The Copyist's Eden
Why the case against intellectual property, taken to abolition, hands the field to whoever already owns the channels — and why creation under scarcity is a problem of institutions, not slogans
Keywords: intellectual property; copyright; patents; appropriability; complementary assets; platforms; pharmaceutical innovation; fixed costs; dynamic efficiency; institutional economics; Kinsella; Boldrin and Levine; plutocracy.
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