Journal of Law, Information and Science
WENDY J. GORDON [*]
The ‘public goods’ characteristics possessed by intangible works of authorship and invention present the basic market failure problem usually relied on to justify intellectual property rights. What is ordinarily less emphasized is that such market failure is no more than half of the prerequisite for an economically desirable copyright or patent system: another requisite condition is that there be less costly market imperfections after intellectual property is instituted than there would have been in the absence of the intellectual property regime. Intellectual property rights are best justified in the presence of "asymmetric market conditions", that is where (1) in the absence of intellectual property rights, there will be market failure; and (2) in the presence of intellectual property rights, there will be market success.
The article compares areas of copyright law in which systemic rules have been used to minimize market imperfections and to further market formation with other areas of copyright in which doctrines such as the fair use doctrine and notions of reasonableness have been used to develop case by case solutions. In particular, the article examines whether face to face refusal to license can ever constitute the kind of market failure which would justify refusing to enforce an intellectual property right. The latter problem arises when copyright owners try to suppress hostile or unorthodox adaptations of their work. The device of “asymmetric market conditions” is also used to examine the presumption within restitution law that people who refrain from crossing others’ tangible boundaries are free to take advantage of each other’s labor. The article then uses the same conceptual tools to explain why intellectual property law reverses this presumptive freedom and replaces it with a duty not to copy.
This piece seeks to unify various problems in intellectual property law by making use of a single perspective, that of market failure. Under an economic view, the issue of whether intellectual property protection should be adopted for a given industry or intangible should depend largely on an institutional comparison of costs and benefits. Put differently, we can analyze the question of whether adopting intellectual property rights will produce a net allocative gain by asking: would the market imperfections of a regime possessing intellectual property law be more or less costly than those imperfections which would otherwise be present? This is ultimately an empirical question, but it is possible to suggest some plausible hypotheses. The most fundamental hypothesis is this: for a market-based intellectual property system such as copyright and patent to be economically justifiable, there must be what I call an "asymmetry of market conditions." Put most simply, such asymmetry appears where:
1. In the absence of intellectual property rights, there will be market failure; and
2. In the presence of intellectual property rights, there will be market success.
The reader is probably familiar with the basic market failure problem cited as the usual justification for intellectual property regimes: it is usually termed the "public goods" problem, and is briefly described below. What is less recognized is the fact that such market failure is only half of the prerequisite: the other requisite condition is that there be less costly market imperfections after intellectual property is instituted than there would have been in the absence of the intellectual property regime.
"Public goods" are defined by having two characteristics, inexhaustibility and nonexcludability. Most intangibles have these characteristics to some extent. Intangibles tend to be inexhaustible over a large range of utilization (everyone can sing or play the same song, or build the same design of engine, without interfering with others' use). They also tend to be difficult for proprietors to fence off (once encountered and remembered, anyone can reproduce the song or design). If production of a public good is left entirely to the private market, lack of fencing can lead to under-production, for usually a producer needs a mode of excluding non-payors if he or she wishes to obtain payment for what he or she has made. Without a mode of exclusion, the payoff from investing in creative activities will be low, and incentives will be inadequate to induce production of as many new intangible goods as the public would be willing to pay for.
Some public goods, such as national defense, can be produced through use of a state apparatus. This approach has the virtue of responding to both "public goods" characteristics. State production (1) takes advantage of inexhaustibility by making the benefit available to all, and (2) resolves the problem of underproduction by requiring everyone, through taxes, to pay.
The United States is committed to the belief that sole reliance on state-directed production or bureaucratic subsidy is not the best way to produce inventions and art. In the realm of inventions, probably the most obvious danger of state control is the bureaucratic tendency to resist innovation. In the realm of cultural products, the most obvious dangers of state control are "lack of taste,” and the possibility of censorship. (A history of free enterprise, of course, also plays a role.) Whatever the reason, there is a consensus in the United States that a diversity of private initiatives needs to be enlisted, and that state production-- despite its ability to respond to both inexhaustibility and nonexcludability-- should not be the primary route to follow in regard to inventions and art.
Thus, the United States, like most western nations, opted for primary reliance on use of the private market to generate incentives for the production of inventions and art. Using a market requires curing the excludability problem. Some scholars have argued that significant modes of exclusion are available independent of the law. For instance, even in a legal system without copyright (one might call such a system "copy liberty"), a writer and her authorized publisher could obtain payment through exploiting natural levers such as the advantage of being first in a market or having a reputation for providing authentic, distortion-free texts. But within the first years of the American republic, its Congress decided to provide legally enforceable rights of exclusion by enacting intellectual property laws such as copyright and patent. These laws give individual creative persons the right to forbid copying of their works.
This right provides remuneration, and thus an incentive to produce new works, in two primary ways. First, it provides creators with leverage in dealing with their first and primary customer: it induces potential publishers or manufacturers to pay in order to lift the "no-copy" ban. Second, it makes publishers and manufacturers willing to pay meaningful sums for the privilege of copying because the exclusive right provides some protection against unauthorized competition from outsiders.
Thus, intellectual property law responds primarily to the second "public goods" characteristic -- difficulty of fencing-- and does so by altering that characteristic by legal fiat. The law provides fences, which in turn assist the producers in capturing for their own pockets some of the benefits their efforts generate. The system relies on the premise that such enrichment will induce new investment in creative endeavor and that enough new investment will be created--investment that would not otherwise exist-- that the value produced by this investment will outweigh the extra administrative and other costs of the intellectual property system.
Even before we begin, we know that an intellectual property market cannot be "perfect," if by perfection one means a market where everyone willing to pay more than a good's marginal cost is able to purchase the good. No matter how low the marginal cost of an intangible might be-- and it might be as low as zero-- some people who are willing to pay for the good at that level or above will not have access to it, for intellectual property law allows producers to demand payment in excess of marginal cost.
This is desirable for incentives, for it is hoped that the price above marginal cost will cover research and development expenses and induce other potential producers to make new intangibles. Nevertheless it is clear that a right of exclusion, although the core of intellectual property law, is only a partial response to intangibles' public goods characteristics, for such law leaves the public unable to take full advantage of the inexhaustibility of intangibles. Instead, the failed promise of inexhaustibility merely exaggerates the deadweight loss that is a cost for all monopolies.
In addition to providing a right of exclusion, a successful intellectual property system should also be tailored to take as much advantage of inexhaustability as possible. If something can be copied at no cost, there must be some instances in which allowing free copying will be Pareto-superior. As will appear, the American legal system makes some effort, albeit minor, to recapture the lost promise of inexhaustability.
For an intellectual property regime to have even a chance of producing more allocative gain than a "copy-liberty" regime, the intellectual property regime must produce resource packages that are tradable. It must also minimize its deadweight costs and other imperfections. In the following, the article will explore some of the devices that American copyright law employs to keep its imperfections at a minimum.
Some of these devices are system-wide responses, where sharp rules are set out. Adjudication under sharply defined rules is sometimes known as using a "property" or "formal" approach. Copyright law also has another kind of response: employment of open-textured rules which require case-by-case substantive inquiry. The latter kind of devices might be said to employ a "nonformal tort" or "reasonableness" approach; in them, a court typically makes a substantive judgment as to the desirability of the defendant's challenged behavior.
In American common law, violation of most property or personal rights will be termed a "tort." Yet torts themselves tend to fall into two broad categories: intentional torts like battery or trespass, which are ordinarily actionable without proof that the defendant's specific behavior was socially undesirable, and unintentional torts, which are ordinarily actionable only if the plaintiff shows that the defendant's behavior was negligent or otherwise unreasonable. As a matter of categorization, the first type of tort can be viewed as following a "property" or "formal" model and the second as following a "nonformal tort," "reasonableness" or "substantive" model.,
In the "property" or "formal" model, the courts defer to the property owner as if she were a mini-sovereign, making no inquiry into whether the owner's decision to exclude a defendant was proper or improper, or whether the defendant's use of the owner's resource was harmful or productive. A classic example under American law is trespass to land. Someone who enters land reasonably but mistakenly thinking he has the right to do so will be liable as a trespasser, as will someone who entered the land out of a pressing need to save time by a shortcut. Following out the analogy to sovereignty, the primary relevant question in these cases is essentially jurisdictional: inquiring into whether the defendant crossed a boundary over which the owner possessed an exclusive right and did so without obtaining the owner’s consent. If so, the defendant has broken the relevant rule and is liable.
By contrast, in the "nonformal tort" or "reasonableness" model, a court does not assume as a prima facie matter that deference is owed to the decisions of the property owner. Instead, the court makes its own substantive inquiry into the desirability of the defendant's boundary-crossing; further, the burden of proof will likely be placed on the plaintiff to satisfy the court that the defendant's behavior was wrongful. A classic example in the United States is negligence law: in unintentional auto accidents, unless a defendant is found to have lacked "due care," she will not be required to pay for the damage she caused.
American copyright law follows an uneasy middle course between the rule-bound "formal" model and the principle-based "reasonableness" model. On the one hand, virtually any unauthorized substantial copying of a protected subject matter is subject to a prima facie prohibition. For example, even "unconscious copying" gives rise to liability; similarly, if a clever plagiarist convinces a magazine publisher that a short story is original, the publisher's good-faith belief that she had permission to print will not help the publisher avoid liability in a copyright infringement suit. This partakes of a formal or "property" approach.
On the other hand, the fact-finder (usually the jury) has both latitude and significant normative responsibility in deciding how much similarity amounts to "substantiality." In any case that involves other than exact copying, a defendant will probably try to argue that his work is not "substantially similar" to plaintiff's copyrighted work. The openness of the resulting inquiry involves a "reasonableness" approach.
Another and perhaps more important instance of a "reasonableness" model in American copyright law is the doctrine of "fair use," which is available to defendants even in cases of substantial copying. A defendant will have no legal liability for making a copy which is "fair" in her particular circumstances. Although the fair use doctrine appears in the Copyright Act, that statute declines to set out any definite strictures. Formal line-drawing is rejected. As the legislative history recounts, "[S]ince the doctrine is an equitable rule of reason, no generally applicable definition is possible, and each case raising the question must be decided on its own facts." The fair use defense essentially draws the court into deciding the social desirability of the defendant's copying.
At first blush, it is surprising to find American copyright law placing at the virtual center of a plaintiff's case the distinctively nonformal principles of "substantiality" and "fairness." After all, copyright is a form of property, and copyists always act volitionally and deliberately. (For example, they know they are publicly performing, or using the photocopy machine, or playing music on their guitar, even if they don't know that they are copying someone else while doing so.) It seems most logical that such a volitional trespass should be treated under a formal rule, as are other non-accidental violations of property rights. The mere act of nonconsensual copying is arguably like the mere act of stepping onto someone else's land without permission, and arguably should give rise to similar liability. So why is this not the case?
Economics does not yield an immediate answer. The classic article by Calabresi and Melamed tells us that intentional takings of property are prima facie wrongful because people should not depart from the market without a strong justification. Their article tells us that accident law uses a “reasonableness” inquiry because, in accidents, such a justification is present: the participants cannot bargain with each other in advance. An hour before a driver hits a pedestrian, for example, neither one knows she has any reason to deal with the other. In the presence of such complete market failure, we are told, the court "mimics the market" through a negligence inquiry, trying to determine whether the parties behaved efficiently and imposing liability accordingly.
But in copyright the "substantiality" inquiry and the "fair use" doctrine are both available regardless of whether the copyright owner and the copyist have complete knowledge of each other's identity and are otherwise able to bargain. Thus, in the most recent “fair use” case to reach the United States Supreme Court, the parties had the requisite knowledge and transaction costs were minimal. Defendant, the rap group "2 Live Crew," had in fact offered the copyright owners compensation in exchange for permission to make a parody of their song, " Pretty Woman." The copyright owners simply refused to give them permission. This hardly looks like market failure -- the two parties were virtually face to face. Yet the Supreme Court indicated that fair use might nevertheless be available to shield the makers of the parody from liability. How, then, can such a doctrine be squared, either with usual American patterns of tort and property law, or with economic notions of market failure?
The answer lies in the imperfection of intellectual property as a response to the "public goods" problem. In the absence of perfect price discrimination, obtaining adequate incentives for production will necessarily involve a price that is set above marginal cost, and thus a quantity produced that is below the quantity that would be produced by a competitive market. Market imperfection is present in even the most pristine copyright transaction. The issue is how legal institutions cure the imperfections caused by excludability without losing the benefits excludability brings.
Or, one can put the same matter in non-economic normative terms. Copying is not necessarily wrongful. It is how we learn, it can be both harmless and beneficial, and it is the essence of having a common culture. Therefore it would be absurd to make all copiers prima facie liable as infringers.
The trick is to find some means to distinguish wrongful from fair copying. Some of those means are case-by-case, like "substantial similarity" and "fair use." Some are system-wide rules.
Western culture has long recognized the tension between the law's need to speak clearly to cover broad classes of cases, and the desire to do justice in the individual case. In English translations of Aristotle, the term "equity" is typically given to judges' power to modify the law to take account of individual cases.
Sometimes the choice among legal responses is dictated by history, mores, or political pressures. Yet one can also attempt to make an economic judgment comparing the costs and benefits of the systemic (formal) approach, with the costs and benefits of the case-by-case (reasonableness) approach that characterizes the Aristotelian notion of equity. An economic perspective might help a student of comparative institutions parse the various influences that have affected different nations' choices among modes of dealing with a common challenge.
Professor William Powers has well summarized the key costs and benefits of the two approaches. The formal approach "enhances predictability, is easier to apply, and controls bias or other sources of error in the decisionmaking process." A formal approach also has costs, most importantly the "social cost of tolerating uses that do not maximize aggregate welfare in the short run."
The reasonableness approach has the converse characteristics. Thus, its prime virtue is that it allows the law to fine-tune its results to match circumstances, so that socially valuable behavior is not cut off or penalized by overbroad rules. On the other hand, the unpredictability of a reasonableness approach keeps this virtue from being fully realized. An equitable approach can chill desirable behavior, because potential actors do not know as clearly as they would under formal rules what behavior is permissible and what is impermissible. A reasonableness approach also entails higher administrative costs than a formal approach. This expense is largely due to the extensive fact-finding required to make a judgment of social desirability.
It is sometimes thought that market dealings will be best facilitated by definite rules. However, work in the economics of entitlements by Jason Johnston suggests that in some circumstances parties bargain more efficiently in response to uncertain reasonableness rules than to certain ones. So it is yet unclear whether an equitable, nonformal lack of definiteness is desirable or undesirable in this regard.
As for the likelihood of error, Professor Powers is correct that the nonformal reasonableness approach, because of its imprecision, is prone to error. For example, in copyright cases, judges often disagree among themselves about whether a given use is "fair." Also, because of the latitude a reasonableness approach allows, it is capable of being perverted by bias. Yet these errors or biases will be less costly than an erroneous or biased judgment enforced on a system-wide basis. A formal approach may reach technically correct answers to the questions it poses ("Did you copy X's property? Yes or no.") but the questions posed may themselves be incomplete, or altogether the wrong questions to ask.
For example, in property law, a formal approach is economically justifiable so long as Adam Smith is correct that owners act automatically as stewards of the social good. But when the invisible hand fails-- which is why "market failure" matters-- trusting owners' self-interest may not achieve efficiency. Under intellectual property law, there is an inevitable market failure--namely, deadweight loss-- and, in many cases, additional problems such as externalities, wealth effects, and transaction-cost barriers. In such instances, it can be quite unwise to assume that formal deference to a property owner's judgment will bring social prosperity. Systemic rules (such as copyright and patent's limits on duration) and case-by-case standards (such as "fair use") work to cut back the intellectual property owner's dominion.
As I have argued elsewhere, the best economic case for intellectual property can be made out when "asymmetric market conditions" are present: market failure in the absence of intellectual property rights and market success in the presence of such rights. The first condition is an old if controversial story, centering around public goods. This article will focus more on the second condition. The second condition is that once a no-copy rule is put in place, licensing will evolve. In other words, this second condition is met if, in the presence of a copyright or some other rule restraining strangers from copying, markets will succeed, not fail.
Markets are believed to be generally a more accurate and less expensive way to allocate resources efficiently than are courts or administrative agencies. Market formation is also relevant to incentives, for monetary payments will not come to creators unless potential users are able to bargain around the law's restrictions and pay for licenses or copies.
No matter how otherwise desirable it may be to have a copyright, patent, or misappropriation system, the arguments in favor of that system from an economic perspective are empty unless markets come into being. Without markets in which to sell their work, the people who own the intellectual products will be unable to obtain fees, they will therefore lack incentives, and, as a result, fewer new works will come into being. In addition, unless markets are forthcoming-- a publisher contracts to distribute copies, a movie maker is licensed to adapt the book-- the public will be denied even the use of the intellectual products that have been made.
In what follows, the article will review systemic devices that the copyright system uses to enhance market trades in intangibles, notably the dynamics of the exclusion right. It will then examine systemic devices that the copyright system uses to reduce costs such as deadweight loss. It will then return to "fair use," which is, as mentioned, a case-by-case inquiry into the substantive desirability of particular uses.
Intellectual property laws use several system-wide rules to enhance markets. Most such laws define ownership, prescribe modes of transfer, and, by requiring notices, providing registries, and the like, make it easier for potential buyers to find potential sellers or licensors. Further, intellectual property laws define boundaries by stipulating those subject matters which are capable of being owned and those rights over which an owner is granted exclusivity. Also, rights are assigned to what Clifford Holderness has defined as a "closed class," a definite group (authors and inventors, or their employers or assignees) rather than to the public, an "open class" from which determinate transfers can be made only with extreme difficulty. Further, the copyright statute provides some market bridges to assist in circumstances of market difficulty. Compulsory licenses are a prime example. At the center of the system lies the grant of exclusive rights that empower copyright owners to forbid others from certain behaviors that involve use of the copyrighted work. That will be the focus of the next section.
Intellectual property law is largely a matter of internalizing positive externalities. That is, it seeks to create a legal structure that will direct payment for the benefits a creative work generates into the pockets of persons who created the products so that they will have an incentive to create more such beneficial products in the future. Restitution law in the United States similarly connotes restoring to someone a payment for a benefit they have generated: it is generally known as the law of unjust enrichment. Like intellectual property, restitution law is also a law of benefits, only it is governed primarily by case decisions, unlike copyright and patent which are statutory. A more important difference between the two doctrines is in their baseline assumptions.
In the United States, a basic restitution doctrine refuses to award payment to "officious intermeddlers" and "volunteers." This doctrine provides that persons whose labor makes others better off will ordinarily have no legal recourse if they labor without advance agreement. Yet intellectual product producers can sue to obtain payment for the "fruits of their labor" from copyists who never agreed to pay. Since restitution law contains no presumption that there should be recovery for benefits generated, it forms a useful contrast with copyright.
First, consider a homely example to illustrate the different treatment a laborer without a contract will receive under the two areas of law. First, imagine someone paints the roof of a building while its owner is away. After the owner returns the painter presents himself at the door and says, "pay me for this wonderful benefit I have given you," pointing at the new paint job which (we will assume) has increased the value of the building. In such a situation, the building owner is entitled to say something quite rude. Next, in contrast, imagine that the home owner, using her own sweat and paint, does a mural on one of her building's exterior walls, copying onto it a photograph which has a valid copyright. Perhaps the mural increases the value of the building more than a new coat of pure white paint would have; perhaps the mural is an eyesore. In either event, if the owner of the copyright comes to the door and says, "Pay me or I'm going to sue you for a very large amount of money," the building owner had better be very polite.
In neither situation has the building owner agreed in advance to pay for use of the other's resource. Yet the photographer's labor embedded in a visual pattern must be paid for, and the painter's labor embedded in new roof pigment need not be.
To prevail in restitution, persons whose voluntary actions provide benefits to others must ordinarily show one of a few very narrow justifications for departing from the market: mistake, coercion, request, or a narrow range of exigent situations, such as danger to life and health. Even then, their ability to recover will often be further restricted by the courts' desire to be sure that the defendant really was benefitted and that forcing him to pay or disgorge will not leave him worse off than he would have been in the status quo ante. Similarly, the Restatement of Restitution is not hospitable to persons who generate benefits as a by-product of self-serving activity. Thus, the Restatement states that:
A person who, incidentally to the performance of his own duty or to the protection or improvement of his own things, has conferred a benefit upon another, is not thereby entitled to contribution.
For example, a mine owner whose drainage efforts clear both her mine and her neighbor's mine of waters is not entitled to contribution from the neighbor.
A person who writes a book and publishes it is certainly operating in the furtherance of his or her own interests. Except as to someone who has bargained with the author for production of the work (such as a patron, granting agency, employer, or contract-publisher), the author is a sort of volunteer. When a book is mass-marketed, many strangers will come across it. If a stranger makes copies of the book for sale, copyright law will give the author a right of action against the copyist even if the author "volunteered" to send the work into the stream of commerce. Since that right of action will be available whether or not the copyist had a contract with the author promising to refrain from copying, and whether or not the copyist's actions harm the author, it is clear that, under copyright law, a unilateral transfer of "benefits" is sufficient to trigger liability. There are many reasons for the difference between the two fields' basic rules; one obviously lies in the active or passive role of the person using the benefit. That is not simply an issue of autonomy. Activity or passivity also has implications for market formation. To see this, consider what results would follow if a legally enforced right to payment were given to the claimants in the two situations.
In the restitution context, the active parties are the benefactors, the volunteers. Systematically allowing volunteers to sue for the benefits they have given would reduce their desire to make contracts with those who might want their services. Admittedly, the people who are the best at painting houses would have no desire to sneak around and do it behind the backs of their customers. However, the people who make messy jobs of it would probably start to paint and then ask for money after the fact-- and could do so in disregard of whether the building owner preferred a different supplier, thus ruining the market even for those who would otherwise be willing to make contracts. In the volunteer context, then, a rule that encourages contract formation-- and thus market formation-- is a rule that denies to the benefit-generator (the potential volunteer) any right of recompense independent of contract. If a volunteer thinks the law will not give restitution, then she will seek to make a bargain by asking the potential recipients for contributions before the project begins.
In the intellectual property context, the likely impact of a right to recover is quite the opposite. This occurs largely because the identity of the active party-- and thus the party who has superior access to information, who is otherwise better able to enter transactions, and who is better situated to respond to the law's messages-- is different there.
In the volunteer context, the recipient may be ignorant until the deed is done. It is the benefactor-- like the house painter-- who has the greater access to information; he knows where and when he will act. In restitution, the rule of law that speaks to this active party and encourages him to seek out consensual market arrangements is therefore a rule of "no monetary recovery without contract." In the intellectual property situation, by contrast, the recipient-copyist is the active party: he can better initiate the transaction. After all, the copyist knows what he is copying, whereas the plaintiff-owner may be hundreds of miles away and have no idea copying is being contemplated. The copyist will also find it fairly easy to identify the author or copyright owner from the by-line, while the copyright owner has no such source of information.
Even if the author can identify the potential copyists, she faces strong strategic behavior problems in making them pay. A baseline rule that denied copyright-- that gave a benefit-generator no recovery unless he had a prior contract with the copyist-- would leave the party with the best ability to contract (the copyist) with little motive to do so. He would probably prefer to free ride. Therefore, the law has to give the benefit-generators (the copyright owners) a right of recovery independent of contract.
Both restitution and intellectual property law give the party who has the information and ability to internalize the incentive to do so. The party simply happens to be different in restitution than in intellectual property. In each case, legal rights are arranged to facilitate the consensual transfers of resources. The party best positioned to alter the use to which a resource is put is required to do so by a systemic choice of a liability or no-liability rule. 
Systemic rules are not only used to enhance tradability. They are also used to distinguish between areas where legal protection is desirable and areas in which it is not. Industries may exist where perhaps the need for government to provide exclusion rights is less (because there is no significant market failure under copy-liberty), or uses which are more expensive to restrict, or other areas where the costs of copyright might outweigh its value. Thus, as to subject matters, not all beneficial products of human ingenuity are capable of being owned. For example, the copyright act grants ownership only in works of authorship fixed in a tangible medium of expression, and even among such works the act denies protection to whole classes (such as typographic design) which, it has been judged, would prove more costly than beneficial to propertize. Similarly, copyright does not give owners of copyright a right to control all uses of their works. Rather, they have control over certain kinds of enumerated uses--those uses whose control is desirably centralized in one entity. Thus, for example, the composer of a song has exclusive rights over reproduction and public performance, but not over private performance.
There are also system-wide rules that work to decrease deadweight loss, and to take better advantage of intangibles' inexhaustability. The key example here is that of duration. To see this, we must backtrack to consider a conceptual matter. At one point, lawyers foundered when asked how to assess, even conceptually, the value of an intellectual property system. The empirical questions are hard enough, but there appeared to be a paradox where deadweight loss was concerned. True, the exclusive right that the copyright or patent owner receives from the law confers a kind of monopoly power (of varying effectiveness, depending on the competing intellectual products available to the audience.) Also true, this monopoly power can then cause deadweight loss as the intellectual property owner imposes a price above marginal cost and the quantity effectively available to the public is reduced. But lawyers were hard put to assess the significance of this deadweight loss, because the item whose quantity was being reduced by the intellectual property law might never have come into existence without that very law.
The way out of this apparent paradox was to make a conceptual distinction between those works which needed the intellectual property law to induce their authors to create them, and those works which did not. As to the former items, any production is due to the legal regime's grant of intellectual property rights. As to the latter items, the intellectual property law merely functions as a restriction.
Thus, one would credit the intellectual property system with all the value of the works that would not have come into existence without intellectual property rights. As to these works, any production of the item would count as a plus, and deadweight loss would be irrelevant. Then, from this aggregate positive value would be deducted the deadweight loss in markets for works that would have been produced even in the absence of intellectual property rights.
The value of the intellectual property system is the net of these two numbers. As Professor Stanley Liebowitz has made clear, imposing system-wide durational limits-- limiting how long particular types of intellectual property rights will last-- can serve to maximize this net value.
In the ideal case the economist would want to separate all works into two classes, those which required intellectual property law as an incentive and those which did not. But even in the real world, some rough guesses are possible.
First, as duration increases, the number of new works attributable to the copyright system's incentive will grow smaller. The usual law of diminishing marginal utility would seem to govern; as the duration of a copyright or patent is made longer and longer, the incentive effect of additional length is likely to decrease. To illustrate: extending the duration of copyright from one year to five is likely to so increase the expected rewards of writing a book or designing a poster that the increase will induce some new works to be made which would not otherwise be created. By contrast, extending the duration from 101 years to 105 is not likely to have as strong an effect, so such further extensions will bring less to the "plus" side of the ledger.
Second, as duration increases, more and more works will not "need" the extra years to come into existence. As the reader will recall, years of protection that are not needed for incentives constitute unnecessary restrictions, and as to them "deadweight loss" should be counted on the debit side of the social ledger. For example, works which were called into existence by the promise of a 56-year reward are "pluses" to be credited to the copyright system for only their first 56 years. The copyright system's grant of exclusive rights for years 57 and following deserves no credit for those works’ creation. Equally importantly, any deadweight loss in markets in years 57 and following becomes a cost attributable to copyright.
Thus, as duration grows longer, the incentive value of an added durational restriction grows less, and the deadweight loss grows larger. The economic goal in choosing a durational limit is to maximize the difference between the two measures, incentive value and deadweight loss. Thus, duration can be custom-designed, providing a set of system-wide sharp rules that limit the periods of protection for intangibles protected by a given regime, such as the seventeen to twenty years that U.S. law gives to utility patents, as compared with the fourteen years of design patents, the "life plus fifty years" that inheres in most copyrights, or the ten years of protection applicable to semiconductor chip mask works. These various systemic limits can function to minimize the cost attributable to deadweight loss.
Sometimes markets do not evolve for a particular creative work or use-- say, for example, that bargaining is impeded by problems such as externalities, or high transaction costs in identifying or communicating with the copyright proprietor. If the copyright laws prohibited copying in that area it would simply be preventing copying without yielding creators any monetary advantage. That would be undesirable. Not only would copyright then fail to perform its primary function, but if users cannot reach market deals with creators, copyright would impose more costs and generate less benefit than would a regime without copyright. For though incentives may be low in a world without copyright, at least copyists and other users would have access to whatever works happened to be created; by contrast, in a world where there is copyright but no markets, incentives are low and the public has no access. Therefore, as discussed in the initial sections of this article, the ability of users to form markets is crucial to copyright's economic mission of encouraging the production and use of new work.
This observation has implications for policy in individual cases. If a defendant faces market failure in the face of copyright, then, in his case, the economic foundation for copyright has crumbled. That is a good argument (if not a complete one) for not enforcing the copyright against him. Thus, it can be argued that “fair use” has evolved as an equitable response to market failure, to ensure that socially desirable uses will not be blocked.
For example, consider photocopying by individual scholars. The transaction costs in contacting a copyright owner for permission to photocopy might well outweigh the benefit the scholar expects to reap. In such a case, enforcing the copyright would merely eliminate the photocopying, rather than generate any license fees for the copyright owner. In such an event, granting “fair use” treatment to the scholar will not impair the copyright owner's potential income stream, and will allow a socially beneficial use to go forward that the transaction costs barrier would otherwise have blocked. High transaction costs are, of course, a classic cause of market failure.
The market failure approach is consistent with the great bulk of “fair use” precedent, and in recent years this sort of argument has even found its way into the courts' explicit arguments. For example, in a recent fair use case involving corporate photocopying of scientific journals, the courts clearly had a market failure model in mind. The opinions of both the District Court and Court of Appeals discuss what economists identify as "transaction costs," and examine the extent to which the defendant's employees-- if required to stop photocopying-- could find other avenues through which to obtain the desired material. Similarly, in the most recent fair use case before the Supreme Court, the opinion indicated that “fair use” can be justified in part as a response to situations in which copyright owners are unlikely to give permission at virtually any price.
The latter position, advanced in a case involving a song parody, might strike the reader as inconsistent with the usual economic assumption that one must take preferences as a given. If one takes this notion seriously-- it is sometimes known as the assumption of "consumer sovereignty"-- then it seems the Court should have accorded to the copyright owner's desire not to be parodied as much respect as any other value. After all, in theory, an unwillingness to sell or license merely indicates that the potential buyer/licensee is not the highest-valued user. So it may seem wrongheaded of the Supreme Court to suggest that it may be appropriate to give a parodist -- a disappointed licensee -- the liberty to copy for free on the ground that the owner would not sell him a license. Is the Court under-valuing the owner's preferences? Not necessarily; there are several explanations of the Court's approach that are consistent with the traditional economic deference to individual preferences.
When a copyright owner refuses to let someone adapt her work for purposes of parodying it, or refuses to give an ideological opponent permission to quote lengthy passages, or insists on suing anyone who quotes passages of her memoirs that reflect unfavorably on her, she is using her copyright as a tool of suppression. The question of whether authors should be entitled to refuse permission to those users of whom they disapprove is a complex one. For example, there can be practical problems in distinguishing improperly-motivated suppression from a refusal to license motivated by a desire to maximize financial return. More important than the practical problems may be a conceptual one. If the proper way to look at these problems is economic, then, as mentioned, the principles of consumer sovereignty would seem to dictate that governmental decision-makers should not question why someone refuses to sell or license. Economics "assum[es] that man is a rational maximizer of his ends in life," and a desire to suppress would seem to be as rational an end as a desire for fame or fast cars.
Additionally, Ronald Coase has persuasively emphasized the importance of transaction costs by showing that, in their absence, the ultimate allocation of a resource will be efficient regardless of how entitlements are initially assigned. So long as the parties can meet face to face, as in copyright a copyright owner and potential parodist or critic could often do, why should there be any need for the judiciary to do anything but enforce whatever property right is before it?
Whether suppression would or would not be economically desirable will depend in most cases on empirical analysis of the particular fact pattern. But some general observations can indicate preliminarily why, when copyright owners seek to use the copyright law to enforce attempts at suppression, neither consumer sovereignty nor the Coase Theorem suggest that judges give the owners formal deference.
At least four reasons suggest that the market cannot always be relied upon to mediate attempts at suppression and that it might be economically desirable to refuse authors an entitlement to suppress. The four reasons are the "suppression triangle"; pecuniary effects; managerial discretion; and endowment effects. The four reasons are interrelated, and to explicate them let me begin with the "suppression triangle.”
In all these examples, remember: so long as there is the possibility that the social interest will be better served by refusing to enforce the owner's copyright, the case is made for using a nonformal, reasonableness mode of inquiry. One would then need to compare the costs of the extra suppression that would result from adhering to a formal pro-owner result, with the administrative and other costs that would be necessary in employing a "reasonableness" or "fairness" inquiry. In the United States, with its strong history of prizing free speech, the costs of improper suppression of news or cultural material is viewed as very high, tending to outweigh the administrative cost consideration
I use the term "suppression triangle" to point to the fact that in cases involving the suppression of information or other intellectual products, at least three parties are affected: (1) the person who seeks or threatens to make the contested use (for example, the potential parodist), (2) the copyright owner who wants to keep the material from being copied or adapted (the potential suppressor), and (3) the person or persons who would want to see the material (the potential recipients). This is the triangle of affected interests. Yet in the suppression transaction typically only two parties are present: the potential user (such as a parodist), and the copyright owner. Whether an attempt to suppress is likely to be value-maximizing will depend, inter alia, on how well the interest of the omitted third party, the class of potential recipients, is represented by the two immediate participants.
Theoretically, the more valuable the parody or other use is to the public, the more the public should be willing to pay for it, and the more the parodist should be willing and able to bid for permission. Thus, the notion of the Invisible Hand expects that any market participant will be in a position to reflect the interests of affected third parties (that is, the public audience). Nevertheless, the Invisible Hand often falters, and the possibility of misallocation remains.
Consider a hypothetical novelist or moviemaker who wants to keep the world from knowing what a hostile critic or parodist has to say about his work. Assume also that the critic or parodist wants to quote from the work or use its imagery and that use of the quotation or imagery is somehow essential to the comprehensibility or believability of the criticism or parody. If the law required the critic or parodist to purchase licenses to quote or paraphrase, how sure could we be that the "highest-valued" use would ensue?
For purposes of mathematical example, assume that the critic or parodist stands to earn at most a thousand dollars profit from even the best-written product. Assume that the novelist or film-maker would lose fifty thousand if the criticism or parody is published. Since the copyright owner would charge at least fifty thousand for a license to criticize or ridicule his work and the critic or parodist stands to gain only one thousand from publishing, it may look like the copyright owner holds the "highest valued" use when compared with the parodist or critic. But that may be an illusion resulting from the fact that the third party (in the owner/user/public triangle) is not being counted as part of the deal.
The publishing of the review or parody might benefit the public (who would thus be warned off from, let's say, a much-hyped romance novel that doesn't really excite anyone who reads past page five) to the tune of that same fifty thousand, or perhaps even more. On these hypothesized facts, requiring the publisher to buy a license from someone who would not sell it is a bad idea, and giving the publisher (the critic or parodist) free use is a good idea. And both are consistent with economic measures of value. If the critic had been able to capture the full value that the review gave to the audience, then the novelist's fifty thousand minimum asking price would have been met.
A parodist may similarly be unable to capture the full value that the work holds for the audience. This can occur for many reasons. There may be significant positive externalities and surplus in the market for parodies, for example. There also may be other complications in the markets for reviews and parodies, such as pecuniary losses that diverge from societal economic losses.
Much of the loss that can come from a critical review will often be merely pecuniary, reflecting not a net loss to society but rather a shifting of revenues from one novelist to another and possibly better one. It is as if the triangle now were a geometric figure with four points (the criticized novelist, the critic, the public, and the better novelist). If one could add to the price offered for the "license to criticize" an amount reflecting the monies that the better novelist would reap, it might be enough to make the difference. Since this cannot happen, mere pecuniary losses may take on an importance they should not have and they might prevent socially desirable licensing.
Another possible complication has to do not with the potential buyer's inability to raise the appropriate amount of capital, but with the potential licensor's potential inability to know even a good deal when it comes along. This complication I will label managerial discretion, by which I mean to embrace all those things that may make managers in complex corporations sometimes arrive at decisions that are less value-maximizing than they could be. I would include here, for example, personal risk aversion, bureaucratic structure, group dynamics, and laziness. Thus, the officials of a company that owns a given copyright may refuse to license simply because the license is in an unfamiliar field and their particular bureaucratic structure penalizes unlucky risk takers more than it rewards lucky ones. When critical, parodic, or otherwise controversial licenses would be at issue, the human desire to "play it safe" might prevent value-maximizing transfers from occurring. Managerial discretion is just one of many agency problems that can prevent the parties from dealing with each other like the unitary participants in the classic Coasian transaction.
All of the above are reasons why socially desirable "licenses to be critical" are not likely to be granted if left solely to the devices of copyright owners. One additional and probably most important factor remains to be discussed: the difference between willingness to pay and willingness to sell, sometimes identified with “endowment” or “wealth” effects.
The concept here basically refers to the fact that giving someone an entitlement makes that person richer, and this may change how the holder values both the entitlement and other resources, and this in turn may affect how entitlements are eventually allocated once bargaining between that person and other persons is completed. Wealth effects do not retard resources from moving to hands in which, given a particular entitlement starting-point, has the highest-valued use for them. Nor are they often strong enough to make a difference; in instances where fungible commodities are sold in markets populated by many buyers and sellers, "buy" prices and "sell" prices probably tend to converge. But, when wealth effects do have an impact, they have the potential of rendering the meaning of "highest-valued" use indeterminate in the sense that the location of the highest-valued use is not independent of the law. Where wealth effects are strong, everything depends on the legal assignment of entitlements that form the transaction's starting point. As a result, in such cases the search for the highest-valued use cannot provide a good basis for assigning initial entitlements.
Professor Coase showed that in a world without transaction costs, resources will be traded to their highest-valued uses, so that, as between any two users of a resource, if A can use the resource more productively than B, A will end up with it. Therefore, many scholars argue, in a real world full of transaction costs that can impede bargaining, it often makes sense to "mimic the market" and assign legal rights to the highest-valued user in the first instance. This is a core insight of Law and Economics.
Yet the Law and Economics argument largely depends on there being a stable highest-valued user. The injunction to "seek efficiency by mimicking the perfect market" only makes normative sense if the perfect market allocation is a constant. If the allocation of rights significantly affects the monetary valuation that parties place on a resource, then there may be no stable economic reality for the law to seek to mimic.
There is indeed a rare class of goods which lack this stability. These are the precious, personal, irreplaceable, crucial goods one thinks of as "priceless.” Examples are many: the Dead Sea Scrolls; family heirlooms; one's children; one's health; one's reputation; one's peace of mind. The monetary value a person places on one of these goods may well depend on whether the person has a legal entitlement to it (whether she "owns" it) or whether she must purchase it.
Consider health, for example. It is plausible that most people would be unwilling to sell their organs at any price, so that Jane Smith might turn down an offer of five million dollars from Billionaire X for one of her kidneys. Similarly, if Jane Smith has kidney failure and one of her dying relatives wills her a healthy kidney, she might well be unwilling to take the billionaire's five million dollars in exchange for her entitlement to it. If so, Jane Smith looks like the kidney's "highest-valued user." But should she have no entitlement to the kidney from the recently-deceased person (perhaps because the relevant jurisdiction does not recognize such bequests as enforceable), Jane Smith's own budget and health insurance will place a limit on how much she can spend pursuing the transplant. It is highly unlikely she will be able to outbid Billionaire X for the kidney. If so, Billionaire X will appear to be the "highest-valued user." One can draw from such a pattern no reliable information about whether the resource has its highest value in the hands of the billionaire or Jane Smith. This phenomenon might be called the "pricelessness effect.”
The pricelessness effect is a subset of the category that economists call “endowment effects” which in turn is related to “wealth effects”: since assigning an entitlement to someone makes that person wealthier, it can affect the valuation the person puts on resources. For example, often "ask" and "offer" prices differ from each other. Many people hedge the Coase Theorem by noting it does not apply when significant wealth or endowment effects are present. But usually the wealth or endowment effect is so minor that it does not impair the reliability of using a market mimicry approach to model efficiency.
The "pricelessness effect" deserves having its own name precisely because the subcategory of effects it denotes are likely to be significant. The "pricelessness effect" comes into play when the entitlement at issue pertains to a good that (1) an individual or group values very highly and (2) which is virtually irreplaceable, and (3) when it is the allocation of that very good which is at issue. As to such items, the initial placement of the entitlement is likely to have a sharp effect on the price and allocation of the resource, even in the absence of transaction costs.
In cases of parody or criticism-- both areas where “fair use” treatment tends to be awarded to defendants-- reputation may be at issue. To many, reputation is priceless in the sense we have been discussing. For example, a novelist who fears that a journalist will use extensive quotations from her book to bolster a hostile review will be most unlikely to sell the journalist a license to copy those quotations-- regardless of the price offered. But that does not mean the author's preference is the "highest-valued use" in any meaningful sense, since that same author may be unable to buy silence if the law gives the journalist a "fair use" liberty right to publish. A similar analysis can be made of parody: since most people intensely dislike being ridiculed, the legal right may determine where the highest-valued use lies. In such cases, the market is useless as a guide, and formal deference to owners' market powers is inappropriate.
For example, assume A is a novelist, a copyright owner who has an entitlement not to license and who is otherwise financially comfortable; she has perhaps $4000 in the bank and a two-year old car and a prospect of steady royalties. She may be tempted by B's offer of, say, $10,000 for a license to use her work, but she can afford to say no without altering her lifestyle. If B's project is an ordinary commercial project and A will not be sacrificing more than $10,000 from foregoing alternative uses of the work, she will probably license. (It might also happen that B's project would not require an exclusive license and would not otherwise interfere with A's other licensing opportunities. If so, granting B permission to go forward would have no opportunity cost at all for A. She would be even more likely to license such a use.) However, if B's project is hostile toward A's work as a whole, A may well refuse the license, either to protect her long-term economic interest (which may be a mere pecuniary loss, remember), her aesthetic reputation, or her feelings.
If however the law gave novelist A no entitlement to prevent B's use, then she would have to persuade B not to publish (cf., blackmail.) The most she could offer B to persuade B not to make the critical use planned is the amount in her bank account, plus whatever she could sell her car for, plus whatever she could borrow on the strength of her expected royalty stream. The total may well be less than $10,000, and A will probably demand a price in excess of $10,000. Give A the entitlement and the highest-valued use of the contested expression is in her hands; give B the entitlement and the highest-valued use is in that licensee's hands. The locus of the "highest-valued use" has shifted as a result of where the law places its entitlement.
In such cases, looking to the results of consensual transactions will not give us any information about who "should" have the right.
Another way to put the point is this: Economics is sometimes used as a normative guide for good social policy. When it is used in this fashion, its primary claim to legitimacy stems from the links between economics and utilitarianism. The more that income distribution restricts the expression of individuals' preferences, the more shaky the link between economics and utility becomes. This linkage has the potential for completely breaking down in cases of "pricelessness.” Though in such cases the parties' preferences may remain constant, both in their objects and in their intensity, a shift in who owns the entitlement may effectively disable one of those parties from effectuating that preference. Thus a legal regime that is committed (even in part) to utilitarian consequentialism would be unwise to rely upon a money-bound market model for normative guidance in cases of pricelessness.
In sum, refusing to allow a copyright owner to suppress a hostile use of the copyrighted work, in a case where the "pricelessness effect" is likely to make a determinative difference, does not necessarily contravene economic principles. In such an instance, it is appropriate for even an economically-oriented court to refuse to defer to the copyright owner, and instead make an independent weighing of how enforcing the copyright in the given instance would affect welfare, and any other relevant consequentialist or nonconsequentialist policies.
This article has employed the notion of asymmetric market conditions to tie together many aspects of intellectual property law. The primary focus has been on explaining why American copyright law uses a nonformal, equitable, substantive reasonableness approach like "fair use" in contexts where other forms of American intellectual property law use approaches that are formal, rule-bound, and system-wide. The article also examines the dynamics behind the key systemic choice made by copyright law, which is to reverse the presumption of restitution law that people who refrain from crossing others' tangible boundaries are free to take advantage of each other's labor. In intellectual property law, this presumptive freedom is replaced by a duty not to copy.
Economics does not capture everything of importance. Nevertheless, a model of asymmetric market conditions can provide a useful method for unifying a number of disparate questions in the economics of intellectual property law.
[*] Copyright © 1997 by Wendy J. Gordon. Professor Gordon is Professor of Law and Paul J. Liacos Scholar in Law, Boston University School of Law, 765 Commonwealth Ave, Boston, MA 02215 USA. She can be reached by internet at email@example.com.
An earlier version of this paper was published in Germany as Systemische und fallbezogene Lösungsansätze für Marktversagen bei Immaterialgütern (translated into German by Elisabeth Haberfellner) in Ökonomische Analyse der rechtlichen Organisation von Innovationen, Beiträge zum IV. Travemünder Symposium zur ökonomischen Analyse des Rechts, J.C.B. Mohr (Paul Siebeck) Tübingen, 328-67 (Claus Ott & Hans-Bernd Schafter eds., 1994).
The author wishes to note that along with new material, this article contains certain focal points and short extracts from some of the articles she has previously published. She suggests you refer to those other articles for fuller thematic and doctrinal development.
The author is grateful for suggestions to Ron Cass, Alan Feld, Elly Leary, Richard Markovitz, Hans-Bernd Schäfer, and participants at the Travemunde Fourth Biannual Law & Economics Conference, the Boston University faculty workshop, and the Olin Law & Economics Workshop at the University of Toronto. She also appreciates the research assistance provided by Sophy Chen and Douglas De May.
Above all she is grateful to the late Sam Postbrief.
 The first condition is that authors and inventors would not be able to obtain much payment for their work in the absence of a rule that restrained strangers from copying, and, as a result, potential creators produce fewer works than the public would have been willing to pay for if payment had been required. This shortfall evidences the market failure.
The paradigmatic form of market failure is a prisoner's dilemma. Roughly speaking: where the payoff from copying is very high and the penalty for engaging in independent creation when the other party copies is also very high, little independent creation is likely to occur even though significant positive social and individual payoffs might have resulted if both parties had chosen the route of independent creation. See Wendy J. Gordon, Asymmetric Market Failure and Prisoner's Dilemma in Intellectual Property, 17 U. DAYTON L. REV. 853 (1992) [hereinafter Asymmetric Market Failure].
 Market success occurs when, inter alia, parties who wish to trade can deal with each other at fairly low levels of transaction costs. See id.
 This is explored more deeply in Wendy J. Gordon, On Owning Information: Intellectual Property and the Restitutionary Impulse, 78 VIRGINIA L. REV. 149, 230-38 (1992) [hereinafter On Owning Information]; Gordon, Asymmetric Market Failure, supra note 1; and Wendy J. Gordon, Of Harms and Benefits: Torts, Restitution and Intellectual Property, 21 J. OF LEGAL STUD. 449 (1992) [hereinafter cited as Of Harms and Benefits].
 See Wendy J. Gordon, Fair Use as Market Failure: A Structural and Economic Analysis of the Betamax Case and its Predecessors, 82 COLUMBIA L. REV. 1600, 1611 (1982) [hereinafter cited as Gordon, Fair Use] and sources cited therein.
 In practice, these properties are not fully present. Thus, the value of a song can become zero through saturation-level repetition, and some exhaustible physical products-- such as a phonograph record or a radio set-- may be necessary to afford access to the intangible. Practically speaking, then, songs may not be infinitely available to all as a valuable good. The same partiality characterizes an intangible's nonexcludability. Thus, for example, though songs are easy to copy once heard, the initial score may be easy for the composer to keep private. Or the composer may be able to extract no-copy promises from his early, small audiences. Even after a song is popular it may be desired in a format that favors an authorized producer's distribution networks.
 The extent to which public provision of public goods is indeed necessary in various contexts is, of course, a matter of debate. For example, Ronald Coase has shown that although lighthouses are a classic public good (their light can be used by a virtually unlimited number of ships within range, and no ship can be practicably excluded from their light), some lighthouses have in fact been built through non-governmental arrangements. See R. H. Coase, The Lighthouse in Economics, 17 J. LAW & ECON. 357 (1974).
 Or, more precisely, there is a need for taste to evolve outside the state apparatus in order for individuals to maintain some degree of genuine self-determination. See, e.g., C. Edwin Baker, Property and its Relation to Constitutionally Protected Liberty, 134 U. PA. LAW REV. 741 (1986).
 Scholars who are critical of copyright, or who doubt the wisdom of its expansion, typically argue that copyright may not be necessary for creators to obtain payment for their work. See, e.g., Stephen Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 HARV. L. REV. 281, 350 (1970); Tom Palmer, Intellectual Property: A Non-Posnerian Law and Economics Approach, 12 HAMLINE L. REV. 261 (1989).
 The U.S. Constitution empowers Congress to enact copyright and patent "for limited times" to further "the progress of Science and the useful arts." U.S. Const. art. I, § 8, cl. 8. Congress enacted its first copyright statute in 1790.
 Also, American patent law prohibits even duplication of the patented invention that happens to result from completely independent efforts.
 The incentives for the creation of new work provided by an intellectual property system must be weighed against the deadweight loss and administrative costs of the system; the economic goal is to obtain the highest net sum. See William M. Landes & Richard A. Posner, An Economic Analysis of Copyright Law, 18 J. LEGAL STUD. 325, 326 (1989). See also Stanley J. Liebowitz, Copyright Law, Photocopying, and Price Discrimination, 8 RES. L. & ECON. 181 (1986) (edited by R. Zerbe).
What matters is not the absolute level of administrative costs involved in an intellectual property system, but rather a comparison among the costs of the various potential systems. Even a system of no intellectual property rights will have significant administrative costs. For example, assume there is a legal regime that rejects patent and copyright and recognizes only individually-negotiated contracts as a limitation on the public's ability to copy. In such a context, for example, inventors may spend a good deal of money on policing the secrecy of their inventions, composers may spend a good deal of money on obtaining contractual promises-not-to-copy from people who seek entry to concerts, and the like.
Similarly, it should not be imagined that intellectual property rights are the only way that costly decreases in public access occur. A regime without intellectual property rights will afford far from unlimited access to the public. To use the prior examples: the inventor unprotected by patent may be unwilling to trade information with rival firms, and composers unprotected by copyright may be unwilling to allow radios or television to broadcast their music to general audiences.
In short, it should not be imagined that underproduction is the only cost of the system without intellectual property rights. For further analysis of the many possible alternatives to copyright law and their costs, see Wendy J. Gordon, An Inquiry into the Merits of Copyright: The Challenges of Consistency, Consent, and Encouragement Theory, 41 STAN. L. REV. 1343 (1989) [hereinafter Merits of Copyright].
 The marginal cost of producing an extra unit of an intangible may be zero, because of inexhaustability, or it may be some positive sum corresponding to the cost of the intangible's physical embodiment (such as the cost of the plastic that goes into a phonograph record).
 It might be argued that in cases of true Pareto-superiority, the law would not need to provide a safe-harbor for free copying: if the owners of the patents and copyrights were truly unharmed by the copying (as the notion of pareto-superiority assumes), they would allow the copying to proceed without hindrance.
However, humans are both envious and insecure. The copyright owner might refuse permission not because he is suffering tangible harm, but because he is irritated that other people are getting a free ride or because he has irrational fears about future harm. A society may well choose to consider itself entitled to disregard envy and insecurity as legally relevant harms.
 Cf. Clifford Holderness, A Legal Foundation for Exchange, 14 J. LEGAL STUD. 321 (1985).
 The other two types of rights that can be sued upon in American civil (non-criminal) courts are those arising out of contract law and restitution.
"Restitution" is concerned with benefits rather than harms. See Gordon, Of Harms and Benefits, supra note 3. See also Saul Levmore, Explaining Restitution, 71 VA. L. REV. 65 (1985). Like negligence law, restitution involves a case-by-case mode of adjudication.
In restitution a person brings suit on the ground that the defendant has been unjustly enriched and that this enrichment came either at the plaintiff's expense or by violating some right of his. There is an obvious need for case-by-case adjudication in order to decide what enrichments are "unjust." Restitution also functions as a remedy following on the violation of other rights (e.g., restitution may require a trespasser to return the profit he made by trespassing on a plaintiff's land). Restitution is discussed further in this article at pages 24-29, below.
 I am indebted here to the work of William Powers. Comparing trespass with negligence, for example, Professor Powers notes:
Ownership embodies a formal methodology, since . . . questions concerning appropriate use are answered wholly by asking whether a proposed use has been sanctioned by the owner. A decision by the landowner . . . concludes legal debate under the ownership model. On the other hand, a duty of reasonable use embodies a nonformal methodology because it makes direct, ad hoc reference to efficiency [or other measures of social desirability]. Under this model, a decision concerning the landowner . . . would depend on a comparison of relative costs and benefits in the specific case.
William C. Powers, Jr., A Methodological Perspective on the Duty to Act (review essay), 57 TEXAS L. REV. 523, 526-27 (1979).
 Also consider the distinction between “rules” and “principles” in Ronald Dworkin, TAKING RIGHTS SERIOUSLY (1977).
 Admittedly, even in intentional torts American courts may take cognizance of excuses (such as incapacity) and justifications (such as necessity or self-defense). This does not undermine the distinction between intentional and unintentional torts, however. Not only is the burden of proving such intentional-tort defenses typically on the defendant, but these defenses also permit a court far less latitude than does the broad balancing of costs and benefits which a court engages in under a reasonableness inquiry. Thus, a person taking a shortcut through another’s land can take advantage of the “necessity” defense only if an imminent danger made the shortcut imperative.
 See Powers, supra note 16.
 There are many verbal formulations as to the meaning of "substantial similarity," also known as "illicit copying," but none does very much to define the jury's task. Consider, for example, a classic case regarding music infringement, Arnstein v. Porter, 154 F.2d 464 (2d Cir. 1946), cert. denied 330 U.S. 851 (1947). First, the court valuably noted that, "Assuming that adequate proof is made of copying, that is not enough; for there can be 'permissible copying,' copying which is not illicit." Id. at 472. But then it foundered when it had to distinguish rightful from illicit copying:
The proper criterion on that issue . . . is whether defendant took from plaintiff's works so much of what is pleasing to the ears of lay listeners, who comprise the audience for whom such popular music is composed, that the defendant wrongfully appropriated something which belongs to the plaintiff.
Id. at 473.
Aside from the reference to the lay (non-expert) audience, this formulation offers no more than vague references to quantity ("so much"), to market value, and to a conclusory notion of wrongfulness.
 If there is a great deal of quantitative resemblance, it is highly likely that substantial similarity will be found. The range of discretion is usually more limited on questions of substantial similarity than on questions of “fair use.” The latter doctrine is discussed immediately below.
 See 17 U.S.C. § 107. As originally enacted, section 107 provided as follows:
107. Limitations on exclusive rights: Fair use
Notwithstanding the provisions of section 106 [which set out a copyright owner's exclusive rights of reproduction, adaptation, public performance, and the like], the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include -
(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
(2) the nature of the copyrighted work;
(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
(4) the effect of the use upon the potential market for or value of the copyrighted work.
17 U.S.C. § 107 (1976). The section was recently amended to make clear that the unpublished status of a work should not be determinative.
For my further examinations of the doctrine, see, e.g., Gordon, Fair Use, supra note 4, (using an economic model to unify fair use cases and give some precision to the notoriously open-ended doctrine). See also Wendy J. Gordon, A Property Right in Self-Expression: Equality and Individualism in the Natural Law of Intellectual Property, 102 YALE L.J. 1533 (1993) [hereinafter Property Right in Self-Expression]; Wendy J. Gordon, Reality as Artifact: From Feist to Fair Use, 55 LAW & CONTEMP. PROBLEMS 93 (1992).
 The legislative history of section 107 indicates that, despite the statutory recognition accorded fair use, the nature of the doctrine remains to be defined by case law: "The bill endorses the purpose and general scope of the judicial doctrine of fair use, but there is no disposition to freeze the statute. . . . " H.R. REP. NO. 94-1476, at 66 (1976), reprinted in 1976 U.S.C.C.A.N. at 5680 [hereinafter House Report]. See also S. REP. NO. 94-473, 1st Sess. 62 (1975) [Senate Report]. The courts have recognized their freedom to continue the development of fair use doctrine. See, e.g., Triangle Publications, Inc. v. Knight-Ridder Newspapers, Inc.,  USCA5 1529; 626 F.2d 1171, 1174 (5th Cir. 1980) ("Congress made clear that it in no way intended to depart from Court-created principles or to short-circuit further judicial development. . . . ").
 House Report at 65, reprinted in 1976 U.S.C.C.A.N. 5659, 5679.
 See Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and Inalienability: One View of the Cathedral, 85 HARV. L. REV. 1089 (1975).
 See Cambell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994). The case was remanded for further proceedings in light of an opinion that, inter alia, made clear the copyright owner's refusing to license might not work against a fair use finding. See id. (parodists are unlikely to be able to obtain consents). See also id. at 585 n.18 ("we reject Acuff Rose's argument that 2 Live Crew's request for permission to use the original should be weighed against a fair use finding.").
 See Harold Demsetz, The Private Production of Public Goods, 13 J.L. & ECON. 293 (1970).
 See Benjamin Kaplan, AN UNHURRIED VIEW OF COPYRIGHT (1967)
 If copying is harmless, allowing copying produces a Pareto-superior result: no one is hurt and the copyist and her customers gain.
 See Gordon, On Owning Information, supra note 3; Gordon, Property Right in Self-Expression, supra note 22.
 In Aristotle, equity is a means of correcting for the law's inevitable overinclusiveness. Aristotle writes,
When the law speaks universally, then, and a case arises on it which is not covered by the universal statement, then it is right, where the legislator fails us and has erred by over-simplicity, to correct the omission . . . .
ARISTOTLE, 10 ETHICS, IN THE NICOMACHEAN ETHICS 133 (D. Ross trans, revised by J.L. Ackrill & J.O. Urmson, Oxford, 1984).
 The legislative history of the U.S. copyright act, including the struggles among various industry groups, is recounted in the following two articles by Jessica Litman, Copyright Legislation and Technological Change, 68 OREGON L. REV 275 (1989), and Copyright, Compromise, and Legislative History, 72 CORNELL L. REV 857 (1987).
 Powers, supra note 16, at 527.
 Usually equity is used to soften the law's effect, and is sometimes seen as a form of mercy. Yet case-by-case approaches can also be used to extend the law's reach, to inhibit socially undesirable behavior. This use of the equity approach is much more controversial, for it may involve use of the state's coercive power without fair warning as to what is prohibited. The danger of chilling the exercise of justifiable liberties thus makes the use of equity as a “sword” for plaintiffs more questionable than the use of equity as a “shield” for defendants.
Nevertheless, in the United States law of intellectual property, equitable approaches have been used not only to assist defendants (as in "fair use"), but have also been used to create new causes of action for plaintiffs. Thus, the U.S. has a general privilege (liberty) of free competition, so that competitors are free to imitate each other so long as they do not violate copyrights, patents, trade secrets and trademarks. Yet courts sometimes invoke their equitable judicial power to forbid counter-productive competition, usually under the tort labelled "misappropriation," in cases where the liberty to copy results in a strong enough market failure.
The classic and still controversial misappropriation case was decided by the United States Supreme Court during World War One, International News Service v. Associated Press,  USSC 191; 248 U.S. 215 (1918). In that case the International News Service (INS) had been barred from the front by certain European censors, and INS began rewriting the war news gathered by its competitor Associated Press (AP) and providing that news to INS's customer newspapers. The resulting pattern of incentives much resembled a prisoner's dilemma game: productive behavior (productive news-gathering such as that engaged in by AP) was becoming too costly given the likelihood of defection (copying) by the other party. Yet there was nothing in the law of copyrights, patents, trade secrets or trademarks that prohibited the rewriting of news gathered by another. The court nevertheless enjoined INS from so using its competitor's news.
For further discussion of this case and the doctrine of misappropriation it spawned, see Gordon, On Owning Information, supra note 3; Gordon, Asymmetric Market Failure, supra note 1; Douglas G. Baird, Common Law Intellectual Property and the Legacy of International News Service v. Associated Press, 50 U. CHI. L. REV. 411 (1983) (discussing INS's impact on the development of intellectual property rights).
 See Jason Johnston, Bargaining Under Rules Versus Standards, 11 J. L. ECON. & ORG. 256 (1995).
 For example, in this century the justices of the United States Supreme Court have decided no more than a half-dozen fair use cases. Yet in two of those cases the justices split four-to-four. One case involved a movie parody that Jack Benny did on his television show. Benny v. Loew's, Inc.  USCA9 470; 239 F.2d 532 (9th Cir. 1956), aff'd by an equally divided Court, 356 U.S. 43 (1958) (fair use found not to shelter defendant's parody; too much had been taken). The other involved a government library that engaged in massive photocopying of medical journals for the use of doctors. Williams & Wilkins v. National Institute of Health, 487 F.2d 1345 (Ct. Cl. 1973), aff'd by an equally divided Court, 420 U.S. 376 (1975) (fair use found to shelter defendant's photocopying).
In cases where the Supreme Court is evenly divided, the result of the lower court decision becomes binding.
 See Gordon, Asymmetric Market Failure, supra note 1 (from which some of this discussion is drawn). See also Gordon, On Owning Information, supra note 3, at 222-23, 230-38 (presenting the concept of asymmetric market conditions - but there called “asymmetric market failure” - and describing its economic bases). Asymmetric market conditions, as a prerequisite for property rights, also has roots in corrective justice. See id. at 180-221. However, investigating that aspect would take us too far afield.
 I call the confluence of the two conditions "asymmetric" because where they obtain, markets are more likely to exist under one legal rule (copying prohibited) than under its opposite (copying permitted).
In a world where lack of legal restraint on copying leads to market failure, authors cannot easily get paid. Yet if in a world that has copying restrictions copyists can form markets, they may not be stymied. Rather, licensing may evolve.
 American copyright law previously required a notice, c in a circle followed by name and date, on all published written copies; more recently, the use of notices has been made largely optional. Nevertheless, the notices provide a useful service in assisting readers to know if the material they are using is in current copyright, and, if so, whom they should contact for licenses.
 See Holderness, supra note 14.
 Thus, for example, the compulsory license which allows musical groups to make "covers" of records was adopted in response to a perceived monopoly problem: at the time phonorecords were to receive copyright protection, virtually all recording copyrights in songs were held by one company, Aeolian Co. See Robert Stephen Lee, An Economic Analysis of Compulsory Licensing in Copyright Law, 5 WESTERN NEW ENG. L. REV. 203 (1982).
 17 U.S.C. § 106 (granting "exclusive" rights "to do and to authorize" reproduction, public performance, public display, adaptation, and the like, for various subject matters.”)
 RESTATEMENT OF RESTITUTION § 2 (1937). See also 17 U.S.C. §§ 106, 112. It is sometimes said that when recovery is denied, plaintiffs tend to be called "intermeddlers," but when they win, they are more likely to be called "volunteers." Both words refer, however, to the same basic pattern: conferring benefits on someone who has not asked for them. This article uses the terms interchangeably.
This material is taken primarily from Gordon, Of Harms and Benefits, supra note 3; see also On Owning Information, supra note 3.
 Under American copyright law, the work's creator has a right to exclusively control the rights of reproduction, copying, adaptation, performance and the like, see 17 U.S.C. § 107; and therefore, can extract monies and obtain injunctions when someone does these things without permission.
 RESTATEMENT OF RESTITUTION, supra note 43, §§ 6-69.
 Id. §§ 70-106.
 Id. §§ 107-111.
 Id. § 112.
 See, e.g., id. § 40, cmt. b, at 109.
 Id., supra note 43, § 106. There are situations in which protecting one's own interests does not bar restitution, but these tend to be associated with coercion, as where a property owner discharges another's duty when that is the only way to prevent a third party from lawfully taking the property. Id. § 103.
 Id. § 106, illus. 2.
 Sometimes the absence of harm may make it easier to obtain fair use treatment, however. See the fourth factor in section 107, set out at footnote 22.
 Even when there is a market failure in the restitution context, so that the potential benefactor and the potential recipient are unable to identify or negotiate with each other, there are only very few circumstances in which payment is ordered through the courts -- mistake, request, coercion and a narrow range of emergencies justify recovery. This narrowness of recovery in the restitution context reflects the fact that if we give the benefit-generator a right to legally enforce recompense, it would tend to erode markets. See Levmore, supra note 15.
 Many examples exist of “inernalizing benefits” by contract. Thus, in many shopping malls, where small stores are likely to benefit from the propinquity of large department stores that draw masses of customers, the small stores may be willing to pay extra rent to subsidize the larger stores' entry. Something like this also happens in oil exploration: neighboring lessees will learn a great deal about whether or not it is worthwhile to drill under their own land from the results of their neighbor's drilling. So "dry hole contribution agreements" have come into being: contracts by which the neighbor who stands to benefit from the information agrees to pay a share of his neighbor's drilling costs should the hole come up dry.
Similarly, assume that landowners are likely to benefit from a venture like a resort complex locating nearby, but the resort is so expensive to build that it cannot afford many externalized benefits-- e.g., it will come to the area only if it is subsidized by the existing landowners or can itself capture most of the benefits generated. Even in such a case there may be no need to allow the resort complex to sue the benefitted owners after the fact in restitution. The developers can try to persuade these neighbors, in advance, to pay them something to encourage them to build nearby. Admittedly, there could be hold-out problems and other strategic maneuvering making this difficult. So the active party has another option: the owner of an attraction could quietly buy the land on which the beneficial spillovers will fall. This is apparently what the Disney organization did with Epcot and Disney World: it bought up surrounding land and built on it enough hotels and restaurants to capture much of the benefit their tourist attraction generates. Where this is possible, benefits are again internalized without the need for restitution suits.
 This is explored at greater length in Gordon, Of Harms and Benefits, supra note 3, and in Gordon, On Owning Information, supra note 3.
 In a world without intellectual property rights, an author may want to bargain with her audience for payment, but the audience is likely to be a wide and uncertain one, and the benefits are those that will flow from an as yet undisclosed intellectual product. Even if the author could somehow identify all the potential recipients, it would be expensive and awkward to reach simultaneously all of the persons who will eventually want access to the work. Even if this were possible, what would happen when the creator tried to negotiate for a payment from them all in exchange for disclosing the work? Many of those audience members might be tempted to hold back in the hope that others' monies would be sufficient to draw the work into the marketplace where they could then make a cheap copy. The larger the group of potential purchasers, the better the odds on the gamble may seem. Also, the work's contents are largely unknown at this stage; the less certain the benefits, the less seems to be risked if the gamble does not pay off. Good odds in favor of winning, and low perceived cost in the event of a loss, make the gamble very tempting. If enough people take this gamble in the hope of taking a free ride, the requisite funds may not be forthcoming. "Chicken," "prisoner's dilemma," and other free rider games illustrate analogous dynamics.
The presence of a publisher does not much alter the desirability of granting intellectual property rights to resolve potential bargaining stalemates. Admittedly, in a world without intellectual property rights the author may find it easier to deal with a publisher than with an undifferentiated audience (only one party; low transaction costs), but then the publisher must deal with the audience. The author's problems with information, transaction costs, and free riders would simply be passed on, one step further down the line. How much would a publisher pay for a book that could be lawfully copied by all comers once it appeared on the market? Unless the publisher has a lead-time advantage or some other sort of real-world clout that can discourage copying, the rate the publisher would offer the author in such a world might be too low. If the anticipated rate of payment is low, otherwise-desirable works may not be created.
 This is a variation of the phenomenon Dean Calabresi referred to as looking for the "best briber.” Guido Calabresi, THE COSTS OF ACCIDENTS (1970).
 The Coase theorem works only where resources can practicably be transferred to their highest-valued uses see R. H. Coase, The Problem of Social Cost, 3 J. L. & ECON. 1 (1960).
 See, e.g., 17 U.S.C. §§ 102, 103.
 This is a matter of legislative history; the statute itself does not explicitly mention typography. See House Report supra note 23, at 53-57. The reason protection was not given to typographic design may have been a fear that such protection might be misused by reprinters as a way to fence off public domain literature such as Shakespeare or the Bible.
 Industry pressures undoubtedly play a role here as well.
 For more on the property formation role of copyright's particular provisions, see Gordon, Merits of Copyright, supra note 11, and Gordon, Fair Use, supra note 4.
 In the discussion of duration that follows I am indebted to the work of Stanley Liebowitz. See, e.g., Liebowitz, supra note 11, at 183-188.
 That is, the person who owns copyright in a particular book will have a monopoly over that book, but not over competing titles by other authors.
 The discussion here puts aside what Edmund Kitch calls "prospect effects," namely, those positive effects on ease of exploitation that can occur when property rights are centralized in one entity. See Edmund Kitch, The Nature and Function of the Patent System, 20 J. LAW & ECON. 265 (1977).
 Liebowitz, supra note 11.
 His graphical representation is particularly helpful. See id. at 187.
 This point is probably made most wittily by Lord Macaulay, in his speeches before the British parliament protesting their extending the duration of copyright. Lord Macaulay argued that while copyright might be necessary to ensure a "supply of good books," the monopoly that it imposed was at best a necessary evil. "For the sake of the good we must submit to the evil; but the evil ought not to last a day longer than is necessary for the purpose of securing the good." THOMAS MACAULAY, Speech Before the House of Commons (Feb. 5, 1841), in 8 THE WORKS OF LORD MACAULAY 195, 199 (Lady Trevelyan ed. 1866) (discussing a bill which would have extended the duration of copyright protection).
 35 U.S.C. §§ 154-157.
 35 U.S.C. § 173.
 17 U.S.C. §§ 302-305. There is a different duration for works made for hire, etc.
 17 U.S.C. § 904.
 I have advanced this argument in Gordon, Fair Use, supra note 4, 1614-15, 1627-41. For further development, see id. at 1614-57 (proposing a 3-part test for fair use, and comparing such test with the case law results) and Wendy J. Gordon, Toward a Jurisprudence of Benefits: The Norms of Copyright and the Problem of Private Censorship (reivew essay), 57 U. CHI. L. Rev. 1009, 1042-43 (1990) [hereinafter Private Censorship]. See also, e.g., Landes & Posner, supra note 11; Sheldon Light, Parody, Burlesque and the Economic Rationale for Copyright, 11 CONN. L. REV. 615 (1979).
 That is, even if the scholar were willing and able to pay whatever price the copyright owner demanded, the scholar might not be willing to both pay that price plus bear the time delay, hassle, and secretarial costs involved in securing a permission.
 Note that this analysis is dependent upon the relative size of the applicable transaction cost barrier. If a clearinghouse or compulsory license system exists which reduces the transaction costs, then the scholar may not require fair use treatment in order to allow her use to go forward.
 See Gordon, Fair Use, supra note 4, at 1627-36.
 See American Geophysical Union v. Texaco, Inc., 802 F. Supp. 1 (S.D.N.Y. 1992) aff’d, 60 F. 2d 93, 929-31 (2d Cir. 1994).
 See Acuff-Rose, 510 U.S. 569 (1994). In assessing the plaintiff's claim that the parody would impair their potential market, the Court responded: "[T]he unlikelihood that creators of imaginative works will license critical reviews or lampoons of their own productions removes such uses from the very notion of a potential licensing market.” 510 U.S. at 592.
The Ninth Circuit Court of Appeals made a similar point in Fisher v. Dees,  USCA9 1324; 794 F.2d 432 (9th Cir. 1986) ("The parody defense to copyright infringement exists precisely to make possible a use that generally cannot be bought."). For other cases involving similar anti-dissemination motives on the part of copyright proprietors, see Gordon, Fair Use, supra note 4, at 1632-33.
 Similar instances also appear in the corporate realm. For example, when a newspaper expanded its TV coverage it told its readership about the extended service in an advertisement that pictured a copyrighted TV Guide cover for purposes of comparison. TV Guide then sued for copyright infringement. Presumably the suit was motivated by something other than a desire for license fees. The comparative advertising was held to be a fair use. See Triangle Publications, Inc v. Knight-Ridder Newspapers, Inc., 626 F. Supp. 1171 (5th Cir. 1980).
 For example, it can be difficult to distinguish suppression from an attempt to direct the work into the most valuable derivative work markets. See, e.g., Paul Goldstein, COPYRIGHT Vol I at 571-73 (rights over derivative works can affect the direction of investment and the type of works produced).
Similarly, in regard to unpublished works, it can be difficult to distinguish cases of suppression from cases of economically motivated refusals to license. An author accused of suppression may be simply trying to keep the work out of the public eye temporarily until it reaches its mature form and can be published.
Even if some practical means existed to distinguish all dissembling "suppressors" from those copyright owners who are genuinely motivated by financial return, some cases will present instances of truly mixed motives. For example, the owner of copyright in an out-of-print collection of letters might sue a biographer who extensively quotes the letters, not only out of a dislike for the biographer's message or perceived inaccuracies, but also out of a desire to preserve the reprint market for the letters. See Meeropol v. Nizer, 417 F. Supp. 1201, 1208 (2d Cir. 1976), rev'd and remanded USCA5 1654; , 520 F. 2d 1061 (2d Cir. 1977), cert. denied, 434 U.S. 1013 (1978).
 Richard A. Posner, ECONOMIC ANALYSIS OF LAW 3 (3d ed. 1986).
 See Coase, The Problem of Social Cost, supra note 58. The Coase Theorem is effective at least in the absence of factors such as transaction costs, wealth or income effects, and strategic behavior. See id. (transaction costs). See also, e.g., Donald Regan, The Problem of Social Cost Revisited, 15 J. LAW & ECON. 427 (1972) (strategic behavior). Compare RONALD H. COASE, Notes on the Problem of Social Cost, in THE FIRM, THE MARKET, AND THE LAW 157, 170-74 (1988) (suggesting that income effects are unlikely to be significant, at least in contexts not involving irreplaceable goods).
 Even if one interprets copyright's economic goal as being solely the use of incentives to "promote knowledge," so that satisfying the copyright owner's personal tastes would not count as an independent value, the empirical answer to suppression questions would not be easy: in a given case enforcing any particular type of suppression would both keep some knowledge secret, and yield long-term incentives that could aid knowledge in the long run (because authors who can suppress have a copyright worth more than authors who cannot). Cf., Frank I. Michelman, Property, Utility & Fairness, 80 HARV. L. REV. 1165 (1967) (the effects of demoralization on productivity). Which of the two potential effects on knowledge would be greater (the loss from enforcing suppression or the gain from long-term incentives) cannot be determined a priori.
 For a fuller discussion of this issue, see Wendy J. Gordon, The Right Not to Use (unpublished manuscript on file with the author).
 Additional reasons might include, e.g., the potential nonmonetizability of first amendment values. See Gordon, Fair Use, supra note 4, at 1631-32.
 I base this theory in part on the work of James Lindgren in the blackmail area. See James Lindgren, Unraveling the Paradox of Blackmail, 84 COLUM. L. REV. 670 (1984) (discussing the three-party structure involved). For an economic analysis of blackmail stressing other aspects of blackmail activity, see Ronald H. Coase, The 1987 McCorkle Lecture: Blackmail, 74 VA. L. REV. 655, 673-74 (1988).
I am indebted to Warren Schwartz for suggesting the potential relevance of the blackmail literature to this problem.
 Information can implicate different issues from literary expression and other intellectual products; for purposes of this very general discussion, however, I shall group all together under the rubric "information."
 There is another factor that may be at work here as well: the idea/expression dichotomy. Since under current law copyright owners cannot prevent others from using their ideas, it could be argued that little suppression of note could occur; it might be suggested that a critic deprived of the privilege to quote could nevertheless communicate effectively.
For simplicity's sake, therefore, assume that in the following examples whatever the defendant has taken from the first artist's work could be considered copyrightable expression rather than simply “idea” and that the use of the copyrighted expression is somehow essential to the effectiveness of the planned derivative work.
 As economist Michael L. Katz writes of the similar problem in the research and development area:
In the absence of perfect discrimination, the firm conducting the R & D will be unable to appropriate all of the surplus generated by the licensing of its R & D, and the firm will sell its R & D results at prices that lead to inefficiently low levels of utilization by other firms.
Michael L. Katz, An Analysis of Cooperative Research and Development, 4 RAND J. ECON. 527, 527 (1986).
 See Richard A. Posner, Conventionalist Defenses of Law as an Autonomous Discipline 17 (September 21, 1987) (unpublished manuscript, on file with the University of Chicago Law Review) (using pecuniary effects to explain why landowners who create certain positive spillovers are not entitled to payment from those who benefitted).
 Journalistic ethics undoubtedly prohibit reviewers from accepting subsidies for doing hostile reviews.
 There is a fairly extensive literature on the controversial question of whether managerial discretion exists and if so what impact it has and what should be done about it; all I mean to suggest here is the simple possibility that managers in complex corporations do not always make the same decisions that an individual owner of a corporation would.
 In an individual, a taste for risk or laziness might be a legitimate part of her utility curve, but a manager is supposed to act unselfishly on the part of the corporation. There is a large literature on these agency problems.
 It might be argued that a taste for laziness or risk aversion are simply preferences that deserve the same respect under the notion of consumer sovereignty as other desires. However, we are not talking here about the risk aversion or laziness of the copyright owner, but of some person who is fortuitously placed within the licensor organization to be able to control licensing decisions. Whether gratifying such a person's taste in regard to laziness or risk serves greater economic ends (as, e.g., a form of compensation) is an even more complex question than the question of how economics should analyze an owner's taste for such things.
 Of course, such licenses might be granted; I offer here only an abstract analysis which would need to be empirically verified.
 Wealth effects are, roughly, the impact on one’s preferences brought about by a change in wealth, including the change brought about by being given, or being denied, an entitlement. See, e.g., E.J. Mishan, The Postwar Literature on Externalities: An Interpretive Essay, 9 J. ECON. LITERATURE 1 (1971) (the allocative impact of wealth effects illustrated at 18-21, though not explicitly in the context of the Coase theorem).
 For an excellent numerical example, see id. at 18-21. It is well recognized that a divergence often exists between the price that a potential buyer would be willing to pay for a resource he does not own, and the price that the same person would demand before he would sell that same resource if the law had initially awarded its ownership to him. What is less clear is what terminology, explanations, and characterizations are best employed for discussing the phenomenon. For a valuable discussion suggesting, inter alia, that traditional “wealth effects” do not fully explain divergence between willingness-to-accept and willingness-to-pay, see Elizabeth Hoffman & Matthew Spitzer, Willingness to Pay vs Willingness to Accept: Legal and Economic Implications, 71 WASH. U.L.Q. 59 (1993).
 For a dramatic hypothetical example, see Alfred C. Yen, Restoring the Natural Law: Copyright as Labor and Possession, 51 OHIO ST. L.J. 491, 518-519 (1990) ("flip flop" of rights).
 See Coase, supra note 58.
 See, e.g., Calabresi & Melamed, supra note 25, for a classic explanation of market-mimicry.
 For further exploration, and for citation to relevant literature, see Hoffman & Spitzer, supra note 97.
 The impact of endowment or wealth effects is sometimes exaggerated. See Coase, Notes on the Problem of Social Cost, supra note 82 at 170-174 (discussing arguments re the presumed effect of changes in legal position on the distribution of wealth and on the allocation of resources).
Professor Coase argues that the impact of wealth effects can be overstated because, among other things, if the legal rules are known in advance, the prices of applicable resources will likely alter in a way that minimizes such effects; in addition, he suggests, contractual provision for contingencies may be available to mitigate some changes in legal rules. See id. at 157. See also id. at 170-174. Neither of these devices are likely to eliminate the wealth effect-- here "pricelessness"-- in the context of authorial suppression of embarrassing criticism, however.
 That is, while I predict that the law's assignment of rights in organs is likely to have a distinct effect on a kidney's allocation, it is a more complex question whether the law's assignment of rights in organs will have much of an effect on the allocation of other resources.
 These points are also explored in Gordon, Private Censorship, supra note 73 at 1042-43; also see Gordon, Fair Use, supra note 73, at 1632-36 (anti-dissemination motives).
 I am indebted here to Alan Feld.
 This belief is rather controversial. See, e.g., such classic sources on the debate as the Symposium on Efficiency as a Legal Concern, 8 Hofstra L. Rev. 485 (1980) and Richard A. Posner, THE ECONOMICS OF JUSTICE (1987) for further discussion of the question of whether utilitarianism and economics are truly linked in this way.
 For further examination of the author's views on the interplay of economic and non-economic modes of normative reasoning, see Wendy J. Gordon, Truth and Consequences: The Force of Blackmail's Central Case, 141 U. PA. LAW REV. 1741 (1993).