidence of exchange. In a free market, money earned by production records the fact that one has created value for others. It is not the root of social decay. It is the abstract form of unconsumed achievement. It permits the producer to defer consumption, compare opportunities, invest, lend, give, save, and plan. To denounce money as such is to denounce the social recognition of productive service. The parasite hates money honestly earned because it exposes the difference between production and demand.

Yet this defence applies only to honest money. Corrupt money becomes false moral accounting. It allows claims upon production to be expanded without corresponding production. It allows political actors to spend without immediate taxation, financial actors to profit from proximity to issuance, and debtors to be relieved at the expense of savers. It blurs the relation between effort and reward. It permits the unearned to masquerade as earned. In Randian terms, it is the weaponization of unreality against the producer, because it forces productive men and women to denominate their lives in units controlled by those who may debase them.

This is why inflation is not merely an economic inconvenience but a moral event. It changes the relation between past production and future command. The saver who abstained from consumption did so in reliance upon the monetary unit as a claim on future goods. Debasement alters that claim. It transfers purchasing power without open contract. It punishes the temporally responsible and rewards those positioned to receive new money, leverage assets, or shift risk. The injustice is often hidden because no thief enters the house. The balance remains visible. The number may even increase. Only the command over reality has been reduced, and that is precisely why the crime is so politically useful.

The same moral logic applies to protocol mutability in digital monetary systems. Those who adopt a rule-based monetary system do so because the rules define the bargain. If those rules can later be altered by a governing class, then the adopters are no longer participants in a monetary order but subjects of monetary administration. The issue is not whether the administrators are clever or benevolent. The issue is that discretion has replaced principle. A system advertised as rule-bound but operated as committee-bound commits a fraud against the concept of sound money, because it borrows the moral authority of immutability while retaining the political convenience of change.

The demand for stable money is therefore not nostalgia for metal, hostility to technology, or romantic attachment to nineteenth-century banking. It is the demand that measurement be separated from manipulation. It is the demand that economic calculation be protected from those who benefit by altering the calculator. It is the demand that monetary institutions serve production rather than rule it. Menger, Mises, and Hayek converge on this point from different directions. Money emerges from exchange, makes calculation possible, and transmits distributed knowledge. To politicize or arbitrarily mutate it is to attack all three functions simultaneously.

The moral defence of distributed digital cash must be built on this foundation or it becomes empty marketing. Digital money is not morally superior because it is digital. It is not superior because it has enthusiasts, slogans, conferences, or market capitalization. It is superior only if it better preserves the functions of money: exchangeability, calculability, rule stability, and resistance to arbitrary debasement or alteration. If it fails those tests, then it is merely another speculative instrument dressed in revolutionary language. A casino with cryptographic chips is still a casino. A committee-controlled token is still committee-controlled. A protocol altered by its managers is still managed money.

This is where the Austrian tradition becomes more radical than most technological rhetoric. It does not ask whether the user feels sovereign. It asks whether the structure of action permits sovereignty. It does not ask whether the word decentralization appears in promotional language. It asks whether power over the monetary rules is actually dispersed, constrained, or eliminated. It does not ask whether the nominal supply is impressive. It asks whether economic calculation can proceed under stable and predictable rules. It does not ask whether the crowd believes. It asks whether reality supports the claim.

VII. Reconstructing Monetary Understanding

The reconstruction of monetary understanding must begin by disciplining language. Wealth should mean productive command over real goods, services, and capital, not merely high nominal balances. Value should mean the significance attached by acting individuals to goods under conditions of scarcity, not a mystical substance embedded in money. Price should mean an exchange ratio expressed in monetary terms, not an oracle detached from institutional context. Inflation should mean a corruption of purchasing power and price signals, not merely an inconvenience to be massaged through statistical categories. Decentralization should mean the absence of discretionary control over the monetary rules, not the mere multiplication of network participants who remain subject to a governing core.

This linguistic discipline is not pedantry. It is the beginning of thought. A civilization that uses words carelessly will be governed by those who exploit the ambiguity. Call monetary expansion stimulus, and theft acquires a doctor’s coat. Call developer governance decentralization, and administration acquires the romance of rebellion. Call asset inflation wealth creation, and exclusion becomes prosperity. Call nominal GDP growth progress, and the consumption of capital hides beneath national celebration. Every false monetary term is an unpaid debt to reality, and reality collects.

The Darcy paradox should therefore be taught not as literary trivia but as the first lesson in monetary epistemology. The question is not how rich Darcy sounds to us. The question is why modern readers are so easily deceived by the persistence of a symbol across changing monetary conditions. That mistake trains the mind to ask the correct question in every monetary context: what does the number command? This question cuts through propaganda with almost indecent efficiency. What does the wage command? What does the pension command? What does the asset command? What does the currency command? What does the token command? What does the monetary rule allow its holder to know, plan, and exchange?

From that question follows a second: who controls the unit in which command is measured? In state money, the answer lies in central banks, fiscal authorities, banking systems, and legal tender regimes. In governed digital systems, the answer lies in developer groups, foundations, exchanges, miners, validators, custodians, or other concentrated actors whose coordination determines rule evolution. In sound rule-based money, the answer must be: no discretionary authority controls it. Rules may be enforced, but not opportunistically rewritten. Participants may compete, but not redefine the unit for others. The difference is the difference between a constitution and a committee memo.

The Mengerian test asks whether the money emerges through exchangeability and marketability rather than mere decree. The Misesian test asks whether it supports economic calculation through reliable prices and a meaningful unit of account. The Hayekian test asks whether it allows dispersed individuals to coordinate through signals superior to centralized direction. The constitutional test asks whether its rules are stable enough that users can plan without pricing the ambitions of those who might alter them. A monetary system that fails these tests may still be profitable, popular, or politically useful. It is not sound money.

The modern tragedy is that many people no longer demand sound money because they no longer know what money is. They demand rising portfolios, easier credit, higher nominal wages, cheaper borrowing, larger fiscal transfers, and protection from every correction caused by prior distortion. They want the scoreboard raised without asking whether the team has played better. They want to be richer by declaration. They want reality to yield to accounting. When this psychology dominates, monetary corruption is no longer imposed from above; it is invited from below. The public becomes complicit in its own deception because the lie is pleasant and the truth requires production.

A serious society would reverse this hierarchy. It would teach that production precedes consumption, that saving is deferred choice, that capital must be maintained, that prices carry knowledge, that money measures rather than creates, and that the unit of account must not be treated as a toy of policy. It would treat monetary debasement not as clever management but as falsification. It would treat protocol mutability in digital money not as innovation by default but as a constitutional danger. It would refuse to call systems decentralized when identifiable groups can alter their defining rules. Above all, it would insist that economic language answer to reality rather than power.

Conclusion: The Game Beneath the Scoreboard

The scoreboard is not the game, and the refusal to understand this is the root of modern monetary stupidity. A rising number may reflect genuine achievement, but it may also reflect debasement, speculation, leverage, monopoly privilege, regulatory distortion, or mass delusion. The number itself does not absolve the mind of judgment. It demands judgment. It demands the hard question that every inflationary culture tries to suppress: what real thing stands behind the symbol? Without that question, economics becomes numerology with institutional funding.

Menger shows that money arises from the market as a solution to exchange, not from political incantation. Mises shows that money enables calculation among heterogeneous goods and across complex capital structures, making rational production possible under scarcity. Hayek shows that money and prices transmit dispersed knowledge through society, coordinating plans no central mind could comprehend. Together, they establish a conception of money as an informational, calculative, and institutional achievement. Money is not wealth but the language through which wealth speaks in exchange. To corrupt that language is to corrupt economic thought itself.

The Darcy paradox endures because it exposes the modern mind at its weakest point. The modern reader sees £10,000 and thinks in nominal terms. Austen’s world understood command over reality. That lost instinct explains why societies can be fooled by rising asset prices, manipulated aggregates, inflated wages, debased savings, and digital systems that call themselves decentralized while preserving rule-making elites. In each case, the error is the same. The symbol is treated as substance. The measurement is treated as the measured. The committee is treated as the market. The scoreboard is treated as the game.

Distributed digital cash is important only if it restores the rule-bound nature of money. If its protocol can be altered by developers, administrators, or coordinated insiders, then the system is governed and controlled in the economically relevant sense. It may be technologically interesting, but it is not decentralized money in the Hayekian or Misesian sense. Sound money requires rules that do not bend to managerial convenience. It requires a unit that permits calculation rather than demanding political interpretation. It requires law, not policy; measurement, not manipulation; exchange, not theatre.

The central battle is therefore not between old money and new money, metal and code, banks and networks, or paper and tokens. The central battle is between reality and evasion. Money must either serve as an honest instrument of economic calculation or become a weapon of illusion. It must either measure wealth or be used to counterfeit the appearance of wealth. It must either transmit knowledge or corrupt it. No civilization can evade that choice indefinitely, because production is not fooled by numbers. Goods do not appear because ledgers expand. Capital is not maintained because valuations rise. Wealth is not created because a committee changes a rule and calls it progress.

Reality is patient, but it is not negotiable. A society may lie about money for a long time, and the lie may enrich those closest to the machinery of alteration. It may decorate the lie with equations, mandates, white papers, governance forums, development roadmaps, and speeches about innovation. Yet the final audit is always conducted by production itself. Can the money measure? Can the prices inform? Can the rules be trusted? Can individuals plan, save, exchange, and calculate without submitting to the discretionary will of those who control the unit? If the answer is no, then the system has failed no matter how loudly its defenders chant the slogans of freedom.

Money is not the game. It is the scoreboard, the language, the measuring rod, and the instrument of calculation by which the game is made intelligible. The wealth of society lies in the minds, machines, fields, factories, networks, laws, skills, and productive actions of human beings. Money at its best honours that reality by allowing it to be exchanged and calculated. Money at its worst conceals that reality beneath a fog of manipulated numbers. The task of monetary theory is to defend the former against the latter. The task of a serious civilization is to know the difference before the scoreboard glows triumphantly over an empty field.

References

Austen, J. (2008). Pride and prejudice. Oxford University Press. (Original work published 1813)

Böhm-Bawerk, E. von. (1959). Capital and interest (G. D. Huncke & H. F. Sennholz, Trans.). Libertarian Press. (Original work published 1889)

Hayek, F. A. (1941). The pure theory of capital. University of Chicago Press.

Hayek, F. A. (1945). The use of knowledge in society. The American Economic Review, 35(4), 519–530.

Hayek, F. A. (1976). Denationalisation of money: An analysis of the theory and practice of concurrent currencies. Institute of Economic Affairs.

Kuznets, S. (1934). National income, 1929–1932. National Bureau of Economic Research.

Menger, C. (1892). On the origin of money. The Economic Journal, 2(6), 239–255.

Mises, L. von. (1949). Human action: A treatise on economics. Yale University Press.

Mises, L. von. (1953). The theory of money and credit (H. E. Batson, Trans.). Yale University Press. (Original work published 1912)

Mises, L. von. (1990). Economic calculation in the socialist commonwealth. Ludwig von Mises Institute. (Original work published 1920)

Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash system.

Smith, A. (1981). An inquiry into the nature and causes of the wealth of nations. Liberty Fund. (Original work published 1776)