Are Negative Spaces Likely to be Fragile

Back in what now seems the distant past-2006-Kal Raustiala and I wrote on innovation in the fashion industry. 2 In that article, The Piracy Paradox: Innovation and Intellectual Property in Fashion Design, Kal and I described how fashion design firms in the United States had long been and remained highly innovative despite having virtually no copyright protection for their designs. We argued that in fact the relative freedom to copy in that industry did not impede innovation, but instead helped fuel it. Coining a term, we observed that the fashion industry operated in copyright's "negative space"-i.e., fashion design was creative work that copyright could regulate but, for mostly incidental reasons, did not. As a consequence, we saw a lot of copying. And yet we also observed a lot of fashion design creativity, season after season, decade after decade.

In the closing pages of The Piracy Paradox, we offered some thoughts on whether other areas of creativity could also flourish in IP's negative space. Other scholars also explored that question, 3 and the scholarship eventually grew into a established sub-field within IP, one which has produced additional studies of the fashion industry, as well as studies of cuisine, fan fiction,

The author thanks Susy Frankel, Graeme Dinwoodie, Maggie Chon, Jens Hemmingson Schovsbo, Barbara Lauriat, and Temitope Kuti for organizing this volume and for inviting me to contribute. Thanks also to Rochelle Dreyfuss for the enormous contribution she has made, over many years, in the IP field, and also in taking the time to comment insightfully and constructively on my work and to help expand and improve it. Those negative spaces showother factors beyond formal IP that support creativity: market incentives, cognitive psychology, social norms, first-mover advantages, path-dependency, or even plain happenstance. The challenge ahead is to map out these factors across fields that vary enormously in terms of markets, participants, practices and norms.

Professor Rochelle Dreyfuss was one of the early commenters on the negative space work, first in a short 2007 commentary, Fragile Equilibria, 6 and then in a longer essay published in 2010, Does IP Need IP? Accommodating Intellectual Production Outside the Intellectual Property Paradigm. 7 As is typical, Dreyfuss was generous with the work of (mostly, at the time) junior scholars trying to make sense of an interesting innovation puzzle. And as is also typical, when she focused on the arguments Kal and I had made in The Piracy Paradox, she adumbrated them with far greater economy than we had managed:

As the authors tell it, negative space works for the fashion industry because the goods send status (positional) signals, these signals decay over time (obsolescence), and both the status and the decay are, in a sense, caused by the The result of these dynamics was, in both our telling and Dreyfuss', an equilibrium permitting rapid innovation in the US fashion industry with little support from copyright protection. That said, although Dreyfuss credited our core argument, she also immediately posed some difficult questions about it. She asked first about how far the negative space arguments could go: Could we imagine non-IP innovation incentives displacing IP? She wondered also what the domain of the negative space argument was: Were there likely to be other important areas of creativity that could function in the negative space of copyright or other IP rights? And finally, there was the question that animated the title of Dreyfuss's 2007 commentary: Would negative space equilibria endure? Or would they tend to break down in favor of IP-focused innovation?

In this essay I'll offer a few thoughts on each of these questions, and particularly on the final one, which is (at least to me) the most interesting. The bottom line is that the questions Dreyfuss raised about the negative space scholarship are not yet resolved. As I shall explain, however, some of what we have seen in the years since suggests that the potential domain of IP's negative space may be somewhat broader than Dreyfuss initially suspected-in part because non-IP incentives are unlikely, on the whole, to displace IP rules, but rather to complement and, to an extent, limit them. And on the third question, there are some developments that suggest that negative space equilibria are more likely to endure than Dreyfuss may initially have thought-and also, on the other side of the coin, that IP-based innovation equilibria may be fragile in the right (or, maybe more properly, the wrong) circumstances.

Dreyfuss raised important questions in the early days of negative space scholarship, when some were wondering (perhaps too optimistically) whether non-IP incentives could displace IP altogether across most or even all of the full breadth of creativity currently protected by some form of IP. This evangelizing aspect of the early negative space scholarship is reflected in the title of her 2010 essay, which includes a phrase that has become attached to this branch of scholarship: "Intellectual production without intellectual property."

In fact, almost every field that has been studied in the negative space scholarship reflects a mix of IP and non-IP incentives. That was true of the fashion industry as Kal and I described it in The Piracy Paradox. The US industry enjoys very little copyright protection. Copyright covers fabric designs as well as illustrations or appliques printed on or attached to garments. But it does not product the shape or cut of a garment; i.e., the central elements of a fashion design. And yet IP is not absent from the fashion industry. Fashion firms police trademarks heavily, even as their designs are open to copying.

As Dreyfuss noted, although we did argue in The Piracy Paradox that "relationships in the fashion industry are cemented by trademark protection," 9 we did not describe the role of trademarks in any detail. 10 We remedied that in a follow-up piece, The Piracy Paradox Revisited: 11

In The Piracy Paradox we did not focus on trademarks per se, but they played an important subsidiary role in our analysis. Labels and marks are essential to many apparel buyers, and not only in relation to the paradigmatic consumer search cost and producer reputation concerns that undergird the economics of trademark…. The central point is that while fashion design operates in a low-IP regime, it does not operate in a no-IP regime. Rather, while copyright protection for fashion designs is almost entirely absent, the apparel industry is strongly protected by, and invested in, trademark law. The existence of powerful brands is an important aspect of the economics of fashion. Once a trend becomes widely adopted, its days are usually numbered. But in the interim, as the trend is building, buyers face a plethora of similar designs. To a large degree consumers choose among these based on price, quality, and the like. But they also choose based on labels. 12 Thus, the presence of trademark protection supports innovation in the absence of copyright protection. Trademark law helps channel gains from innovation, even if it does not make them fully and exclusively recoupable by the innovator. What we get is a delicate balance between IP and non-IP incentives, one which preserves a wider freedom to copy than a fully-IP-driven 9 Fragile Equilibria, at 3. 10 See DIPNIP, at 1450 ("Raustiala and Sprigman make a strong case for why fashion can function in the absence of copyright, but they largely ignore the role played by trademarks."). equilibrium would likely present, but which also provides some protection to firms that invest in innovation over the long-haul in the form of reputational returns protected by trademark.

Thus the correct response, in my view, to Dreyfuss's first and second questions is to note that her skepticism of the most ambitious low-IP account has merit, and to direct the main thrust of the negative space scholarship, instead, toward describing how non-IP incentives can complement IP or limit the need for it. This is the narrative that emerges, for example, when we compare the low-IP US regime for regulating databases with the higher-IP EU regime. We see in the US only marginal regulation of databases by copyright: i.e., copyright does not protect the factual content of most databases but only narrowly protects original schemes of selection, coordination, or arrangement-and generally only against exact or near exact copying. This narrow IP protection complements technological and contractual protections (themselves only partial and subject to leakage), as well as state misappropriation law. Again, this illustrates a low-IP arrangement, but not a no-IP one. In contrast, the 1996 EU Database Directive imposes a community-wide sui generis right in the contents of databases. This high-IP regime was intended to spur database development in the EU, which was seen as lagging the US industry. Those benefits have failed to materialize, 13 and the Directive is now under review and may eventually be narrowed or scrapped. 14 The story in the field of open educational resources, or "OER," is similar in some important ways. OER, which started in the early 2000s as a global movement to make inexpensive, openlylicensed instructional materials widely available, has now gone mainstream, with some of the most widely-used K-12 instructional materials in the US available under one of a variety of open licenses. One example is Eureka Math, also known as "EngageNY," which is licensed under a Creative Commons license permitting free non-commercial use. Introduced in the 2013-14 school year, a RAND Corporation 2016 survey identified Eureka Math as the most widelyadopted math curriculum in the United States. 15 That said, while OER resources have transformed the education publishing market, it would be wrong to say that OER has "displaced" traditional educational publishing and its reliance on copyright protection. Both forms of publishing co-exist and compete. Indeed, Great Minds, which created Eureka Math, now markets a wide array of curricula made available under a variety of licenses. 16 Some Great Minds curriculum is openly-licensed. Other curriculum, including the new Great Minds math curriculum, is made available under a standard copyright licensing arrangement. 17 OER has not displaced copyright in the market for curriculum, but it has disrupted that market in ways that have made it more competitive and innovative. The market had previously been dominated by a small group of large incumbent publishers. The entry of low-IP players seems to have unlocked a lot of dynamism. The story in databases is, at a general level, similar, although we are comparing legal rules in two different jurisdictions, rather than different business strategies within a single market. Does the success of low-IP approaches suggest that we don't need any form of propertization in either databases or curriculum? No. The US approach to databases is "low-IP," not "no-IP." And OER is, like open-source software, a form of private ordering, rather than a legal rule imposed from above. Low-IP innovation in this field may serve as a competitive check on firms that engage in IP-driven innovation.

Thus a valuable role for low-IP innovation to play is not to displace IP, but rather to discipline its use. Indeed, that is a central part of the story with open-source software. Many, including Dreyfuss, have noted that the most important open-source project, Linux, is subsidized by a titan of IP-driven innovation, IBM. 18 That is true, but it is also a fact that, at least in my view, shows why low-IP innovation remains vibrant in the software field: because it is a way to prevent important markets from being dominated by firms with dominant IP. That is, IBM supports Linux because the alternative is Microsoft's dominance in server operating systems, which would threaten IBM's ability to compete both as a provider of software and 15 Nicole Gorman, Survey Reveals New York Math Curriculum Most Widely Used in the Country, Education World (Apr. 21, 2016), https://www.educationworld.com/a_news/survey-reveals-new-york-math-curriculum-most-used-country-492126618 16 Note that the author serves on the Board of Directors of Great Minds, and has provided legal counsel to that company. 17 See generally www.greatminds.org. billion in US revenues) has long operated within a low-IP equilibrium and continues to do so. That a large part of the US software industry (approximately $375 billion in revenues) also operates in a low-IP environment-a situation that, again, has been stable for decades-also stands against the notion that low-IP equilibria are inevitably fragile, even when, as in software, low-IP and IP-focused innovation co-exist and compete. We see a similar story of stability in other long-standing negative space creative fields.

Comedians have been operating in an informal, norms-based low-IP equilibrium ever since they began the move from the no-IP "corn exchange" of free appropriation in the mid-1960s. Research, 8 Innovation Pol'y & Econ. 1, 11 (2007).

24 Jon Marcus, Think universities are making lots of money from inventions? Think again, The Hechinger Report (Jan. 17, 2020), https://hechingerreport.org/think-universities-are-making-lots-of-money-from-inventions-think-again/ ("US colleges and universities are producing a surprisingly small proportion of the nation's patents and startups and making so little money from licensing inventions that, at many schools, it doesn't even cover the cost of managing them."). that newspapers had long operated according to a strong norm that allowed morning papers to appropriate stories from editions sold the previous evening and vice versa, but did not allow papers published at the same time to copy from one another. The thirst for immediate access to news created by World War I, however, led to the break-down of that norm. The lawsuit succeeded in creating a misappropriation-based property right in so-called "hot news." As Dreyfuss has observed, there is an alternative explanation, based not in changing norms but in technological shifts:

[P]rior to the time of INS, news was distributed on large networks of leased telegraph Because the cost of maintaining these networks was so high, a newspaper that owned them enjoyed a strong competitive edge. But in the early twentieth century, technological improvements reduced distribution costs. As a result, the advantage of owning leased lines disappeared, and the industry shifted its attention to the ownership of content.

[T]he INS case was constructed in order to create a property right in news, which all of the papers then needed to recover their costs and earn a profit. 28

This is a plausible story, but it's perhaps too simple. For as we look at the newspaper business from the perspective of 2022, we see all around us the ruins of the IP-focused equilibrium that INS built. And again, it was technological change that undermined the equilibrium. This happened on several fronts, all of which were connected to the growth of the Internet. Craigslist snatched away much of the classified advertising business that supported newspapers. Google and social networks like Facebook made off with much of the commercial advertising. And perhaps most significantly, Internet distribution of newspapers themselves means that elite newspapers like the New York Times, Wall Street Journal, and Washington Post are now effectively competing with second-tier papers, not to mention third-tier newspapers in hundreds of smaller cities and regions. The result is a concentration of power in a small number of outlets, a decimation of the bulk of the market, a collapse of local news coverage, and, arguably, a democratic crisis. If that isn't a story about the fragility of a high-IP equilibrium, I'm not sure what is. In fact, we now see efforts in many localities-The Texas Tribunefoot_2 and Charlottesville Tomorrowfoot_3 are two prominent examples-to turn local and regional journalism toward a non-profit, grantfunded model in which journalism is framed explicitly as a public service and copyright is basically irrelevant. That is, we may be seeing low-IP green shoots emerging from the burnedover soil of the high-IP equilibrium in journalism.

For the moment, the comparative fragility of low-IP innovation incentives is far from selfevident. That said, as this short essay has tried to make clear, many of the questions Dreyfuss raises about the reach of the negative space scholarship, and whether low-IP equilibria can operate in markets that traditionally have been governed by IP, are still pending.