Category Archives: Uncategorized
Jun 10, 2016 Dotan Oliar
Zahr Said’s Reforming Copyright Interpretation puts its finger on an important, yet little studied, aspect of copyright law: judicial interpretation. It pushes the ball quite a bit by providing a descriptive taxonomy of courts’ interpretive approaches in copyright law, advancing and defending an interpretive approach that it considers best overall, and applying and exemplifying its framework and arguments with a good number of cases while situating it all within a larger body of law and literature scholarship. For me, that’s tons of progress in one article, and the reason why I like it lots.
In resolving copyright disputes, judges must make interpretive decisions. Decisions regarding interpretation are often outcome-relevant – for example, when they lead a judge to decide whether an issue is a matter of law or fact or whether expert testimony may be admitted or not. These decisions can also be outcome-determinative – for example, when a judge makes an interpretive decision that resolves a case on summary judgment or finds an allegedly infringing use to be fair. The interpretive judgment that these decisions involve often flies under readers’ radars. Said draws our attention to judges’ interpretive choices and to the systemic effect that they have, or could have if they were to be conducted appropriately, on copyright law.
Descriptively, Said organizes judges’ interpretive choices along two axes. First, judges decide what to focus on in their interpretation. In doing so, they choose a point on a Text/Context spectrum. By analogy to patent law, this could be thought of as the proper mix of intrinsic and extrinsic evidence in claim construction. “Text” here stands for any interpretive decision drawn out of, or relying on, the copyrighted work itself—be it textual, pictorial, musical, etc. “Context” stands for interpretive decisions relying on evidence outside the work such as the author’s intent, expert opinion, or readers’ perception or response.
For example, a court may need to decide whether a work is substantially similar to another. It may make this determination by comparing the two texts. Or, it can rely on the author’s (or alleged infringer’s) intent, the lay reader’s view, community perceptions, expert opinion, or even the court’s own subjective impression of the works. Each choice vests the interpretive authority over the text in a different source. This decision can be outcome-determinative: a judge reviewing a defendant’s motion for summary judgment may grant it based on reviewing the text alone and conclude that no similarities exist. But if the judge’s interpretive approach gives weight to the author’s intentions or to audience perceptions, then she may deny the motion.
Second, judges choose an interpretive method along an Analysis/Intuition axis. The interpretive method helps the judge reach her legal conclusion from her chosen source of authority over the work. “Analysis” stands for careful explanation of the judge’s reasoning, applying doctrine to facts, a practice that may constrain a court’s future analysis or subject it to appellate review. An example of analysis is Mannion v. Coors Brewing Co. (SDNY 2005), where the court identified three potential loci of originality in photographs generally, and then moved to determine whether any of the three were present in the photograph before it. Alternatively, courts can assert conclusions about the works at issue that appear to be based on nothing more than intuition. An example of intuition is Roth Greeting Cards v. United Card Co. (9th Cir. 1970), where the court found infringement by noting that the defendant’s greeting card copied the plaintiff’s “total concept and feel.”
Underlying Said’s perspective is the notion that interpretation is inevitable. Courts engage in it whether they acknowledge it or not. Said wants judges to be aware of the interpretive phase of their decision-making process, and to choose and defend their points of choice along the two axes deliberately. Said suggests that it would be better if doctrine provided judges with an interpretive roadmap – perhaps like patent doctrine does – rather than assuming that judges already own the enterprise of formulating interpretive methodology. An agreed-upon methodology would enhance the enterprise’s predictability and fairness, and subject it to more effective appellate review.
Said laments judges’ tendency to view artistic, expressive, “nontechnical” copyrighted works as “self-interpreting” or “facially clear” and thus semiotically accessible and in no need of interpretation. This is different from the case of more technical works, such as software, where courts recognize the need for interpretation and for following an interpretive framework. Said resists this technical-artistic distinction for interpretation. Every text is complex, and meaning is not clear on the face of the work. Accessing works requires an act of interpretation.
Normatively, Said provides two prescriptions for a desirable approach to copyright interpretation. First, she suggests that judges lean toward the text side on the Text/Context continuum. Text is an easy-to-locate, objective source that both parties have access to. Focusing on text tends to minimize the amount of evidence presented in trial, and thus minimizes litigation costs and trial uncertainty. Second, Said prefers that judges use analysis over intuition because of current doctrinal confusion with the “total concept and feel” test, because different judges have different intuitions, and because intuition allows little ground for appellate review. Analysis, on the other hand, tends to create precedents, interpretive roadmaps, and case law consistency.
Said acknowledges that her prescription-favoring, text-based formalism may resolve most infringement cases, but that it may not suffice for cases turning on issues such as contested authorship or transfers of copyrights. Here, courts’ decision-making will often have to go outside the four corners of the work, such as by considering the parties’ intentions and community norms. Said acknowledges fair use as another context in which text-based formalism may prove inadequate, and where courts will have to look at contextual considerations, such as social meaning in finding a work transformative.
Said’s work is rich and intriguing. Paralleling its insight and direction, it also opens up new questions and possible extensions, and here I would like to mention two. First is a possible empirical extension. As for the descriptive portion of the article, it would be interesting to know what more could be said to portray a more accurate descriptive model of courts’ interpretive choices. Take the Text/Context axis, for example: what ideal points do courts tend to choose in deciding cases? If I had to guess, a uniform distribution would seem unlikely: while I can see courts choosing points ranging from the text end of the spectrum up to about the midpoint or even beyond, I would be surprised if many cases gave exclusive or predominant weight to context, and little or no weight to the text. The article’s descriptive insight could be augmented here with an empirical component. Further, one may wonder about whether courts’ choices of ideal points along the two axes in particular cases are independent of each other. Again, if I had to guess (and as the article recognizes in passing), I would think that there is some correlation. It seems that choosing a point involving a high degree of intuition and little analysis on that spectrum would correlate with a preference of text over context. Again, mapping these correlations would enhance the article’s descriptive power.
A second possible extension may be tying Said’s work on copyright interpretation to other bodies of IP. The framework might be extendable to patent and trademark laws, for example, perhaps with some adjustment. In patent law, the Text/Context axis seems readily applicable, as it can be used to model courts’ and commentators’ preferences as between intrinsic and extrinsic evidence in claim construction, for example. What about the Intuition/Analysis axis? How many patent law summary judgments could be described as founded on little more than intuition as to the meaning of patent claims? Can patent courts be characterized as distinguishing between “simple” and “complex” technologies, so that patent cases reflect (an implicit) judgment that the former require little interpretation of claims while leaving more room for interpretation in the latter category (paralleling copyright courts’ distinguishing between technical and nontechnical works)? Trademark infringement’s test calls for a mix of textual and contextual considerations, such that its choice and direction could be mapped using that axis as well. Perhaps the framework could be used across all IP fields, even if the normatively desirable “ideal point” may differ among them, and perhaps it is even of broader application in interpretation generally. These are for sure laborious and non-trivial questions that go well beyond the scope of Said’s article, which already does a lot. They are, however, a testament to the article’s utility and insight.
May 6, 2016 Lisa Larrimore Ouellette
Jake Linford,
Are Trademarks Ever Fanciful?, 105
Geo. L.J. (forthcoming), available at
SSRN.
Trademark law protects distinctive marks: ones that identify the source of goods or services and distinguish them from others in the marketplace. But how should courts determine whether consumers view a mark as distinctive? In an attempt to provide some analytical rigor to this essential question, courts have developed a complicated two-prong test: they look to both “inherent distinctiveness” (i.e., linguistic uniqueness) and “acquired distinctiveness” (i.e., whether consumers have come to see the mark as distinctive of source). Inherent distinctiveness for word marks is based on the so-called Abercrombie spectrum (named after the 1976 2d Cir. case that most famously articulated it), which classifies marks from most to least distinctive as fanciful (KODAK cameras), arbitrary (APPLE computers), suggestive (COPPERTONE suntan lotion), descriptive (AMERICAN airlines), or generic (“apple” for apples). Marks like AMERICAN can become strong, protectable marks only by developing “acquired distinctiveness”; marks like COPPERTONE and APPLE are presumed to be protectable at birth; and coined marks like KODAK are the strongest of all.
As Barton Beebe notes in his casebook, Abercrombie‘s “influence on U.S. and even foreign trademark law cannot be overstated.” But Abercrombie‘s foundation has been slowly crumbling. For example, in a 2009 study, Thomas Lee and colleagues found that consumers are far more influenced by how and where a mark is placed on a box than by where the mark falls on the Abercrombie spectrum. Rebecca Tushnet has explained that Abercrombie “lacks empirical foundation” and is out of step with basic marketing knowledge, such as that an ostensibly fanciful mark like VIAGRA is already imbued with “suggestions of virility, viability, and Niagara Falls (a classic sexual image).” And now, in an impressive trifecta of recent articles, Jake Linford has further dismantled the key theoretical assumptions underlying the Abercrombie classification scheme for word marks.
Linford has tackled Abercrombie from all sides: First, in A Linguistic Justification for ‘Generic’ Trademarks, Linford argued that categorically denying protection to “generic” marks ignores processes of language change through which once-generic terms can acquire source significance. He then moved to the middle of the spectrum with The False Dichotomy Between Suggestive and Descriptive Trademarks, which drew on “semantic shift” research to explain that suggestive and descriptive marks are likely to confuse consumers in similar ways. Linford argued that suggestive marks, like descriptive marks, should be protected only upon a showing of acquired distinctiveness rather than receiving automatic protection. Most recently, in Are Trademarks Ever Fanciful?, he attacks the other end of the spectrum, arguing that courts are wrong to assume that made-up words are empty vessels without source significance that should receive automatic, robust protection.
The key to Linford’s argument on fanciful marks is research on “sound symbolism,” or the strong connection between word sound and word meaning. For example, most people think that a table named MAL is larger than one named MIL, and that a product named TAKETE is spikier than one named MALUMA. Want your product to seem small and fast? Choose the vowel /i/ over /u/, voiceless consonants like /t/ over voiced ones like /d/, fricatives like /s/ over vocal stops like /k/. One recent review of this literature concluded that “the sounds of words can convey meaning apart from their actual definitions, and this meaning can systematically bias perceptions and judgments.”
Marketing research has demonstrated that sound symbolism can be used to choose more effective trademarks. Consumers prefer fictitious brand names that match perceived attributes of a product, such as GODAN over GIDAN for a dark beer, NELLAR over NULLAR for a fast Internet service, KUTUM over KITUM for male deodorant, GOMMEL over GIMMEL for SUVs (but vice versa for convertibles). Marketers use this research when coining fanciful trademarks. By cataloging the sound symbolism literature, Linford makes a strong case that fanciful trademarks are not empty vessels: rather, they “trigger powerful associations in the minds of consumers, which can be used to transmit both source- and product-related information.”
What are the implications of sound symbolism for trademark law? As Linford explains, this research undermines the justifications courts have offered for granting strong protection to fanciful marks, and it can impose unjustified costs on competitors. “Courts often find similarity between a fanciful mark and a competitor’s mark to indicate bad faith adoption by the competitor,” but Linford argues that “sound symbolism provides a reason for the competitor to adopt a mark with some similar characteristics: to communicate product features to consumers.”
At the very least, in light of Linford’s work, courts should discount the use of similar sound symbols by alleged trademark infringers rather than presuming bad faith. Linford also tentatively proposes “[t]wo more radical shifts”: requiring every mark to show acquired distinctiveness in order to receive protection, and holding that some sound symbols are “essentially functional” and thus not protectable. He acknowledges the concern that weaker protection for fanciful marks would lead to more firms choosing descriptive marks, which might ultimately impose even greater costs on competitors, but he argues that “sound symbolism is a phenomenon with real power” such that there may be little competitive difference. And while abandoning Abercrombie would be a big change in practice, there are few trademark scholars who strongly defend this doctrine. Rather than advocating artificial linguistic distinctions, scholars have increasingly argued that trademark law should be more attuned to how consumers actually view different types of marks. (An excellent recent example is Alexandra Roberts’s explanation of why Abercrombie does not reflect consumer perceptions of whether hashtags like #IceBucketChallenge are source-indicating trademarks.)
Perhaps a bigger concern with Linford’s sustained attack on Abercrombie is that it would increase administrative costs. Under current law, choosing an “inherently distinctive” suggestive, arbitrary, or fanciful mark is a shortcut around the costs of developing evidence of acquired distinctiveness. And the categorical bar on generic marks makes it easy for courts to explain why certain marks are undeserving of protection no matter how much consumers have come to view them as source-indicating. But the lower administrative cost of a bright-line rule over a more flexible standard is of little comfort when the error costs associated with the rule outweigh that administrative cost savings, and Linford has provided numerous reasons to suspect high error costs at each point of the Abercrombie spectrum.
Would abandoning Abercrombie and just looking for evidence of acquired distinctiveness actually lead to lower error costs? It would at least focus courts on the right question, but trademark cases rarely involve reliable evidence for assessing how consumers actually view a mark. Consumer surveys are expensive and unreliable, so courts often turn to circumstantial evidence such as advertising spending on the asserted mark, but such evidence is only a very weak proxy for how distinctive the mark actually has become to consumers.
What courts really need is a simpler way to tell whether consumers associate a mark with a certain product. In The Google Shortcut to Trademark Law, I argued that one cost-effective way to do this is to look at online search results, which allow us to collapse the two prongs of the distinctiveness test: if a word mark is distinctive (i.e., if consumers tend to primarily associate it with a certain product), then most of the top search results will use that mark to refer to that product. I don’t know whether this test will gain traction, or whether better consumer surveys will be developed, or whether creative scholars (or judges) will come up with an entirely new way to assess distinctiveness. But Linford does a terrific job explaining why it is a mistake to stick with the status quo.
Apr 12, 2016 Christopher J. Buccafusco
Andres Sawicki,
Risky IP , Univ. of Miami Legal Research Paper No, 16-18 (2016), available at
SSRN.
Intellectual property laws govern activities that are inherently risky. Authors and inventors can only estimate the consumer demand for their contributions. And many creative activities run the risk of infringing existing IP rights. Accordingly, it is essential for policymakers and scholars to understand how creators think about risk.
To date, most people who have written about IP law and risk have assumed that creators will be risk averse. In a new paper, Andres Sawicki challenges these accounts and argues that the kinds of people that IP law typically regulates—creative people—tend to be risk seeking. Accordingly, where others saw the risk inherent in IP as a problem, Sawicki sees it as potentially beneficial.
Sawicki begins by discussing the various features of IP that involve risk, including the fundamental riskiness of all market-based creative ventures as well as IP doctrines such as patent law’s definiteness requirement and copyright law’s fair use standard that introduce additional uncertainty. When previous scholars have noted these issues, they have tended to assume that creators faced with risk will tend to be risk averse, because, after all, most people are risk averse. Thus, most people prefer receiving $100 to playing a lottery in which they have equal chances of receiving $200 or nothing.
Sawicki reviews a large and growing body of literature from psychologists who study creativity that suggests that, at least for creative people, this account is not entirely true. Research suggests that more creative people tend to have a greater tolerance for risk than do less creative people and that a fundamental feature of successful creativity is the ability to recognize and capitalize on risks. Sawicki is careful to note that the literature in this area is not entirely conclusive, but he makes a plausible case that successful creators are likely to view risk and uncertainty differently from other folks.
Based on this claim and a set of important simplifying assumptions, Sawicki then offers a number of testable hypotheses about how creators’ risk preferences will influence their behavior. He suggests that creators will tend to seek out opportunities to engage in activities with uncertain benefits over otherwise equivalent ones with certain benefits. And he suggests that risky environments are more conducive to creativity.
If these claims are correct, Sawicki argues, creators’ higher risk tolerance could make IP-style incentives a more efficient proposition than previously recognized. Relative to a stable salary, prize, or tax incentive of equivalent value, the probabilistic and uncertain value of IP rights might ultimately generate higher creative output. In addition, Sawicki argues that efforts to make IP rights more certain may undermine some of the value that we get from risk-seeking creators.
As always, and as Sawicki recognizes, the devil will be in the details. Sawicki’s claims rest on various simplifying assumptions that may turn out to be wrong (a prospect he openly admits). For example, he assumes that a person’s risk preferences in one domain (such as artistic creativity) will be similar to those in another domain (such as financial risk). And he tends to assume that the framing effects noted by Daniel Kahneman and Amos Tversky may have minimal effects on creators’ decisions. Whether or not these assumptions are correct is a matter for future empirical research, and Sawicki’s paper has done a terrific job of pointing out the most promising directions for that work.
If Sawicki’s descriptive account of creator behavior is correct, IP law may be able to take advantage of creators’ risk tolerance to generate more incentive bang for its buck. But creativity and risk taking are not always valuable. When creators enter a field they face a variety of existing IP rights that may impinge on their goals. They can choose to license those rights or adopt the more risky strategy of designing around them. To the extent that downstream creators are excessively risk seeking, then, they will tend to choose the latter option at inefficient rates. Instead of licensing existing solutions, they will waste R&D resources in risky ventures. Ultimately, however, IP scholarship needs an accurate account of how creators are likely to respond to these and other risks.
Mar 17, 2016 Michael W. Carroll
Eric Priest,
Acupressure: The Emerging Role of Market Ordering in Global Copyright Enforcement, 68 SMU L. Rev. 169 (2015), available at
SSRN.
Corporate copyright owners based in the United States have been frustrated by the prevalence of piracy in China and in certain other fast-growing markets, and that frustration has led to three primary responses. The copyright industries have (1) supported proposed legislation that would impose enforcement obligations on U.S. parties, such as the Stop Online Piracy Act; (2) advanced expansive interpretions of the enforcement jurisdiction of the International Trade Commission; and (3) deployed technological protection measures.
In his new article, Acupressure: The Emerging Role of Market Ordering in Global Copyright Enforcement, Professor Priest identifies two additional strategies that seem to have promise. These strategies rely on pressuring certain intermediaries that hold the power to deny infringers access to the markets they seek to serve. Presenting these as case studies, he then abstracts away to model how and when market-based pressure on intermediaries or customers – the “Acupressure” in the title – are likely to be effective. He concludes by revisiting familiar critiques of copyright enforcement through private ordering and integrates these into his analysis of the public policy ramifications of these new developments.
After explaining why attempts to police foreign conduct in U.S. courts has largely failed, Professor Priest introduces the first case involving the online video market in China. In a very short period of time, two of China’s largest sources of online videos changed their advertising-based business model from one that relied on infringement to one that relies on licenses from copyright owners. Why incur these substantial costs when these service providers enjoy certain limits on liability that closely resemble those that U.S. intermediaries enjoy and when the threat of judicial enforcement for transgressions is negligible?
According to Professor Priest, these services’ most valuable advertisers – multinational corporations – faced internal and external pressures to dissociate their brands from platforms identified as pirate sites. He supports this assertion with some original reporting based on interviews with executives at the two companies. He emphasizes that after these two companies became licensees, their incentives aligned with the advertisers and the copyright owners, which led them to use legal process to pressure their unlicensed competitors, such as Baidu. He also persuasively explains why alternative theories of causation are less plausible.
His second case involves use of theories of unfair competition in the United States to block market access to foreign manufacturers that use infringing software in their operations. The theory is that companies that rely on copyright infringement for a critical input into their business operations enjoy an unfair advantage that allows them to artificially lower prices. The theory is accepted in “anti-IT theft” statutes in force in certain U.S. states and in case law that deems this an unfair trade practice under a state’s mini-FTC Act. Professor Priest cites examples in which foreign manufacturers agreed to license their software in response to threatened loss of market access.
From these examples, Professor Priest derives models of alternative enforcement. Most prominent is the “market pressure” model applicable “[i]n jurisdictions in which the state’s influence over infringers is weak, the state may be replaced in the enforcement chain by a powerful ‘third party’ with greater influence over the infringer.” The essential elements are that a third party with sufficient leverage exists, that the enforcement targets belong to a relatively small, tight-knit group to form compliance norms that can be internally monitored and enforced, that enforcement is ongoing either through the third party or other community members, and that the complying members are incentivized to invest in enforcement against defectors.
This model better conforms to the Chinese video market case study than to the use of unfair competition in the United States because the users of infringing software are a large and diffuse group. Professor Priest highlights Washington state’s law as having an innovative feature because it targets retailers in the United States who sell products derived from uses of infringing software, analogous to regulations prohibiting the sale of products produced with child labor or in violation of certain environmental norms. By shifting the target, this strategy pressures the retailer, such as Walmart, to exert copyright compliance pressure throughout its supply chain. But, query whether this law would survive a challenge under the Dormant Commerce Clause.
Professor Priest is more sanguine than I am about the desirability of these new enforcement strategies from a public policy perspective, but his article does a great service by bringing these to light, by engaging with concerns about non-judicial enforcement of copyright norms, and by analyzing the key ingredients that contribute to these strategies’ relative success.
Feb 15, 2016 Pamela Samuelson
Sarah Burstein,
The Patented Design, 83
Tenn. L. Rev. ___ (forthcoming 2016), available at
SSRN.
Ornamental designs of articles of manufacture have been patentable subject matter in the U.S. since 1842. About 400,000 such patents have issued in the years since the birth of this regime, two-thirds of which have been granted since 2000. Scholarly interest in design patents has historically been quite modest, but has been heating up lately. This is due in no small part to the epic battle between Apple and Samsung over Apple’s claim that Samsung’s phones infringed some of Apple’s design patents. Samsung has asked the Supreme Court to consider whether the designs at issue are really “ornamental” and thus properly covered by design patents. In addition, Samsung wants the Court to review the award to Apple of its total profits on the sales of the infringing phones in the amount of $399 million.
The Supreme Court has not reviewed a design patent law since 1894. The Court’s 1871 decision in Gorham v. White articulated a test for infringement that is still influential today. Gorham did not raise difficult issues of patent scope because the defendant in that had embodied a clearly ornamental patented design for silverware in directly competing products.
The question about ornamentality in Apple v. Samsung is important. One can only hope the Supreme Court will take the case and answer this question. Professor Burstein’s new article poses two other questions about design patent scope. One is whether a design patent can be infringed if the design is embodied in a different product than that depicted in the patent. The other is whether a design patent can be infringed by a visual representation of the design not embodied in an article of manufacture. She concludes that other-product-embodiments and visual-representations should not infringe design patents, both as a matter of statutory interpretation and as a matter of policy.
Burstein convinced me on both points. What I loved best about the article, though, was the rich discussion about what the word “design” means in the context of design patents. Burstein observes that the word design is “mercurial,” serving both as a verb and as a noun, a process as well as a product. Design has a dual meaning: the look of products and the preparation of instructions for the production of those products. Designs for articles of manufacture are about blending form and function. When articles of manufacture are beautiful, that beauty results from this blending. The product and its design are inseparable. These insights about design inform Burstein’s interpretation about the proper scope of design patents.
Burstein discusses some cases that recently posed other-product-embodiment and visual representation issues. The cases are so colorful that one might have imagined them to be improbable professorial hypotheticals.
In Kellman, the plaintiff owned a design patent on a novelty foam hat in the shape of a wing-nut so that fans of the Detroit Red Wings team could express how nuts they were about their favorite team. The wing-nut visual pun caught on. One firm printed a wing-nut design on t-shirt, and Coca-Cola impressed a wing-nut design on bottle caps. Kellman sued both entitites. Under the Gorham test for design patent infringement, neither claim was plausible.
Gorham directs the trier of fact to consider whether an ordinary observer would the designs at issue to be so substantially the same as to induce customers to purchase the defendant’s product supposing it to be the plaintiff’s. No one who wanted a wing-nut hat would be deceived into buying Coca-Cola because of the wing-nut on the bottlecap. The t-shirt claim was arguably stronger because both products are items that humans wear. Yet, it was implausible that anyone who wanted the hat would be deceived into buying the t-shirt instead.
The PS Products case was even stranger. The design patent at issue was for a stun-gun designed to be worn like brass knuckles. Its owner sued the maker of a videogame that included visual depictions of similar weapons. The District Court dismissed the case on the ground that no reasonable person would purchase the videogame believing he/she had purchased a stun-gun.
These decisions reject the view that design patents protect designs per se. Burstein argues that these cases were rightly decided and offers statutory, doctrinal, and policy reasons in support of her argument. What convinced me most, however, was her discussion earlier in the article about the inseparability of patented designs and their embodiments in articles of manufacture. If a designer thinks that a design can, in fact, be embodied in more than one kind of product, he/she should apply for more than one design patent. For it is a requirement that an applicant identify what kind of product is to embody the design.
Burstein would not, however, limit the scope of design patents to only those products specifically identified in the patent. She would favor a rule that extended design patent protection to products in the same general category. This part of the paper is less developed, but is consistent with her overall thesis.
Design patents are not copyrights; they do and should not protect designs per se. Having spent a great deal of time cogitating about the proper scope of copyright’s derivative work right, I appreciated Burstein’s project to do a comparable analysis of design patent scope. Bravo, Sarah.
Jan 15, 2016 Christopher J. Sprigman
Michela Giorcelli & Petra Moser,
Copyright and Creativity – Evidence from Italian Opera (2015), available at
SSRN.
Today opera fans in the United States are rich, old, and increasingly rare. But it wasn’t always that way. In the Eighteenth Century, opera was the closest thing to mass entertainment, especially in Italy. And that fact provides a platform for economists Michela Giorcelli and Petra Moser to say something interesting about the effect of copyright law on creativity. Giorcelli and Moser’s Copyright and Creativity – Evidence from Italian Operas, is a paper I liked, lots.
Giorcelli and Moser’s paper is a natural experiment using historical data surrounding an “external shock” – viz., Napoleon’s invasion and occupation of northern Italy between 1796 and 1802. The northern Italian states of Lombardy and Venetia adopted copyright laws in 1801, as a direct consequence of French rule. Six other Italian states studied by Giorcelli and Moser only began adopting copyright laws during a period that began a quarter-century later. Giorcelli and Moser collect historical data on 2,598 operas that premiered across the eight Italian states in question between 1770 and 1900, the most fertile years of Italian opera production, and a period that both precedes and follows the adoption of copyright by Lombardy and Venetia.
Comparisons across the period reveal a statistically significant increase in new operas produced in the states that adopted copyright in 1801. Giorcelli and Moser estimate that Lombardy and Venetia produced an average of 2.12 additional operas per year after 1801. This increase is relative to a baseline of 1.41 operas per state per year before 1801, thus yielding an apparent increase of approximately 150%, versus an increase in production of approximately 54% in the states that had not adopted copyright.
The authors then inquire whether the increase in number was accompanied by an increase in overall quality of the operas produced. Using other historical data sets, Giorcelli and Moser estimate a 4.6-fold increase in the production of historically popular operas in response to the adoption of copyright, and a ten-fold increase in the production of durably popular operas (i.e., those for which full-length recordings continue to be available on Amazon in 2014).
The data also shows that after Lombardy and Venetia adopted copyright, opera composers began to immigrate to those states. “Between 1801 and 1821, 43 composers who were born outside of Lombardy premiered an opera in that state. Another 13 composers born outside of Venetia premiered an opera in Venetia,” the study states. “By comparison, all other Italian states together only saw premieres by 5 composers who were born outside the state premiered their first opera in other Italian states without copyrights.”
In sum, assuming that the production of operas is a reasonable proxy for artistic and literary production generally – and I’m not so sure it is, but more on that later – the Giorcelli and Moser study suggests that the adoption of some reasonable copyright term provides a incentive that produces more creative output versus an environment in which there is no copyright protection and creative output is consequently more vulnerable to appropriation.
However, and importantly, the study also suggests that subsequent extension of the copyright terms from life of the author plus 10 years to life plus 40 had no clear effect on either the number or quality of operas produced. This second conclusion is particularly important because our contemporary debate is usually not whether to have copyright at all, but rather whether in extend already very long copyright terms. On that question, Giorcelli and Moser provide evidence that a bit of copyright is enough, and more copyright doesn’t necessarily lead to more creative production.
Of course, longer copyright terms do have a cost. They perpetuate monopolies, with all the social costs that monopolies typically produce, plus extra costs that arise when the spread of knowledge goods like books or even operas is restricted by monopoly’s high costs and scarcity. Cheap books help spread literacy. Cheap operas help spread cultural literacy. Both forms of learning enrich society. Copyright terms that are too long limit the spread of knowledge while producing little, if any, additional creative output.
I love the Giorcelli/Moser paper, but like any piece of empirical work, it raises methodological and data-quality questions. I want to focus on the latter for a moment.
The paper’s conclusions are only as strong as the data on which the researchers rely, and I would have liked to have seen more discussion of why we should trust the completeness of the records listing opera premiers during the period. Giorcelli and Moser consult multiple sources, but these are historical records from a often chaotic time and place in which it is not intuitively obvious that all premiers would have been recorded. If the rate at which the premiers failed to be recorded varied across the Italian states, that variance could cause a significant perturbation in the data, with corresponding effects on the results.
To be fair, Giorcelli and Moser use more modern sources to account for omissions from the historical records, but it is difficult to judge from the paper the confidence one reasonably can place in the amended records.
Let me close with a word on a methodological issue, which is, as I alluded to earlier, whether data suggesting that the adoption of copyright boosted the production of operas is representative of copyright’s likely effect on the production of other, more modern and commercially important, forms of creativity. My guess is that the answer depends on how closely the economics of a particular form of creativity resembles the economics of opera production. I would characterize opera as a form of creativity with very high fixed costs – authoring an opera represents a large and sustained creative effort, and the production of an opera is also costly (hiring singers, musicians, and opera halls is expensive, as are the often-lavish sets and costumes). Opera is, in short, a paradigm of the sort of creativity that is unlikely to flourish without some way to prevent copying, because the originator needs an extended period of exclusivity in order to earn back those high fixed costs.
Like opera, some modern forms of creativity feature very high fixed costs. For example, blockbuster movies. But other commercially important forms of creativity – for example, pop music – are produced with relatively low fixed costs. It’s cheap these days to write and record pop music. Is the sort of lengthy monopoly created by copyright law necessary to stimulate the production of this lower-cost creativity? On that question, the jury is still out.
