copyright main image

Hope v Reality

stan liebowitz is the Ashbel Smith professor of managerial economics at the Jindal School of Management at the University of Texas at Dallas.

Illustrations by Roman Muradov

Published January 21, 2014.


The basis for American copyright laws is laid out in the Constitution: "To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."

Happily, the founders' reasoning conforms nicely to the economic analysis of the underlying issue. But it has been a matter of debate whether the current law conforms to this constitutional instruction – or more generally, whether the law creates economically efficient incentives to create and distribute the property in question. In fact, the contemporary debate over how best to reconcile the some times conflicting goals of copyright regulation is really a modern rendition of a centuries-old argument. Here's a primer on how to think about the question.

What Does Copyright Protect?

Copyright is intended to protect the expression of ideas, not the ideas themselves. It does this by allowing the owner to determine who is allowed to make copies of the work being protected. Copyright was originally applied to books. But as technology changed, protection was broadened to cover creative works ranging from recorded music to film to visual art. There are often layers of rights – in a movie, for example, the screenplay, the book or magazine article it's based on and background music may be separately protected. Untangling these layers of rights provides employment for an army of lawyers.

Unlike patents, copyright is supposed to protect an author only from someone avoiding the costs of creation by copying the author's work. Copyright does not bar independent creation of similar or even identical work.

To illustrate this narrow range of protection, consider the copyright that allows J.K. Rowling to prevent her stories from being printed, translated or turned into movies without her permission. It does not allow her to bar others from publishing fiction about, for example, boarding schools that teach magic, even if they are piggybacking on the same customers.

In principle, then, it should be fairly easy to determine whether a work infringes on others' copyrights. If you publish a 200-page story that is word-for-word identical to mine, that's surely an infringement because there is an essentially zero probability that you could independently create an identical work. Indeed, it is virtually impossible to create even a single identical paragraph without copying. On the other hand, if you create a very similar story, but devise your own plot and characters and write each paragraph yourself, you're not infringing.

That said, there are some hard cases to judge. The main tune in a song may have less variety than that found in a typical paragraph. Two almost identical tunes might be independently created, and copyright infringement cases for songs often revolve around the issue of whether the writer of the second song heard the first song.

Parodies of works are allowed under copyright, although there can be a fine line between a parody and a copy (or derivative work). A cause célèbre among critics of copyright was a book published a few years ago called The Wind Done Gone. It borrowed the characters from Gone With the Wind, but retold the story from the perspective of the slaves. Margaret Mitchell's estate accused the author of infringement, but the courts eventually ruled that the newer work was a parody, not a copyright infringement. The publicity over the court case, it's worth noting, was probably good for sales of both books.

Fan fiction, where fans of, say, Star Trek create works using characters from the original but provide original plot lines, is usually considered a derivative work. Derivative works are often considered infringements of copyright, although many copyright owners seem willing to overlook the transgressions because they believe the derivative work serves their own interests by keeping the fans happy.

In light of the narrow nature of the legal protection, I believe it is a mistake to say that copyright provides a monopoly, as the term is used in economics. The "monopoly" it provides is no different than the monopoly everybody has on his or her own talents, or the monopoly that any firm has on its own branded products.

copyright image 01

A Capsule History

The U.S. Congress passed copyright legislation just three years after the adoption of the Constitution. The initial term was 14 years, followed by another 14 years if the copyright was renewed. But copyright was only for American authors. For most of the 19th century, the United States was something of a rogue in this regard, failing to extend protection to foreigners. (Ironically, this failure gave foreign books printed in the United States a leg up in the American market, since without the need to compensate the author they were less costly to publish.)

Foreign authors could close the loophole by finding an American "coauthor" to write a preface, thereby protecting the entire book. But sharing the credit was more than some authors could stomach: Rudyard Kipling (among others) refused to make this sort of arrangement, while T. H. Huxley acceded.

Canada applied tit-for-tat, denying copyrights to Americans. Mark Twain beat the Canadian system by spending several weeks in Montreal before publication of The Prince and the Pauper, thereby gaining temporary Canadian residency.

Arnold Plant, an early 20th-century economist, examined the state of United States publishing during the late 19th century, concluding that some British authors made more money in the United States without copyright than they did in Britain with copyright. His argument: being first in the market provided a sufficient time cushion to allow most of the benefits of copyright to go the creator of unprotected work. But his case was weaker than it looked since the American market was already considerably larger than the British market, implying that parity in sales was still indicative of weaker protection.

Early 20th-century U.S. copyright law provided 28 years of protection, followed by another 28 on renewal. This has gradually been extended. In 1976, the term was increased to 50 years after the death of the author, and in 1998 it was increased to 70 years after the death of the author. (Works created under contract – "works for hire" – have different copyright lengths since there is no "life" upon which to base the duration.) The 1998 extension, by the way, triggered an academic and legal backlash that I'll detail later.

The scope of copyright has kept up, imperfectly, with new technologies. The owner of a copyrighted sound recording of a musical composition is paid each time the recording is played on the radio. But owners of copyright on the performance of the song on those same sound recordings are not paid when the record is played on the radio (United States law is unusual in this regard). Thus, if the radio plays a recording of Rod Stewart singing a Gershwin song, Gershwin is paid but Stewart is not. If the same recording is streamed over the Internet, however, both the owner of the composition (Gershwin) and the owner of the sound recording (Stewart) are paid.

There is no logical distinction between these cases. In large part they reflect the varying political clout of the stakeholders when the rules were set. And that's no way to run a copyright system.

Limits on Copyright Protection

The most important limitation to copyright protection is a concept known as "fair use," which provides defenses to infringement claims. Originally a common law concept, it was codified in the 1976 copyright law. Simplifying a bit, fair use applies if:

  • Only a small amount of the work is copied;
  • The use is educational and/or nonprofit;
  • The work is nonfiction;
  • The market for the original work is hardly damaged by the copying.

None of these rules are hard and fast. For example, in the Betamax case (Sony v. Universal City Studios, 1984), the Supreme Court held that making a videotape of a broadcast television program was fair use. The first three conditions hardly applied. But because these programs were supported by advertising and the advertising was contained in the videotape copy, it was unlikely that the market for broadcasting would be damaged, implying that the last condition was likely to be met.

Fast-forward 30 years, and assumptions of no financial damage seem much weaker since modern recorders can strip out commercials. Currently, Dish TV, which makes no secret of the fact that its Hopper DVR can remove commercials, is being sued for copyright infringement. Nevertheless, the Betamax fair-use precedent will be difficult to overturn.

Another restriction on copyright is the compulsory license. Such a license compels a copyright holder to make the work available to all users as long as the user meets certain conditions (for example, pays a stated fee).

Have you ever wondered why Pandora, the Internet streaming service, does not allow you to pick an individual song, but only a performer, composer or genre? This limitation is due to a compulsory license that Pandora wants to use. When Internet streams are controlled by the listener, the stream is categorized as "interactive" and the copyright rates (for sound recordings) are set by private negotiations between the parties. When Internet streams are classified as non-interactive or semi-interactive, the streaming entity can pay the compulsory license copyright fee (for the sound recordings), which is currently considerably less than the market payments for interactive streams. Pandora cannot let users cross a line on the control of streaming content or it would be classified as an "interactive" service and oblige the company to pay the higher royalty.

An Economic Analysis of Copyright

Start with a light dusting of jargon. When a typical good is consumed – say, an apple or a box of tissues – it becomes unavailable to others. But with creative works like books, consumption may be "non-rivalrous." That is, my reading of A Tale of Two Cities does not preclude anyone else from reading it. Nothing is used up when you "consume" the work; while a physical book may become more worn, the story it contains is not affected.

Another economic characteristic labeled "non-excludability" is often (though incorrectly) linked to non-rivalrous consumption. There are some items, like national defense, for which it is essentially impossible to exclude individuals living in the country from benefiting from the safety it provides. But intrinsically non-excludable items are far and few between.

It was claimed by no less an authority than Paul Samuelson, a towering figure in the history of economic thought, that television broadcasts were non-excludable. This was true for a while in the United States, but only because of the way over-the-air broadcasting was regulated. It was not an economic characteristic of television broadcasts per se. For example, television broadcasts were excludable in Britain for several decades after World War II. Specialized vehicles with electronic sensors roamed the streets to catch unauthorized television viewing by households that did not pay a user fee to the BBC.

In general, with sufficient resources brought to bear, unauthorized consumption of intellectual products can largely be prevented, just as theft and shoplifting can largely be prevented. Thus, non-excludability is normally a function of legal protection and the failure to invest in protective systems (like signal scrambling for cable TV) rather than a fundamental characteristic of copyrighted products.

But back to non-rivalrous consumption, which is a characteristic of copyrighted goods. To achieve economic efficiency – maximizing the net value to society – once the book, song, play or movie is created, everyone who values it more than the cost of reproduction should be allowed to consume it. By this logic, in a world in which the reproduction costs are close to zero, as is largely the case with digital goods distributed over the Internet, all potential users should be allowed to consume it.

But if there were no charge to consume the product, where would the money come from to pay for its creation in the first place? This conflict between the efficiency of a low price to encourage consumption and a higher price to provide the incentive to create is known as the efficiency tradeoff in the production of non-rivalrous goods like copyrighted and patented goods.

One theoretical solution – "perfect" price discrimination – is to charge each consumer just slightly less than the maximum he is willing to pay. There are a couple of problems, though. First, all of the surplus from the market exchange would go to the seller rather than the buyers, which may seem unfair if you are a consumer. Arguably more important, sellers would need to be mind readers to set prices perfectly. Inevitably, sellers would err, overestimating the value to some potential consumers and driving them from the market.

Because a price of zero provides no revenues to producers and because perfect price discrimination is only a theoretical fantasy, neither approach to maximizing value is practical. A third approach – really another fantasy – would be to have the government pay to create copyrighted works. Were the government to pay creators just enough to persuade them to produce works that consumers would collectively value more than the cost, and then give away the creation, the inefficiency would disappear. But there are problems.

First, it would take a crystal ball to know which efforts should be financed. Second, the funds to pay creators would have to come from somewhere, and taxes create their own inefficiencies. So government funding would trade one type of inefficiency for another. Third, it seems dangerous to put the government in charge of creativity since censorship and bias are likely to arise in its choices.

Most of the time, then, we use a rough-and-ready market-based approach, allowing the copyright owner to pick the price. That gives creators an incentive to create works for which revenues (or personal satisfaction) appear likely to cover the cost of creation. Consumers, for their part, get to consume the product if it's worth it to them at the price. Eventually, the copyright expires and anyone can consume the product for the cost of reproduction.

That still leaves the duration of copyrights to be decided – a decision that evokes the now-familiar tension between incentives for creation and benefits of low prices to consumers. Long-duration copyrights favor creative activities; shorter durations favor consumers.

Implicit in the debate over copyright duration is the assumption that one duration fits all. Yet that seems implausible on its face. An ideal copyright law would provide just enough protection to induce production – a point understood for a long time. In 1841, the historian-politician Thomas Macaulay argued:

It is good that authors should be remunerated; and the least exceptionable way of remunerating them is by a monopoly. Yet monopoly is an 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.

I have argued that an "efficient" copyright system along Macaulay's line would be unfair because it would funnel all the net value to consumers. It's unclear, after all, why we should squeeze creative occupations protected by copyright in this way, but allow other markets to direct much of their surplus to other categories of producers – for example, to athletes, movie stars and entrepreneurs.

The Problem With Efficiency

Our market-based economy does not always – or, for that matter, very often – aim at maximizing economic efficiency over other societal goals. Yet that's the perspective of economic analyses of copyright policy. And it can lead to some uncomfortable implications.

For example, copyright critics following in Arnold Plant's footsteps argue that some creators do not need copyrights to induce them to create. After all, there are all sorts of motives for creation other than money – for example, fame, not to mention the delight in creation. For that matter, some artists may do their best work while starving.

Let's take that last point seriously. What if artists do better work when poor than when rich? What would an efficient copyright system look like under these circumstances? Economics provides a simple answer. In this hypothetical world, giving artists money would be like polluting the air – a negative externality borne by society rather than the buyers and sellers. The solution: tax artists to keep them poor, just as we tax air pollution to make polluters change their behavior.

Let's go one step further. What if making slaves out of artists increased their production even more? What if it were unambiguously clear that their enslavement would generate more gains to others than losses to the artists? That's efficiency – and, plainly, it's not what we would want.

Skeptical of extreme hypotheticals as debating points? Take the very real case of the British-based multinational ARM Holdings, which designs the microprocessors found in most smartphones and tablets. ARM does not produce the chips; its output consists solely of their designs, which are perfect examples of non-rivalrous goods. What would an efficiency-focused analysis have to say about the optimal copyright length for this company?

The analogous logic to determine optimal copyright length would say that ARM should be allowed to generate all the revenue that allows it to cover costs, and not a penny more. Of course, we don't treat these firms this way. Is it really fair to treat copyright industries as sui generis cases where economic efficiency is paramount?

The Never-Ending Debate

Eldred v. Ashcroft (2002), a challenge to the 1998 copyright extension, worked its way to the Supreme Court. Critics of the extension to 70 years after the author's death claimed that it eviscerated the constitutional "limited term" requirement. Certainly, 120 years of protection for an author who writes a work 50 years before death is a long time. But the number is still finite and the term is still of "limited" duration. It's thus surprising that the case made it to the Supreme Court and that the challenge was supported by even two of the justices.

What the case had going for it were groups of influential legal academics and economists, which submitted "friend of the court" briefs. The economists, among them several Nobel Prize winners, argued that the extended duration made copyright too long. Few new works, they concluded, would be brought forth by the extension because the revenues generated by the extra 20 years of copyright life would be too small to increase current output.

They're right that the "present value" of revenues discounted so far into the future would lead to small increases in expected revenues. But the assertion that a small increase in revenues would have only a small impact on current production might or might not be true; it depends on how elastic the supply of new works is with respect to increases in expected revenues.

Moreover, the group failed to acknowledge that, by the same reasoning, the present value of the benefits of increased future consumption if copyright expired in 90 years, say, instead of 110, was also small. In truth, there is not remotely enough information for anyone to claim to know the ideal copyright duration.

Further muddying the waters was a secondary issue that offered a stronger case to critics of the law: the increase in copyright length was made retroactive to works that had already been created. Since the overriding economic purpose of greater copyright duration is to increase the incentives for creativity, the economic argument for increasing duration for works that have already been created simply does not exist.

Of course, the big copyright debate in the last decade has been about the impact of piracy on the sales of copyrighted goods – especially in the context of digital downloads. Almost every claim in this debate is in dispute, with much of the dialogue amounting to nothing short of pure disinformation.

With almost a decade and a half of experience, however, we can now say that the music industry, both domestic and abroad, has experienced an unprecedented decline in revenues – more than 60 percent after inflation. Moreover, most studies conclude that piracy has led to most of the decline.

By the same token, sales of recorded movies have fallen considerably in the last seven years (perhaps not coincidentally, after BitTorrent made movie piracy practical). Every rigorous study of movie piracy that I have seen has concluded that piracy had an impact, but the magnitude of the hit to the movie industry has not been pinned down.

The dispute about what, if anything, should be done about piracy has been particularly nasty in part because the home-electronics industry benefits indirectly from piracy.

It shouldn't be surprising that self-interest drives the debate. But part of the opposition to antipiracy regulation comes from people who seem to view copyright as little more than extortion. One of the things that we learned from Hollywood's recent confrontation with the digital hardware and software interests is that Google and

Wikipedia can scare the living daylights out of politicians if they get their users riled up.

Stay tuned.