Tom Bell on copyright reform; the Hayekian knowledge problem and copyright terms

by Stephan Kinsella on January 6, 2013

I had the pleasure of finally meeting in person law professor Tom Bell (one of the handful of IP attorneys to publicly oppose IP)¹ at Libertopia in San Diego last September.²

In a recent interview by Jerry Brito on the excellent Surprisingly Free podcast (one of my favorites), Bell discusses his views on copyright and his chapter “Five Reforms for Copyright,” in the recent book Copyright Unbalanced: From Incentive to Excess, edited by Brito (which is, ironically, and unfortunately, not published under any kind of Creative Commons license nor made available free in PDF form or otherwise; Mercatus page). [Brito’s book is reviewed by Tom Palmer here.]

Tom Bell on copyright reform

Tom W. Bell, professor of law at Chapman University and author of the concluding essay in Copyright Unbalanced, a new book edited by Surprisingly Free’s own Jerry Brito, discusses the ways in which copyright has evolved over time and why reform is vital.

Bell differentiates copyright from other types of property, arguing that conflating the two terms causes great confusion amongst laypeople and, over time, corrodes the value placed in tangible property rights. According to Bell, copyright is a privilege created by statute that doesn’t exist in a state of nature and is not recognized by common law.

As a special type of economic good, copyright must be treated differently than tangible property rights, according to Bell, who outlines five proposals for copyright reform.

While Bell is not opposed to copyright, he argues that copyright enforcement has gone too far, and lawmakers should structure policies to lead us towards a world in which we conceivably do without it.

In his excellent and concise chapter, Bell argues that copyright has changed so much since 1790 that it has deviated dramatically from the original idea for copyright in the Constitution, and now poses a grave threat to our freedom and liberties. Bell argues for a return to the Founders’ Copyright, among four other reforms:

Reinstate the Founders’ Copyright Act,
Withdraw the US from the Berne Convention,
Develop misuse doctrine into an escape from copyright,
Focus copyright policy on consumers’ costs, not producers’ profits, and
Reconceive “IP” as “Intellectual Privilege.”³

Regarding relabeling IP as “intellectual privilege,” see my post Intellectual Poverty, suggestion other possible terms that could be used to more accurately describe patent, copyright, and other IP laws.

The Founder’s Copyright would be much narrower than the current metastasized copyright system: it would offer an initial 14 year copyright term followed by another 14 year term, if applied for (compared to life of the author plus 70 years, now, and automatic copyright protection); no criminal sanctions; only maps, charts, and books covered, but not movies, photographs, paintings, sculptures, plays, music, architectural works, boat hull designs, and software; only exact, literal reproduction is covered, not quotations of a portion of the work, not derivative works (like sequels to a novel). This would be a radical improvement in copyright and would probably get rid of 95% or more of its current mischief.

Brito is also somewhat skeptical of copyright and IP but is not an abolitionist; he adopts the utilitarian approach which holds that we want something between zero copyright protection and the very strong copyright we have now (see his discussion in Geoff Manne on copyright and Brito: What’s Wrong With a Copyright Alert System?; also Republicans More Radical than Libertarian Copyright Moderates and Stan Liebowitz on copyright and incentives). In Bell’s wide-ranging interview with Brito, he discusses a variety of problems with modern views on copyright. For example, he challenges the idea that copyright should be thought of as property; Brito seems to agree (see Classifying Patent and Copyright Law as “Property”: So What?); he observes the danger of this analogy: that the limits put in place on copyright to prevent its excesses (like fair use) could be imported back onto real property law and weaken real property rights (Tom Bell: Copyright Erodes Property?); and other issues.

The Hayekian Knowledge Problem and the Copyright Term

One other thing worth mentioning—in this interview Bell and Brito talk about the (Hayekian) “knowledge problem” being one reason it’s difficult to know what optimum copyright term to set (Brito and Manne make similar observations in another podcast episode; see Geoff Manne on copyright). And in the introductory chapter to Copyright Unbalanced, Brito writes (my bolding added):

You should be skeptical of Congress’s ability to develop a rational policy given the knowledge problem copyright presents and the public choice pressures at work. [p. 7, Kindle Edition]

… The fact that copyright does not emerge spontaneously does not mean it is illegitimate, or even unwise. What it does mean is that we should be cautious about how we create such a property right and how we define that right.

As F. A. Hayek showed, created orders are handicapped by the limited knowledge of their planners. This is in contrast to emergent orders, like customary or common law, which evolve slowly through trial and error. Such bottom-up processes are the product of human action, but not human design. As a result they incorporate dispersed knowledge and tend to match economic efficiency.

Copyright is a created order. Not only did it not emerge spontaneously, it is in fact a tool to create scarcity through state action. But why would we want to introduce scarcity where there is none? And why would we want to limit use where there is no conflict? [p. 11]

Good question. And, in a section entitled COPYRIGHT’S KNOWLEDGE PROBLEM:

The knowledge problem inherent in legislatively created rights is one reason why a delicate balance is difficult to reach. Not only are the contours of copyright centrally planned, but as Richard Epstein has noted, “There are in fact no ‘natural’ boundaries here, similar to the metes and bounds of land” to guide policy makers.

Conservatives and libertarians tend to easily recognize the knowledge problems in other government programs. For example, consider subsidies for renewable energy, which are also arguably meant to promote the production of a public good.

How does Congress know that the market is not already providing the right amount of investment in renewable energy? Without a government subsidy, there would still be investment in renewable energy technologies. By creating a subsidy, Congress is saying it doesn’t think it’s enough, but it has no way of truly knowing that.

Putting aside how Congress can know that there should be more investment in renewable energy, the other question is, how much investment is optimal? Without a market process to guide such investment, Congress can’t know how much is enough. So when Congress offers a certain amount of subsidy, it’s guessing. It’s likely offering too little or too much, with each error introducing its own inefficiencies.

The same challenges exist for copyright. Without copyright, there would still be songs written and movies made. Congress just thinks there wouldn’t be enough. So, it offers a subsidy in the form of copyright protection to incentivize more creative output. The same questions we asked about renewable energy present themselves: How does Congress know we wouldn’t have “enough” creative works without copyright? And assuming it knows that, how does it know the right amount of incentive to offer?

The first Copyright Act, which the Framers legislated, was limited to maps, charts, and books, which meant that other kinds of expressive works, such as songs and plays, were not protected. The Framers must have thought they would get enough such works without a subsidy. Were they wrong? Today songs and plays are protected, but fashion designs are not. Do we have the right balance now? How do we know?

The first Copyright Act also set a copyright term of 14 years renewable for 14 more, and required authors to register their works before receiving protection. Was 14 years enough, or too little? Today the copyright term is life of the author plus 70 years. Is that too much, or not enough? How do we know?

Also, requiring registration and renewal meant that only those authors who sought a subsidy got one. Only 5 percent of books published at the time were registered for copyright, and a small minority of their authors sought renewal, making the subsidy modest and somewhat self-regulating.

Today there is no requirement for registration or renewal, and anything you write is automatically protected whether you want it to be or not. Does such a massive trade of access for incentive get us closer to striking the delicate balance? How do we know?

This is not to say that because the ideal contours of copyright are unknowable, we should therefore have no copyright. After all, we also don’t know that zero is the right amount of copyright. However, understanding that there is a knowledge problem in copyright is useful. It tells us that while Congress may inevitably have to make an educated guess when it designs copyright, it should be humble and not take lightly the possibility that it will strike the wrong balance and introduce serious inefficiency. [pp. 13-15; endnotes omitted]

This entire approach is problematic, as I discuss in Knowledge, Calculation, Conflict, and Law and my post Knowledge vs. Calculation.

According to Hayek, there is dispersed knowledge in society that is difficult for any central planner to gather (or keep up with as it dynamically changes), and difficult even for the individual actors who possess it to articulate in an objective way, since it is “tacit.” Yet people’s dispersed knowledge affects how they buy and sell, and thus gets “reflected” in the overall price structure of the economy. Even this conception of prices “conveying” or “encoding” knowledge is problematic, in my view (as mentioned in the two linked pieces), but at most, it means that individual knowledge of relative resource scarcity is reflected in prices. This is because the knowledge of individuals, whether tacit or otherwise, that affects the prices they form, is these individuals’ knowledge (or forecasts or opinions) about the demand for, usefulness of, and supply and availability of scarce resources. That is, according to Hayek, the free market economy and its price system provides a way to help economize in the face of scarcity; it helps allocate scarce resources more efficiently, by means of a price system embodying or reflecting or transmitting various knowledge individual actors have about relative scarcities of, and supply of and demand for, scarce resources.

The Misesian case against central economic planning is that in the absence of a free market price system, there is no way to calculate—no way for the entrepreneur to forecast and compare possible plans on terms of profit maximization, because resources are heterogenous and incomparable. The future is always uncertain, but entrepreneurs have no way of comparing contemplated projects in cardinal terms, without some common, numerical frame of reference. They cannot even determine after the fact whether a project was as successful or more successful than anticipated, or than alternative projects could have been, without some cardinal means of comparison. A free market price system permits such calculation because the entrepreneur can calculate in money terms and compare various possible projects and alternative uses of scarce resources and select the project expected (in his entrepreneurial judgment) to yield the greatest monetary profit. And he can use money prices to assess how a project has done (cost accounting). In this way, to the extent entrepreneurs are successful (profit), resources are allocated to their most valuable uses. The Hayekian case against central planning emphasizes the difficulty a central planning agency has in acquiring knowledge that is dispersed and tacit knowledge. But even here, the knowledge is knowledge of relative scarcities of resources, of the supply of and demand for these resources.

But is this the reason the legislature has difficulty knowing what the optimum copyright term (and scope, etc.) should be? No. There are, instead, several other, and far more fundamental, reasons legislators cannot be expected to get the “right” copyright term. First, there is the political process itself: entrenched interests like Disney, Hollywood, the music industry, and so on, push Congress to keep ratcheting up copyright scope and penalties and term, not because of any finding that it is optimum, but for political reasons (which Brito acknowledges, in referencing “the public choice pressures at work”; in observing that “humility and restraint are not qualities readily found in the modern Congress”; in observing that “The process by which the terms and scope of copyright are decided is a political one” prone to “rent-seeking”).

Second, in the case of allocation of scarce resources we know that each resource has an exclusive use and can only be used for use project A or B. If the resource is used for the “wrong” plan then there is waste. Thus, there is a need to have some system that permits optimum allocation of the resource. This is the function of a free market in which private property rights are respected and money is used for calculation. But the copyright monopoly is justified by its proponents as stimulating the production of creative works—essentially, patterns of information, knowledge—that have some “value” to society that is greater than the costs imposed on the economy by the copyright system. Information patterns are not scarce resources. It is difficult to imagine how the price system, even in the Hayekian conception, which is useful in helping allocate scarce resources to their most efficient uses by overcoming the knowledge problem faced by central planners, provides information about how much value a copyright monopoly generates and how much cost it imposes.

Tabarrok’s Innovation vs. patent strength curve

Third, the idea that legislators face a knowledge problem in knowing where the optimum is for copyright strength and term, relative to costs it imposes and benefits it allegedly generates, presupposes that there is an optimum. But there is no reason whatsoever to presuppose this. (See Tabarrok: Patent Policy on the Back of a Napkin.) In the analogous case of subsidies for renewable energy, Brito notes: “It’s likely offering too little or too much”, as if zero is the obviously wrong choice even here. Why? Zero is the right choice for both renewable energy subsidies and for the subsidy of copyright. In the case of renewable energy subsidies, at least the subsidies likely increase renewable energy R&D and technology, since money is poured into it. This still does not mean it’s worth it, since renewable energy tech is developed at the expense of other technology or projects that are now unaffordable since tax dollars were siphoned away to stimulate renewable energy. But in the copyright case, there is no reason to believe that there is more creative work on net stimulated because of the copyright system in the first place; there is certainly good reason to believe that the patent system decreases net innovation. At the very least, the copyright and patent systems lead to severe distortion of their respective fields: culture and creative works, innovation and technology (Destructive Creation; Leveraging IP).

Imagine a similar approach being taken to the prison sentences being imposed for marijuana use. The prohibitionist might say that zero penalty is “obviously” “too low,” but he might also admit that torture and the death penalty for possession of an ounce of pot is obviously “too high”; thus, there is some optimum “in between.” It might be difficult to know what this optimum penalty is: is it a $100 fine? A year in prison? Ten years in prison? And then the prohibitionist could say, “Well, because of the Hayekian knowledge problem, it’s hard to know what the optimal penalty should be. But we know it’s between zero (legalization) and the death penalty; given this; a seemingly-arbitrary 5 year prison sentence is “closer to optimal”, or “better,” than a zero penalty or capital punishment. But the reason you can’t find the “right” punishment is that there is no optimum; or, rather, the optimal punishment is zero. But even if we concede that there should be some finite (non-zero, non-“infinite,” i.e. capital punishment) penalty for marijuana possession, and even if we concede we cannot know what the optimal amount of punishment should be, it is not because of the knowledge problem. For how could the most efficient free market imaginable tell us the “right” or most efficient amount of punishment to impose? The free market is efficient in the sense that it permits the efficient allocation of scarce resources. It is this ideal situation that the “knowledge problem” faced by central planners is contrasted to. What has this to do with coming up with a punishment of people who are engaged in a victimless crime?

And it is similar in the case of copyright, since there is no reason to think that there is a curve with a peak. Brito writes: “we also don’t know that zero is the right amount of copyright”. Well what kind of argument is this? If you don’t have some kind of proof that some copyright does increase social welfare, then why have it at all? Utilitarians seem to admit copyright puts limits on speech and imposes various costs on individuals. They believe this cost is justified if it leads to net gains in cultural innovation. But they have no evidence that it does. Shouldn’t the burden of proof be on them to show that their utilitarian claims are true? Until they do, shouldn’t we refrain from imposing copyright on society? Just consider all the economic and societal interventions Congress could engage in if it could say “well, we are not sure this law is optimal or producing good results, but on the other hand, we don’t know that not having the law is optimal either”! How is this a justification for having a law and then trusting Congress to tweak it? If Brito is right that Congress cannot know what the optimal copyright term is, and cannot even know that zero is not the right term … why have copyright at all, even on utilitarian grounds?

Copyright Duration and the Mickey Mouse Curve

The truth is that, as in the case of drug prohibition, the “optimal” amount of copyright is zero. This is true on both moral and empirical grounds. Moral: since copyright infringes on property rights (see Intellectual Property Rights as Negative Servitudes); empirical, since there is no evidence that patent or copyright lead to any net increase in innovation or artistic creation at all, much less an increase that is greater than the costs of such a scheme (and it is undeniable that IP distorts innovation and the creative fields) (see The Overwhelming Empirical Case Against Patent and Copyright; also Legal Scholars: Thumbs Down on Patent and Copyright). If the “curve” is monotically downward sloping—meaning that the greater the copyright protection, the worse off we are both in rights terms and also in economic terms—then the reason legislators cannot find its optimum is not because of the knowledge problem that plagues central planners trying to efficiently allocate scarce resources, but because there is no peak to find. But they truth is they are not trying to find the optimum; they do not even care. They keep ratcheting copyright terms upwards, enhancing penalties, increasing copyright scope, pushing it onto other countries, at the behest of Big Content special interests. They do not pretend to have the faintest clue that there even is an optimum, much less that they know, or that anyone can know, what it is.

(Update: Mises himself, a utilitarian of sorts (more of a consequentialist, I think) in the field of political ethics, also had a few murky statements on IP, which I note in Mises on Intellectual Property. An Austrian economist friend observed to me that:

The … thing that struck me as I read [your post] was Mises’s view in Human Action (pp. 657-659). He, too, claims that there is no optimum with respect to the configuration of patents and copyrights, but that different configurations of IP render different patterns of resource use. Each pattern of resource use is optimal for a given configuration of IP. But there is no way to compare the pattern of resource use under one configuration with the resources use under a different configuration. This seems to be a generalizable principle for Mises, that the configuration of property rights lies outside the market’s economizing process of resource use and, in fact, is a requisite for it.

In other words, the problem that plagues central economic planning, whether it’s characterized as Mises’s calculation problem or as Hayek’s knowledge problems, and which do not pose a problem for a genuine free market order, is not the reason a legislature cannot determine whether or not to have copyright or what form it is to take.)

Finally: the fundamental reason it’s impossible to compute these things in the first place is the subjective and intersubjectively incomparable and ordinal, not cardinal, nature of value. Not to mention the problem with counterfactuals and fact that there are many unseen, undeterminable costs to any such regulatory scheme.

Utilitarian scholars who argue the copyright system should be reformed because we are on the wrong side of the peak, but that there is a peak and that it’s possible for a corrupt legislature of a criminal state to find the optimum, or even to want to find it, are wrong. The optimum amount of copyright is the amount that is compatible with freedom and property rights: zero.