WHAT’S WRONG WITH COPYING? – Lunney Review

Vol. 7, No. 2 (March 2017) pp. 18-26 

WHAT’S WRONG WITH COPYING? by Abraham Drassinower.
Harvard University Press, 2015, 288 pp., Hardcover, $39.95 

Reviewed by Glynn S. Lunney, Jr. 
Texas A&M University School of Law 
glunney@law.tamu.edu 

The book, WHAT’S WRONG WITH COPYING?,¹ has two objectives. The first objective is to show that the dominant economic approach to delineating copyright’s scope and limitations is fundamentally flawed. The second is to derive from copyright doctrine directly a more persuasive and effective alternative to understanding and ultimately defining copyright’s scope and limitations. Unfortunately, the book fails to achieve either. 

The book presents three principal critiques of the dominant economic approach to copyright. The book’s first critique of the economic approach is that copyright does not extend or deny its protection to any given creative effort based upon the value the effort generates.² Rather, it focuses on the nature of the creative effort, and asks whether the creative effort at issue represents authorship. We cannot, therefore, understand copyright generally, or copyright doctrines, such as originality, specifically, as an attempt to maximize value, or so Drassinower contends. This critique is a strawman. 

As Drassinower himself recognizes, in economics, we distinguish between public and private goods not based upon the value they create, but on the presence or absence of rivalrous consumption.³ An economics or instrumental approach is not thus limited to differentiating things based upon value. In intellectual property, we distinguish between copyrightable works of authorship and patentable inventions. We distinguish them not based upon the value they create, but based upon whether the creative activity at issue results in a work of authorship or an invention. These legal categories are not, however, deontological constructs, but in my view, reflect underlying instrumental concerns. I would argue (and indeed, have argued) that we separate creative activities into patent and copyright because the cost investment and cost recovery structures associated with these two creative activities differ.⁴ Because the cost structures differ, we need differing legal regimes, with differing prerequisites and differing scopes of protection to identify when a market failure is likely, and to remedy it without imposing undue social cost. For example, the requirement that an invention be nonobvious before it is entitled to a patent makes sense given the cost investment and recovery structure for the sorts of creative activities we protect through patent. There is no need for patent to protect original and novel inventions that are also routine, trivial, or obvious. Given the intrinsic difficulties associated with copying the sorts of things that we protect as patentable inventions, the ordinary workings of the market will tend to ensure the efficient development of routine or obvious advances.⁵ Extending patent protection to such obvious advances would impose social costs, without any return benefit. In contrast, the originality standard better defines when market failure becomes likely, in the absence of some special protection, for the sorts of creative efforts we protect as copyrightable works of authorship. Consider the counterfactual: If we required nonobviousness for a movie to receive protection, only a handful of movies over the last twenty years would likely qualify. More importantly, the vast majority of movies, including virtually all romantic comedies, would not satisfy the standard and would therefore be left unprotected. Without protection, it is at least plausible that many of these movies would not be forthcoming. Originality is not therefore a deontological principle in copyright, but an instrumental one. Originality best defines, for the sorts of creative efforts we protect as authorship, when market failure becomes likely.

Similarly, the differences between patent and copyright’s respective scopes of protection also reflect instrumental concerns. For example, patent protects against independent creation of the same invention, but copyright does not. This difference does not reflect some deontological difference in the intrinsic nature of patent and copyright, but real-world instrumental concerns. Specifically, near-simultaneous independent creation of competing inventions is more likely than near-simultaneous independent creation of competing works of authorship. For example, assume that three companies are racing to develop cures for three diseases. They are all using similar methodologies and similar information, yet one company will win the race in each case, coming in a month or two ahead of the others. If the patent infringement standard were “copy,” rather than “make, use, or sell,” then the winning company would have at most a month or two to recoup its research expenses before the others entered the market and near-perfect competition began.⁶ Under such an infringement standard, none of the companies would recover their research investment. Foreseeing that result, none would make the investment. A copying standard would not therefore solve the market failure problem for the sorts of creative activities we protect as patentable inventions. In contrast, the “copying” standard works reasonably well to resolve the market failure for the sorts of creative activities that we protect as works of authorship. Independent creation of identical movies, for example, seldom occurs, and even where two movies happen to explore the same topic at the same time, the movies invariably differ sufficiently that competition between them will not drive prices to marginal cost. Thus, the presence of A BUG’S LIFE or DEEP IMPACT did not reduce the ticket price for seeing ANTZ or ARMAGEDDON. 

The fact that we cannot differentiate between authorship and works of authorship, on the one hand, and a variety of non-authorship creative products, including patentable inventions, on the other, based upon value is simply irrelevant. We provide different prerequisites and scope of protection for inventions (through patent) and works of authorship (through copyright) because the cost investment and recovery structure for these two types of creative endeavors differ. Because of the differing cost structure, different prerequisites identify when a relevant market failure is likely, and different scopes of protection remedy the relevant market failure at the least social cost. 

The second critique the book offers is to point out that the empirical balancing of costs and benefits an economic approach requires can be indeterminate, given the data available. In Drassinower’s view, this occasional indeterminancy renders the economic approach unhelpful.⁷ While true in some cases, this sort of indeterminancy is not a problem unique to the economic approach. Any approach that we might use for drawing a line between, for example, protected expression and unprotected idea will have the same problem. Whatever approach we use, there will be cases where the line can plausibly be drawn on either side. Indeterminancy is not therefore a problem unique to the economic approach. Furthermore, even if this critique somehow cut more deeply as applied to an economic approach, better data and proper application of the burden of proof would seem the proper response, rather than abandonment of the economic approach altogether. More importantly, even if there are hard cases under an economic approach, there will also be easy ones – cases, such as Google Books, where the expected welfare gains from finding fair use far exceed any plausible estimate of the welfare losses.⁸ We should not get so caught up in the search for a perfect line-drawing tool that we miss the obvious: The economic approach does a better job than any conceivable alternative both in identifying the salient characteristics of an easy case and in identifying the right outcome. 

The book’s third critique of the economic approach, and perhaps the most interesting, is to suggest that by embracing the economic approach, proponents of narrower copyright (or “copyright minimalists”) have allowed the proponents of broader copyright (or “copyright maximalists”) to set the terms of the debate and have thereby already lost.⁹ As support for this critique, Drassinower points out that over the last two hundred years the economic or instrumental approach has been the dominant paradigm in Anglo-American copyright jurisprudence and over that same time period, copyright protection, both in scope and duration, has steadily expanded.¹⁰ The weaknesses in this argument are, however, readily apparent. As Drassinower himself admits, this historical correlation does not establish causation.¹¹ We have, after all, seen the same expansion in copyright protection in France and Germany, where non-instrumental, rights-based perspectives hold sway. Indeed, some of the expansion in United States copyright protection has come about due to efforts to harmonize our consequentialist copyright laws with those of deontological Europe. Nor should we discount the substantial, if sometimes hidden, role rights-based accounts play within United States copyright law itself. 

In addition, for much of its history, the direct costs of copyright have been largely hidden, if not altogether lost, in the more general welfare losses associated with the natural monopoly cost structure and the resulting oligopolistic market structure of the copyright industries. Given the technology and market structure in 1950, it did not matter whether the copyright on a book lasted five years, fifty years, or five hundred years. For virtually all books, there would be an initial print run, after which the book went out of print. Given the technology and market structure at the time, there was essentially no chance that another publisher would pick up the book and reintroduce it, whether the book was under copyright or not. The repeated extension of copyright’s term through 1998 did not therefore reflect a failure of the consequentialist approach, but an accurate assessment that the costs of longer copyright, given the technology and market structure at the time, were essentially zero. True, the benefits were also essentially zero, and personally, I would have preferred Congress to strike a different balance. But at the time, given the technology and market structure, neither the costs nor the benefits of a longer copyright term were particularly compelling. That Congress chose one result, rather than the other, in these historical debates is not therefore a meaningful indictment of the consequentialist approach. 

Copyright has always been driven by changes in the cost of creating and distributing works of authorship. In the analog era, just as we could offer a reasonable instrumental justification for a longer copyright term, we could also offer a reasonable instrumental justification for a broad derivative or reproduction right. With analog technology, costs and market structure often meant that only one film version of a novel would be created. Given that, there was little reason not to require that one film version to be authorized. Only one license would need to be negotiated; so transaction costs would be small. Moreover, given the distribution market for films at the time, requiring a license imposed no deadweight losses. Whether a film was an authorized or unauthorized derivative work, the admission price at the theater remained the same. At this time, a broad derivative or reproduction right made reasonable instrumental sense. 

Digitization and the Internet have fundamentally changed the cost of, and the associated market for, the creation and distribution of works of authorship today. With the technology available today, works that have been out of print for years could be, but for unduly long copyright, reintroduced into the market.¹² Rather than a narrow trail of derivative works, any given novel today might spawn thousands of competing, overlapping derivative works. Requiring licenses for all these uses becomes impracticable, and the deadweight losses if we insist on licensing become substantial. Just as the introduction of the printing press changed the cost of creating and distributing works of authorship in a way that justified the initial creation of copyright, so too today, changes in the cost of creating and distributing works of authorship due to the digital revolution have changed the cost-benefit calculus of copyright. It is this changed calculus that makes today’s copyright both unduly long and overly broad. Consequentialist or instrumental reasoning can justify both copyright’s initial creation and today’s current need to cut it back; a deontological or rights-based approach cannot.¹³

Rather than concede the field to the copyright maximalists, the consequential approach is beginning to push back and to push back strongly. True, we are just at the beginning. Progress in narrowing copyright is painfully slow. Progress is slow not because instrumentalism is the wrong tool, but because judges are people, and people fear change and cling to the familiar (and copyright owners do not want to give up what they perceive as theirs). As a result, as we move forward to a happier world of narrower copyright, there is likely to be some backsliding. Already, however, the knee-jerk tendency to expand copyright to encompass new uses, which might have been appropriate in the analog era, is starting to fade. The defeat of SOPA and PIPA in Congress, and the expanding scope of fair use in the courts over the last fifteen years perhaps best illustrate this slow progress. 

As a consequentialist proponent of minimalist or narrow copyright, I am strongly encouraged by what I have seen over the last fifteen years. We are just now beginning to gather the sorts of empirical evidence that we need to understand the role copyright plays in the creative process, as well as the costs it imposes. We are just now moving beyond the casual assumptions and bedtime stories that have dominated the discussion and helped to drive copyright’s expansion over the last two hundred years. We are just now beginning to win the battle for a limited copyright that actually promotes “the progress of Science.” Rather than celebrate with me these initial signs of success, Drassinower asks me instead to abandon the consequentialist approach on the cusp of victory. 

Now, if he could offer guaranteed victory through his alternative, I might be tempted. But he cannot. All he can offer is an unpersuasive principle derived from a flawed methodology. As his methodology, Drassinower examines some of copyright’s key cases and doctrines in an attempt to discern the underlying copyright principles that animate copyright law. 

As applied to case law, this methodology is intrinsically flawed. In the natural sciences, we can make observations of, for example, objects falling and derive from our observations a hypothesis, and eventually, a theory of gravitational attraction. This methodology works in the natural sciences because each observation is driven by the same underlying natural law. There is no corresponding natural law of copyright. Most obviously, copyright, unlike gravity, is historically contingent. If Drassinower applied his methodology to 16th century copyright (no protection), to 18th century copyright (short, limited protection), or to 20th century copyright (long, broad protection), he would derive very different underlying principles. Indeed, even within the 20th century, applying his methodology to United States copyright law before the Supreme Court’s decision in Feist Publications (sweat-of-the-brow justifies copyright) and after the decision (sweat-of-the-brow does not justify copyright) would paint quite different pictures of copyright. Stare decisis or statutory language may limit the discretion of individual judges in particular cases, but at the end of the day, judicial decisions remain ad hoc resolutions of the particular disputes that parties, for their own idiosyncratic reasons, choose to bring, and the arguments the parties choose to make.

Holding up a mirror to judicial opinions in an attempt to see the underlying principles that bind them is futile. There is no there there. All that Drassinower can see in such a mirror is an imperfect reflection of himself. He is not discovering the hidden principles of copyright through his methodology, but only the values, preconceptions, ideologies, and biases that he brings to the search. This is most evident in his selection of a small subset of copyright cases, focusing on Feist Publications, and the corresponding copyright doctrine of originality as the basis for his methodology. Had he started with some other case, such as Campbell v. Acuff-Rose Music, or some other doctrine, such as Congress’s decision to extend copyright protection to computer programs, his methodology would have led to a vastly different perspective on copyright’s fundamental principles. 

Even if we put to one side the deeply flawed methodology, however, the principle Drassinower derives – that copyright exists to prevent forced or compelled speech – is equally problematic. Most obviously, it is not literally accurate. I am familiar with instances of what I would consider compelled speech. It is compelled speech when a prisoner of war is tortured until he makes a video renouncing his home country. It is compelled speech when, before the Fifth Amendment prohibition on self-incrimination, a criminal defendant could be forced to either admit guilt or perjure himself. It is not compelled speech when a recording of the song “Eye of the Tiger” is played in connection with Kim Davis’s release. True, that may upset the band-members of Survivor, but characterizing that use as “compelled” speech is simply inaccurate. No one is torturing or imprisoning the band to force them to play the song; they are not even physically present. While playing a recording of the song may lead some to believe (mistakenly) that Survivor supports Kim Davis or her position, the First Amendment’s answer of more speech seems fully adequate to redress that concern. 

More importantly, for me, however, Drassinower’s principle misses entirely the point of copyright. I have been the victim of copyright infringement. When I submitted my Fair Use article to law reviews in the spring 2002 cycle, a student editor at a leading law review took the Word file that I submitted, changed the title of the article, substituted his name for mine as the author, and submitted it as his own work to various law reviews. That fall, the article published first under his name in the SEATTLE LAW REVIEW and then came out a few months later under my name in the BOSTON UNIVERSITY LAW REVIEW. When I discovered the infringement, the notion that he was forcing me to speak never crossed my mind. Nor was my concern that he copied. Rather, my concern was almost entirely the moral rights concerns of attribution and integrity.¹⁴ Another was claiming my work as his own, and he had published the initial draft, which I had revised substantially before publication. 

I recognize, of course, that copyright does not exist to serve my idiosyncratic preferences. But I do not believe my experience in this regard is unique. In all my years as a copyright attorney, no client has ever approached me and complained that another, by copying her work, is forcing her to speak. Almost always, the client’s complaint is that someone has taken what belongs to them. The closest I have seen to the forced speech concern is when private (and usually embarrassing) letters, texts, or e-mails are published, but even there, the concern is more privacy than forced speech. 

I would be perfectly willing to put my personal feelings aside, however, if the principle Drassinower offered, despite its imperfect fit with my own experience, would prove more effective at reigning in copyright’s excesses on the key battlefronts of today. Yet, it does not. In my experience as a copyright lawyer, originality is seldom decisive; thus, it strikes me as an odd place to center copyright law. Rather, the key battlefronts, as I see them, are the scope of fair use, whether the copyright term should be extended another twenty years, and the scope of secondary liability. The instrumental approach has helpful, and in my view, very persuasive arguments to offer on each of these issues. It has already shown success in expanding fair use and in beating back attempts to expand secondary liability. In these contexts, the instrumental approach focuses precisely on what we should care about – the great things, such as a searchable database of all books, and perhaps the Internet entirely, we risk losing if we interpret copyright over broadly. 

Drassinower employs his principle of forced speech to argue for similar outcomes on each of these issues. Thus, on Google Books, he insists under his principle that Google is not speaking for itself; therefore Google’s copying should not count as actionable use under copyright.¹⁵ This interpretation is, by no means, self-evident from the principle itself. But, even if it were, it omits entirely the reasons why we should care. It substitutes an abstract principle for the real concerns at issue, and thus lacks entirely the emotional salience and resonance that is necessary to carry the day. 

As a scholar, my life’s work has been devoted, in large part, to correcting errors and pointing out common fallacies in the instrumental case for broader copyright, and to showing that the proper application of instrumental principles points to the need, today, for radically narrower and shorter copyright. Drassinower’s book is a direct assault on my life’s work. Perhaps it is not surprising then that I did not find the book particularly helpful to understanding copyright. 

Suggested Citation: 7(2) The IP Law Book Review 18 (2017).

© 2017 Glynn S. Lunney, Jr. 

---

1. Abraham Drassinower, WHAT’S WRONG WITH COPYING? (2015).
2. Id. at 18 (arguing that the economic approach to copyright seeks to maximize the value generated by works of authorship, conceived as information commodities, but then pointing out that copyright’s principal prerequisite for protection, originality, “is not about value per se”).
3. Drassinower mistakenly identifies nonexcludability as a necessary precondition for the existence of a public good. He further mistakenly asserts that copyright solves the market failure associated with public goods, at least for those public goods that also constitute works of authorship, by making works of authorship excludable.
4. Glynn S. Lunney, Jr., Lotus v. Borland: Copyright and Computer Programs, 70 TUL. L. REV. 2397, 2427-28 (1996) (“I would suggest that the differing standards turn almost entirely on the practical differences involved in copying the creativity embodied in an invention rather than a work. Specifically, the creativity embodied in an invention is usually more difficult to copy as a practical matter, than the creativity embodied in a work of authorship.”).
5. Graham v. John Deere Co., 383 U.S. 1, 11 (1966) (“The inherent problem was to develop some means of weeding out those inventions which would not be disclosed or devised but for the inducement of a patent.”); Glynn S. Lunney, Jr., E-Obviousness, 7 Mich. Tele. & Tech. L. Rev. 363, 408, 412 (2001) (attempting to develop a framework for implementing the inducement interpretation of the obviousness requirement); Michael Abramowicz & John F. Duffy, The Inducement Standard of Patentability, 120 Yale L.J. 1590 (2011) (same).
6. To keep the example simple, I am ignoring the regulatory process required for marketing pharmaceuticals.
7. Drassinower, supra note 1, at 25-26 (arguing that it is not clear whether granting or denying copyright to a collection of facts, such as the telephone white pages at issue in Feist Publications v. Rural Telephone, better promotes the progress of science).
8. Drassinower admits the strong instrumental arguments for noninfringement in the Google Books case, but insists that it is an isolated exception. Drassinower, supra note 1, at 225 (“The gigantic public weight of the Google Books Project is not the kind of force we can mobilize frequently in the name of lawful copying.”). I disagree that it is an isolated exception. Other examples of instrumentalism’s success include: (i) the use of thumbnail images for searches on the Internet, Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir. 2007); (ii) the home taping of copyrighted television broadcasts for later viewing, Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417 (1984); and (iii) the creation of electronic reserves for educational purposes, Cambridge University Press v. Patton, 769 F.3d 1232 (2014). These are all easy cases, from an instrumental perspective, but that is the point. These are the cases where getting the answer right makes a difference. If the costs and benefits of extending or denying copyright protection in a particular case are evenly balanced, then, while I would have a preference for denying protection, either outcome is acceptable.
9. Id. At 147 (“Instrumentalist discourse is, in my view, part and parcel of the very expansion that minimalism seeks to counter.”).
10. Id.
11. Id.
12. Paul Heald, How Copyright Keeps Works Disappeared, 11 J. Empirical Legal Studies 829 (2014) (showing the disparity in availability on Amazon.com for books first published before 1922 and for those still within copyright’s term).
13. In my view, this is one of the main problems with Drassinower’s approach: It does not respond to changes in the costs and economics of copying and distribution. Indeed, Drassinower critiques the economic approach for responding to changes in the economics. Drassinower, supra note 1, at 5 (“It follows, of course, that if we were to come up with an efficient way of demarcating ideas and licensing their use, we would then have to consider reevaluation – if not abandonment – of the idea/expression dichotomy as a doctrine fundamental to copyright.”). But, in my opinion, that the instrumental approach responds to changes in the costs of creating and distributing works of authorship is a virtue, not a vice. As for abandoning the idea-expression dichotomy, I believe Drassinower’s critique misunderstands the instrumental view of copyright. Idea and expression are not self-defining deontological principles. They are labels we attach based upon instrumental balancing. If, on balance, prohibiting the copying of Lotus’s user interface is undesirable, then it’s unprotected “idea.” If technology changes and the underlying cost-benefit analysis shifts, we may redefine whether some particular aspect of a work, whether as user interface or a particular chord progression, constitutes idea or expression. The legal categories, idea and expression, and their legal consequences would remain.
14. Given my circumstances, the connection to copyright’s traditional incentives story is somewhat indirect. I did not receive royalties directly from publication of the article. Rather, I was paid to write as part of my position as a law professor. Nevertheless, publication and the associated reputation obviously plays some role in determining whether I can work as a law professor, and if so, how much I am likely to make.
15. Drassinower, supra note 1, at 225.