Knockoff Economy session III

Chris Buccafusco (Chicago-Kent College of Law)

What level of creativity would we have with formal IP?  The studies can’t tell us.  The problem with the approach is that it doesn’t give us the necessary counterfactual.

Specifying the mechanisms by which creativity and innovation continue to occur in these low-IP fields is very helpful.  What else?  More comparative and historical studies.  IP protects perfume in Europe: do we see innovation?  Historically—Stan Liebowitz has done work on the effects of downloading on production; Paul Heald has looked at how public domain status affects entry/investment.  Experimental evidence: issues associated with attribution.

What about “real art”?  Music, movies, literature.  How much do these studies of marginalized production affect “real” culturally valued production?  Difficult to answer.  Internal v. external incentives.  That may be why we never see “failures” of entire fields—people like creating stuff, and often continue to do so.  The question is: how much more could we be seeing if we did have IP.

Role of performance.  Sometimes it’s possible to preserve the role of individual performance and incentives related to a particular performance: music, food, adult entertainment; less so with pharma, movies, etc.  If chefs are not able to protect the food, they shift to performance elements.  Competition evolves around performance, not content, which might be problematic for social welfare; if we really value chefs for the delicious food they make, then the ones who make good food will lose ground to good performers.  Book on creating country music/manufacturing authenticity: in the 20s/30s, desire to create music that sounded authentically folk/country because it could get IP protection rather than performing public domain standards.  (Isn’t this the music version of churn?  If you can separate composers from performers, I’m not sure why this is a problem, since people who are good at composing can do well without losing out to people who are good at performing; though maybe you can’t.)  In all these industries, there are differences between the incentives of different kinds of players—chefs don’t care about copying, but cookbook authors do; Food Network sends takedown notices to people posting Rachael Ray recipes.  Content is highly copyable and can’t be substituted with performance. Have to make normative decisions about who we care about.

We don’t want innovation no matter what; we want cost-justified innovation.  Every single year, Balenciaga has to invest in creating a new dress, and the ones that are failures represent lost social welfare.  The innovation we get may be more costly than justified.

Bob Ellickson (Yale Law School)

For further work: What does it mean to “thrive”?  We are getting the exact right amount of innovation in fashion, football, etc.? Weaker version: in a world of transaction costs, we can’t do much better. Third, weakest version is a slam dunk: however we’re doing now, patent and copyright would make us do a lot worse.

Overall systems of social control: formal associations that make rules.  Notre Dame Law School makes rules about food in the library: neither law nor norms.  In all the contexts studied, there are formal associations—maybe even tattoo artists.  Question: are these ass’ns involved in policing copying or not?  There is a national ass’n of comedians.  The ass’ns in the fields in The KE are generally weak.  Why do the comedians set up an ass’n but not get it involved in disputes like this one?  What about strong associations, like condominium ass’ns, which establish and enforce rules?

His view of social control: how do we prevent wrongful copying?  First-party controls achieved through socialization of children: the main form of order in our society.  People don’t copy jokes because they think it’s wrong.  Second-party social control: if you copy, the person who’s been copied will retaliate violently or by calling clubs to keep you from being hired.  Third-party: law; norms (ostracization of offender even if the offender didn’t steal your own joke); associational rules.  The last is the dog that’s not barking here.  Comedians themselves think they don’t need/want that because informal second- and third-party sanctions work well enough; ass’n has no comparative advantage.  Homeowners’ associations are stronger because of weaker social ties—it’s possible to survive being hated by your neighbors but not by your comic colleagues, so the HOA needs stronger/more formal associations to govern behavior.

Nicole Garnett (Notre Dame Law School)

How much can we learn about the appropriate level of social control from the case studies? Elinor Ostrum on effective management of common pool resources; too much gov’t can interfere with that.  Turns out there’s a lot of gov’t in the collective management, too.  Example: mountain-climbing shoes: Sanuk Vagabond, apparently very innovative—TM and patent does protect some of that.  Apparently not used for mountain climbing/functionally, but do signal membership.  Crocs immediately copied the Sanuk Vagabond with the Crocs Santa Cruz, upsetting the mountain-climbing community. Will speed of copying spur improvement in Sanuk’s quality?  Now making new/different Sanuks. Separate problem: fake Sanuks.  With more limits on copying, would there be more counterfeiting as the only way to copy?  Will restrictions on copying lead to more creative production—designing further away from the Sanuk/Croc.

[discussion of use of design patents to protect shoes]

Michael Heller (Columbia Law School)

Fashion in academic book conferences: which books get chosen?  Writing entertainingly/clearly for a mass audience is something academics should strive for. 

Pointed us to ND’s law library food & drink policy: historically none allowed.  Change in that in recent years.  Noticed disjunction between formal “law” and actual practice of students.  ND librarian researched the range of policies around the country and incorporated elements from others while making its own. Three distinct zones: no eating & drinking, no exceptions, in special collections. Drinking ok but not eating in computer room (why not the opposite?).  Rest of the library: lengthy policy allowing light prepackaged snacks not damaging to the materials and not disturbing to others, with examples (disturbing = aromatic; prepackaged = pretzels, not burgers).  Initial concern was messy food.  But the only all-caps part: no outside delivery permitted!  What if you tried to have a light prepackaged snack delivered?  (How likely is that?)  The all-caps rule trumps.  So moved beyond the initial concern.  Set of informal norms, law trying to catch up over 25 years.  What’s changed as the law became increasingly elaborate and covered more of the norms?

Well, nothing has changed in what students do.  Students brought in snacks before and they still do; they’d get in trouble for pizza and they still do. What’s changed is the effort in defining what’s ok and publicizing the rule.

One possible response is, so what?  If the actual practice before and after is the same, then there’s a possible invariance theorem: law/IP doesn’t matter.  People simply do what they’re going to do, or work around it. Omri Ben-Shahar showed mathematically that a long adverse possession period led to relatively little individual effort to police, and a short period led to the opposite; the total amount of adverse possession would be invariant as to the law.  (This may be the paper.)  Heller thinks Ben-Shahar is wrong because he doesn’t account for the effects of very specific guidelines in front of people/shaping norms; harder to adapt to new types of food/new types of behavior.  In a world of fast change, formal rules are more troubling.

Me: On how much more we’d have if we had IP in these spaces: Glynn Lunney has done work on this—tradeoffs with other kinds of investment.  Relatedly, scarcity of attention.  How much more can you actually experience? “failed” genres and the reasons they fail other than copying: w/o fully endorsing, let me quote Nicholas Carr: “The oral epic poem, the symphony, the silent film with live musician accompaniment, the dramatic play, the short-form cartoon, the map, the LP. Most of these still exist, particularly on the consumption side, but they’ve all been diminished.”

Garnett: what are the connections between norms and strong associations? If the norm is strong, the association doesn’t need to be, which is good because ass’ns can become highly overbearing—especially homeowners’ associations! 

McKenna: As Avishalom was saying: innovation doesn’t just descend down on consumers with no consumer-side aspects, as if it’s delivered with no question of what they want.  We don’t actually know whether consumers want shoes that are very different from Crocs v. shoes that are close to Crocs at very different price points.  Another complication of the counterfactual.

Sprigman: Came up in Apple v. Samsung. If we had less copying and had more “pioneering,” then Samsung has to make its phones very different—but if consumer demands are supposed to drive innovation (and where those preferences come from is a complicated question).  If Apple came up with a user interface that people want, with variations, but they don’t want something that’s very different.  As if one car company could have a driving interface that was a wheel and everyone else has to use handles.  Consumers may not want that kind of differentiation!

Comedians’ ass’n used to be a lot stronger; wanted to organize over health care.  One thing that weakened it was that the ass’n hired a lawyer to see about the possibility of suing for copyright infringement of jokes, and it weakened the ass’n because people split on whether this was a good idea.

Raustiala: Council of Fashion Designers of America; trying to change norms about copying/knockoffs.  MPAA is trying to do the same thing (Beebe points out that they’re trying to get it into schools, with some success).

Shaver: Slippage: the difference between the way the law is supposed to work and the way the law actually works—higher in some countries than in others through gaming, error, fallibility.  Challenge of defining what’s new/infringing is one of the slippage points.  Law can take on a life of its own and lead to overprotection.

Buccafusco: does it have to be political economy? Couldn’t it be a real economy issue driving the increase in scope, term, etc.  We see unidirectionality but we don’t have the counterfactual.

Perzanowski: Steve Jobs said consumers don’t know what they want until he shows it to them.  Made a lot of money that way—preferences aren’t always known in advance.  Counterfactual about the right level of protection: First we need to figure out qualitatively what kind of output we want, and what the optimal level of production is before we can figure out the optimal level of protection.  There may be overproduction in tattoos; clients could be satisfied by choosing from existing designs—but they want to make the monkey dance/see the artist go through the steps of creating something new. Adult entertainment: we already have plenty; why make someone create a new performance?  If we don’t know the optimal level of production, we won’t be able to talk about optimal level of protection.

Buccafusco: zero-sum investment doesn’t mean zero-sum output—if we shift from one type of media to another, we might get different amounts out of them. There are good reasons we want to live in a world of 21^st century cuisine instead of 14^th century cuisine: taste, variety.

Qian: why no innovation before the entry of counterfeits in her study?  If we think of different brands as having monopoly power within their own niche, their own brand serves as a natural entry barrier/horizontal differentiation allowing people to self-sort, not because quality is higher but from personal preference. Counterfeits break the natural monopoly.

Income inequality positively relates to demand for counterfeits: empirical finding based on studies in China and lab experiments: the link is through power/status signalling. If people are less threatened for being poor, less surrounded by superrich, then they feel less powerless and feel less need to buy counterfeits: that’s an anticounterfeiting strategy without IP.

Samuelson: fast innovation has been problematic in software (settled down now).  Release 1.0, 1.1, 1.2, etc. and dealing with customers with different levels of the product—companies ended up getting penalized by consumers if they innovated too fast. 

User innovation stories of Eric von Hippel and collaborators: companies basically adopt/appropriate pioneering innovations in info tech/sports equipment.  Another way in which you get innovation-as-improvement but without using IP/with very strong sharing norms.

There are a couple of places where we could see some interesting experiments: SCt decided that diagnostic methods aren’t patentable; they aren’t copyrightable; and they can’t be kept as trade secrets.  How much investment was there prior to Prometheus, and how much will there be post-Prometheus?  Even bigger deal is Myriad and patentability of gene sequence; biggest argument against rights is creation of patent thickets, preventing research instead of helping.  If the SCt strikes down the Myriad patent, what happens to innovation in that space?

Copyright side: until early 1990s, the US didn’t protect architectural designs, and then it did.  The architects weren’t enthusiastic, but the treaty required it and the copyright pros wanted to make it look like we were in compliance with Berne and architecture was the easiest sacrifice.  Could look at effects on innovation.

McKenna: the ones who use the system to litigate aren’t high end designers but subdivision builders, not high end builders.

Buccafusco: same with food.

McKenna: certain players who push for stronger TM/copyright may be the ones who benefit most from the current state of affairs.  Do they not get it?

Gosline: When you talk to luxury brands, they stick to the party line, but they do understand that they benefit from infringement and enjoy a particular status within their categories where they are less likely to suffer harm than others in the middle categories.  High end stuff that is less well known is less likely to be counterfeited—the heavily logoed stuff gets copied, but not bags that cost $15,000/bag.  Have resources to create the super-high-end stuff that allows them to survive/benefit from imitation at the lower end.  “Fakes are never in fashion” campaign etc. are trying to control the behavior of people who might be their consumers.  But they’ll never say that on the record.

McKenna: they litigate against people who seem to helping them.

Gosline: yes, against distributors, but they’re not chasing the individual consumer who buys a bag.  Distribution channels are different.

Qian: Provides results to companies that give her data; some have an a-ha moment where they see the benefits of counterfeits—but there are lots of lost sales on the low end. Overall average effect across product lines is still negative.  Also looks at brands at different stages—counterfeits helped brands that were less renowned at the time of infringement.  Once the brand is established, they want counterfeits shut down.  That’s just what Microsoft did. Only when it had a 90% market share did it start fighting piracy.

Gosline: it’s mass exclusivity, which is an oxymoron.  If you pull back the curtain and say everyone’s got some luxury, that undermines it.

Sprigman: this is an industry of lawyers/brand protection experts dedicated to doing precisely this. (As Upton Sinclair said, it is very hard to get a man to understand something when his salary depends on believing the opposite.)

Raustiala: cultural differences?  In Europe, where so many of these firms are headquartered, are the norms different?

Gosline: differences according to whether you have a new upwardly mobile class with disposable income.  Doing some work in India now.  Dynamism: people wanting to signal they now have loads of money.  Europe/US face more stability there.  Notion of couture—pretty common in India to have shirts made for themselves; doesn’t differentiate rich from poor in India, which creates a challenge for luxury manufacturers to convince people that’s why you should pay so much.

Recent entry of luxury manufacturers and media is serving to create value—fashion industry is training people to look forward to these products. 

Qian: experiments in China and US have similar results qualitatively; quantitative scale is larger in China, and field data so far is only from China.  Effects larger for China: advertisement effect. If consumers are unfamiliar with multinational brands they rely heavily on common opinion for knowledge/awareness.  Converse is not famous in US but is in China because of all the counterfeits.

Knockoff Economy session II

Yi Qian (Northwestern University Kellogg School of Management)

Misattribution of jokes to comedian Steven Wright: he says it didn’t affect his show; some are so horrible he was embarrassed by the misattribution, and some are amazing and he wishes he’d written them.  Does this reaction make sense for counterfeits?

Counterfeit sales are huge.  Research agenda: what are the consequences of weak IP laws?  What are the impacts, origins, and remedies of counterfeits?

Impacts of entry by counterfeiters: endogeneity of entry raises ID challenges.  Look for situations where some are infringed and some aren’t to see effects.  Unique features of counterfeits: usually illegal; similar appearance but usually inferior functionality; intent to deceive (intend to fool consumers; or buyers may intend to fool friends/signal status).

Key findings: authentic’s quality, as proxied by cost, increased. IP not necessary, just as it’s not sufficient, for innovation.  Authentic prices increased after initial drops.  Vertical integration of distribution channels was an effective defense. Advertising also became more informative.

Market is self corrective because private parties compensate for lack of public enforcement when counterfeits become pervasive; would establish offices to work with gov’t.  As compared to generic entry, counterfeits cause companies to invest more on improving searchable quality and less on experiential quality, which could be more valuable for some industries, e.g. fashion, than others, e.g. pharma.  Social welfare could be nonlinear as gov’t enforcement increases.

So, counterfeits can stimulate innovation. Low quality entrant may increase category or brand awareness (word of mouth), which may benefit everyone.  Increased purchases in core channels. May enable compensatory consumption: when low-power consumers have no resources, can compensate for feeling lack of power by consuming counterfeits, which is supported by field studies and lab experiments.  Network effects: Bill Gates said if there has to be piracy in China, let it be Microsoft.  Effects differ for different quality tiers.

After counterfeit enters: Average sales of medium-end product decreased mildly; significant drop in sales of low-end products; increase in sales of high-end products.  Even within brands: High-end Nike sales increased and low-end decreased.

Experiment: 62 participants: treatment group members were exposed to images of counterfeits.  Treatment group had lower purchase intent for the authentic low-end shoe when they see counterfeit alongside authentic and higher purchase intent for the authentic high-end shoe.  Stated motivation: “Brand is a guarantee for quality. I don’t know this brand, but it should be good if it has counterfeit followers.”  Some say they like variety; others say they don’t care about authenticity if they product fits needs.

Balance of advertising and incentive effects enhances authentic’s incentives to move up quality ladder; should prioritize enforcement to counterfeits for low-end product lines. Optimum IP protection could vary from country to country, brand to brand, product to product.  (Compare

Constantine V. Nakassis, Counterfeiting What? Aesthetics of Brandedness and Brand in Tamil Nadu, India, 85 Anthropological Quarterly 701 (2012) (discussing meaning of “brand” in Tamil Nadu).

Pam Sameulson (UC Berkeley School of Law)

Drury v. Ewing, old fashion case: Mrs. Drury was designated as the authress/inventress of a chart (the closest thing then in copyright’s subject matter to a dress pattern), copied by Ewing.  Defended on the ground that the drawings were different, but the argument was that the resulting dresses were the same; court found infringement.  From that decision, would have been a small step to say that instructions in magazine/drawing would be copyrightable and infringed by the derivative work/the dress.  Hard to call a dress a chart, but subject matter cases show that words like chart become elastic when occasion arises. Drone’s influential 1879 treatise endorsed Drury as the way to go, so when the SCt reversed in Baker v. Selden the course of the law changed.  Other legislative close calls on fashion followed.  Distinction between writings (copyright) and useful articles (essentially consigned to patent)—quasi-constitutional understanding, in part because of the use of “respective” in Art. I cl. 8 sec. 8.  This understanding was part of why semiconductor chip protection wasn’t pure copyright, even though Arthur Miller compared the layout of a chip to a Mondrian.

The KE is good because we no longer have that quasi-constitutional barrier; robust examples of lack of need for such protection is a good bulwark.  Claims exist for perfume, etc.; examples of thriving industry can show why we should restrict copyright subject matter.  Would a grant of exclusive rights encourage markets or not?  How well would a copyright regime fit onto the kind of subject matter at issue (synthetic biology is particularly terrible here)?  Are the kinds of creators doing this kind of thing like the creators we have in copyright already—are their processes and artifacts similar? Copyright should be reserved to communication of intellectual content to human beings; software doesn’t fit into that well, though it does fit some of her criteria.

Aaron Perzanowski (Case Western Reserve Law School)

Most of us have an impulse to move from the concrete to the abstract: come up with grand unifying theory explaining motivations for and substance of norms.  The KE gives us lessons from case studies, consistent across some industries, but we should still resist the urge to generalize for the time being. Most important lesson: importance and need to inform IP policymaking through facts on the ground.  Abstract theory can distract from that work.  The KE shows admirable restraint in this regard.

World of custom tattooing: unique, one of a kind, bespoke tattoos for individual clients.  One in 5 American adults has a tattoo (how many are custom? Market segmentation is the kind of thing I’m interested in).  Formal assertions of legal rights are incredibly rare.  Not because people are ok with copying or because they can’t establish effective legal claims against copyists.  They’re squarely within the Copyright Act. But they rely on social norms instead.

Norm: favor client autonomy.  They can even reproduce images for commercial purposes.

Norm: disfavor copying custom design by another tattooist.  Also violates norms for tattooer to reuse a custom design for a second client.

Fundamentally, both norms are driven by client demands and expectations.  Clients want custom tattoos, and they want to use them in whatever ways they choose.  Tattooists could defect from these norms, avoiding opportunity costs of designing new ones.  Extract rents from clients who want to post pictures on FB.  In the long run, that undermines the market for the industry as a whole.  Norms protect longterm collective/community interest.  (Though enforcement seems likely to be hard when you can get a big settlement from the producers of The Hangover.)

Norms influence output, but output influences norms. Similar to standup comedy: norms emerge at a time of radical transformation from your grandfather’s tattoos (hearts that say mom) aka flash to the model of more personal, narrative tattooing.  Causation probably runs in both directions.

What does that mean for innovation? We have more creative production/different creative production.  Is it better? Not sure of that answer.  Norms do favor one kind of expression over another.  Before, you could look at a sailor’s tattoo and know their life story—after 5000 miles you get a certain tattoo in a certain place; iconography was shared.  Today, more fractured/harder to interpret.

Also a distrust of lawyers—not edgy to hire a lawyer/lose street cred.  History of suppressive regulation on tattooing; most courts said tattoos weren’t speech but nonexpressive conduct. Both the substance of the norms and the motivation for relying on them is a function of particular facts and circumstances.

Dan Kelly (Notre Dame)

Sees two theses in the book: social norms and other substitutes for IP that may provide incentives to innovate.  Another stronger claim: copying actually accelerates innovation.  Convinced by first, more skeptical of second.  Is there a tension between these two aspects? 

Choices are generally thought to be good: many kinds of cheese are better than just one.  But if I have to wear a different suit to class every day, that may go too far.  (That’s not really the same thing: now you’re talking about pressure on the consumer to buy, instead of choosing her one suit from a wide range.  It’s true that there can’t be the pressure in the absence of the wide range, but we can easily see—I see it in class and in my closet v. my husband’s—that the pressure differs even if there’s a wide range available for women and for men.)  Financial innovation too can be problematic.  Focus on type or kind of innovation incentivized.

Chefs at the high end: they feel the need to innovate. But what’s the effect at the lower end?  Do we see innovators at the low end or are they driven out?  (One could ask the same about comics; there are of course the cruise ship/dinner speaker comics at the lower end.)

“No copying” is drilled into us as a social norm; maybe we should rethink that. 

Buccafusco: how do we think about norms in the shadow of law v. norms in the absence of law—tattoo artists don’t use available law.  Also, chicken & egg question—do people make anticopying claims because copyright law exists? Well, there’s a history of people making anticopying claims about cookbooks in the 15^th century, but it’s a reasonable question.

Tattoos: could this be handled by contract?  Wedding photography is a potential comparator, where contract seems to have come in to define the appropriate relationships.

Sprigman: worst thing for wedding photographer is that they sign contracts, then the mother of the bride digitizes the photos and sends them to everyone—contract is a partial tool, but not enough because that person isn’t bound by the contract.

Buccafusco: what’s the difference between wedding and fashion photography?

Sprigman: depends on where the photographer is on the scale—lower down, they want the models to use the headshots freely because they get attribution; at the top of the scale, you negotiate.

Perzanowski: tattoo artists aren’t particularly interested in ownership; one example involved an assignment (a failed attempt at a WFH agreement).  Mostly they don’t want people hiring lawyers, creating paperwork for everyone to fill out.  Initial reaction is often: “oh, they did that because copyright/law doesn’t really help them, doesn’t protect what they care about—premise of the joke, recipes, etc.”  We assume the law has failed, and thus conclude that they wouldn’t have to rely on poor substitutes if they had law; but here we have a group that has the law and says no thanks.

Kelly: most general level, some of the disagreement between pro and anticopying camp is about how effective Coasian bargaining is.  If transaction costs are low, then people shouldn’t be allowed to copy (why does that follow? If they’re really low, the right will go whereever it’s most valued), but if they’re high they should be.

Buccafusco: turns out even with apparently low transaction costs you get neither contract nor copyright.  (But doesn’t that suggest that the right is with the person who should have it—the consumer?)

Robert Ellickson: good way to get back to Beebe’s question.  Should we treat all this self-presentation with fashion, luxury goods, tattoos as an arms race in self-presentation that’s wasteful?  Possible argument: presents information about people; fashion churn means more information is available from clothes—see something absolute and see if individuals are responding to the churn.

RT: On “no copying” as a social norm: No, it isn’t!  “No copying of certain kinds” is drilled into us (which gets to Samuelson’s point about subject matter and how we really need to think hard about it to make the system work); but try telling your teacher that you’re no copycat so you aren’t going to copy the spelling previous people used, or the way of writing letters they used, or the essay structure, or the method of proving a theorem in geometry … That might make a good Calvin & Hobbes cartoon, but you will fail your grade.  Some things count as copying and others don’t, and the invisibility of the things that don’t affects how we think about “copying” in general.)

Qian: we often see brands launching low-end stores; this can benefit brands through awareness in the same way counterfeits do. Capture low-end consumers in vertical differentiation as well as advertising selves. 

McKenna: consumers are really good at segmenting and still believing that the high end is high quality.

Qian: counterfeiters could be seen as buzz agents advertising for the brand.  Brands don’t internalize the costs.  Can also show them what are viable markets if considering expansion.

Heller: Mike Tyson case?

Perzanowski: People see it as a real outlier.  Plaintiff is still somewhat consistent with the norm; he didn’t sue based on Mike Tyson’s presence in the first movie. It was when it moved to another actor in a commercial context.  It’s an interesting case because it demarcates the outer boundaries of the client autonomy norm.  Mike Tyson can do whatever he wants with his tattoo (except authorize a Mike Tyson-related storyline, apparently! As McKenna then said).

Fagundes: whether it matters in the shadow/not covered at all?  His suspicion is that it doesn’t: people were wrong about what the law was, in weird ways.  Derby girls: called it theft or copyright infringement.  These views were surprisingly persistent.  He could tell people the real rules, and they’d ignore him.  Lay views about the law are more strongly determined by norms/morals than by law itself.  Once they learned what the law was and that it could make an intervention, they were uninterested in using it even if they were infuriated by violation of social norms.  Outsider status was relevant.

Ellickson: hey, we (law professors) ignore copyright as we prepare for class too.  This is universal, not outsider.  If a law prof sued another prof for copying an article for class, that prof would be ostracized.

Raustiala: in the fashion industry, wrong lay beliefs are also common.  People think: if you copy 30% that’s fine.  They were resistant to true information, too.

Kelly: even if the individuals don’t know, presumably the legal department knows.

Raustiala: a lot of them don’t.  It’s all about size. Industry structure is important. Fashion has a lot of tiny little actors—150 in a factory and 2/3 at the top; big theme = we don’t want to get lawyers.

Avishalom Tor: Move from specific to general—look at quadrant where there is copying and success. What about the quadrants where we have copying and no success?  (Or no copying and no success?)  Questions about upfront costs—don’t tell me when it works, tell me when it works and when it doesn’t—what differentiates?

Raustiala: conceptually, if you really had a creative domain with no continued production, we’d want to look for that path. We don’t think that music, for example, is on that path, though a particular model of the music industry is declining.

Sprigman: the claim is that biologics will have this problem because you can’t get patents on them and yet there’s still high investment to produce commercial quantities.  He has no view on this as yet; wouldn’t be surprised to see a sui generis right.  But not an abundant literature on tech failures because of copying.

Tor: Steve Shavell’s book—patents/medieval example?

Samuelson: the semiconductor people had that story 1979-1984; Intel said it costs millions to make a chip and $50,000 to copy it.

McKenna: but that’s the form it usually takes—just a projection. Do we have evidence of a refusal to protect and subsequent collapse?

Samuelson: passage of Semiconductor Chip Protection Act stopped the copying and led to innovation; people inside the industry think that law led to more innovation—still copying, but now with tweaks in a constructive way.  Then it led to very different industry structure over time: design firms and separate mfg facilities where before it had been very integrated.

Sprigman: fab plant is expensive; design is also important—theorizing what you’re doing takes time/research. 

Samuelson: Intel had evidence that they were losing market share despite being innovators.  We were terribly worried about the Japanese.  Then they also went to branding—Intel inside—to guarantee quality.

Roundtable on The Knockoff Economy part 1

Notre Dame Law School Roundtable on The Knockoff Economy by Kal Raustiala and Christopher Sprigman

Welcome and Introductions (Mark McKenna)

Session I

Barton Beebe (NYU School of Law)

The KE raises fundamental questions about what we mean by “innovation” and about what kind of innovation industries without intellectual property are able to produce.  Dave Fagundes’ work on roller derby naming, Aaron Perzanowski’s work on tattoo artists fit into the book’s framework.

Many of the industries in the book aren’t technological but appeal to taste/aesthetics: cooking, standup comedy, fashion, fonts—may be high tech but still based on aesthetic choice.  Football strategies, financial instruments, open source/database—separate from his comments.

Can one’s thinking be directed by the divide between the aesthetic and the technological?  If so, what do we mean by aesthetic innovation?  The Constitution apparently instructs us to seek aesthetic progress.  Slippage in the book between innovation and creativity: do you really want to talk about innovation in standup comedy?  Case studies: innovation seems to mean “more,” diversity, quantitative definition rather than qualitative.  More stuff out there as a result of sharing communities that allow copying or regulate it via internal norms.  Some of the more might be better, but that’s not the focus. 

Maybe that’s no problem but in fact all we can ask for.  Let me push against that: excellence should be part of our definition of innovation.  Something that is new and that has value.

Fashion: troubled by the basic argument that copying allows the churning of fashion and thus demonstrates innovation.  The core example of the book is the fashion industry, but Beebe sees an industry of renovation, not innovation.  Oprah to Ralph Lauren: How do you keep reinventing?  He’s not a pioneer.  He’s a tweaker, but not even a tweaker of fashion.  Vintage-inspired women’s clothing: another example from the book.  Ralph Lauren is a great innovator in branding and the sale of distinction, but not in fashion.  How many pioneers have there been in fashion, or is it just infinite tweaking?  Susan Scafidi would reel off names of true innovators, but he’s not sure he’d call that apparel fashion; at some point it becomes art.

Book recognizes this problem; discusses how design patents don’t work because so many designs are reworking.  Fashion houses don’t claim this season’s offerings are “better” than last season’s, just that they’re different.  If this is the core example of innovation, Beebe considers the thesis not proven.  Creativity, yes; innovation, no.

If there were more copying, would there be more innovation?

Innovation seems to be in branding/commodifiable forms of rarity.  Normatively: Is this the kind of innovation we want?  Standup comedy or better food, sure, but he wonders in fashion whether we should be concerned about it. [I note there’s a gender piece here in what kinds of variety Beebe appreciates.]

Renee Gosline (MIT Sloan School of Management):

The “Real” Value of “Fakes”: Resilience in the Face of Brand Imitations

Power dynamics reflected in the way we signal identity, membership, status, Anglophilia, etc. Counterfeits are democratizing on the surface, but a lot of the walls are resilient to this process.  Brooklyn Museum discovered that some of its antiquities were fake: decided to display them along with the originals.  It showed people how to value the real thing by juxtaposing it with the fake. 

Her research is on the consumer side.  Regulation works with norms and cognitive processes; until those work with the regulation, you don’t truly have an institution.

Does counterfeiting result in contamination for consumers of authentic brands?  Are consumers substituting counterfeits for authentic products?  Her study asked people to figure out whether a product was real or fake—they often had a hard time at least via pictures.  In the real world, you often can’t inspect the lining of another woman’s purse.  When you don’t have social cues, confidence in ability to discern real/fake goes down, and willingness to pay thus goes down. 

When you show the product in context, people are more confident in determining whether or not a product is real or fake.  Authenticity is a product of the social context; who seems inauthentic to the product.  Willingness to pay also is affected by social context.

FB group: “Darling, I can tell by the rest of your outfit that your Louis Vuitton is fake.”  Social shaming: mock imitators.  In the absence of successful regulation, norms still protect the brand.

Consumer interviews: response to raids in Chinatown: consumers say “This is a waste of time for everyone.”  Bostonian said: buying fakes doesn’t put you into the rarefied air of rich people.  People on the ground decide what authenticity is and penalize those who try to step into the “wrong” place.  Having the receipt from the store isn’t enough.   [That famous Pretty Woman scene?]

Many times, counterfeits don’t impact willingness to pay. If there’s no social context, though, there’s a significant impact on willingness to pay.  If the product is counterfeit but the user is high-status, then people do change willingness to pay.  Fertile area for investigation: those are the people who others want to imitate.

What about consumers of counterfeits?  2.5 years of ethnography on a street market (82 transactions, 32 interviews); social network (Tupperware party with counterfeits instead; upper middle-class): 112 consumers at 14 events, 43 interviews.  Finding: Counterfeits aren’t substitutes.  People who use the fake start to realize how far away from the real they really are; the brand becomes more salient, people pay more attention to it, and people experience cognitive dissonance.  They start to think: if my possessions are fake, what am I?  46% dumped the fake and converted to the real, even though they started off thinking they were too clever to pay that much money.  You can come to love the real through the fake; see people with the fake and love your real more. 

Pam Samuelson told a story about a trip to Hanoi with grads, when the richest people on the bus wanted to know where the counterfeits were—she was fascinated with their fascination.  One woman who was among the most adamant was a senior partner at an LA law firm who’d bought a pirate Bridge Over the River Kwai DVD and it ended before the movie did!  Why was it so important: because it was so exciting!

Gosline: Yes, it’s the high going low (and the low going high) that provides the thrill.

Linda Przybyszewski (Notre Dame Department of History)

Ralph Lauren is not an innovator!  “Designer” is the term used in fashion for true innovators; giant conglomerates travel the world for boutiques looking for reference items to copy.  Hungering for the fashion study equivalent of French chef study.  The garment industry is so segmented, enormous, and complicated; many young designers celebrated as innovators get eaten up. One reason is that they don’t know much about finance, but another reason is copying. 

Elizabeth Hawes: American dress designer, active from the 20s; she was first sent to Paris to copy, but was then kicked out for the copying, and then started to do her own designs. In the Depression, she found it impossible to survive on her own, so she started contracting with wholesalers.  Some would give her royalties.  Other designers were destroyed by wholesalers who didn’t keep their promises of royalties.  She found the “churn” of fashion soul-destroying: didn’t have time to design/innovate.  The cry for something new dulls the senses and stops the imagination.  Ended up writing books, some bestsellers—appreciated the copyright rights she got!  Complained: French designers were financed, protected, and publicized by the gov’t and the fabric industry; not so in America. So do we have a way to measure/compare French v. American innovation?  Hawes ends up advising would-be designers to give up the idea.

What is the alternative?  In industries in which copying is rampant, there’s no point in taking entrepreneurial risk, and at least some people will know that and write books instead.

Rebecca Tushnet (Georgetown Law Center) Norms no less than laws are products of power, and in studying nonlegal regimes for defining and defending creativity we should also be attentive to the power relations expressed, challenged, and reinforced in various fields of "IP without IP."

Innovation in sports as example: innovation in football v. basketball, where preference for the current system has apparently been strong enough to prevent the success of the full-court press even though weaker teams can use it to beat more skilled ones.  Compare Pierson v. Post, which I just taught: Post’s problem with someone swooping in to deprive him of the benefits of the hunt could be solved by gearing up the way fishermen gear up, which is to say by innovating (perhaps adding sniper rifles to the hunt), but that would’ve changed the meaning of the hunt in ways inconsistent with its signification of aristocracy.  So what innovations are considered within bounds depends on the operations of power, and the question is regularly whose preferences about the meaning and scope of a particular field will be put into practice, even in the absence of a formal monopoly like AT&T.

Malcolm Gladwell suggests that “When underdogs choose not to play by Goliath’s rules, they win.”  So one question is why they ever decide to play by the rules!  Gladwell talks about the effort of breaking the rules, but there’s also social disapproval/norms.  In Beebe’s terms: someone gets to judge excellence, and sometimes their rules are different from those of the creators.  Sometimes the creators are really bad judges (my 5-year-old does great work … for a 5-year-old giving stuff to her mom), but I am attracted to the benefits of quality neutrality from a regime perspective.

Another important point in the book is that people who the authors call “tweakers”—not the drug addicts, the other kind--are often directly responsible for refining a leap by a pioneer and making it much more productive/valuable, and this both incentivizes tweakers and leads to big debates over credit allocation.  Copyright and patent don’t favor tweaking because of the control they give patentees and copyright owners over follow-on innovation/derivative works.  Compare this with how fan creators in noncommercial spaces solve the problem: they don’t generally make formal IP claims or contest ownership of the core products—though they sometimes do label themselves better custodians of the spirit and meaning of a franchise, like Star Wars, than the formal copyright owners.  But they do follow robust norms of attribution and often even permission within their own communities, thus trying to manage disputes in a space where only law’s shadow reaches.  Fan authors are recognized by other fans as entitled to credit for their particular contributions, though these norms are often contested and changing as new fans come in and new technologies change the means by which fan productions are disseminated. 

One regularly encounters attempts to formalize norms in some way—in the softer sense through best practices statements such as those put out by AU’s Center for Social Media, and in a harder sense through proposals to recognize certain sorts of norms in the law.  Having just reviewed Demsetz for my property class, I kept thinking of the ways in which the formation of explicit property allocations through law can be slowed, or perhaps more importantly changed, by struggles between different groups based on their beliefs about proper allocations, in other words, based on their expressed norms.  Special rights for session musicians in the new digital sound recording performance right is an example of encoding norms about fair compensation into law. 

But when only certain groups show up to the formal lawmaking process, of course, their interests are unlikely to be represented.  The case studies the book illuminate how poorly copyright law’s standard incentive theory describes a large number of creative endeavors.  Why is that?  One key insight is that, in Madhavi Sunder’s words, identity politics interact with intellectual property concepts: intellectual creations regularly have as much to do with the first kind of “IP” as they do with the second.  Put differently, authors understand their works to express and form an identity that is both unique to themselves and part of a larger culture.  This cultural context both incentivizes creativity even in the absence of conventional IP ownership and shapes the content of what gets created. 

Intellectual property, like property generally, allows some claims to power and authority and deauthorizes others.  I’m going to talk here about comedians: modern comics think of their jokes as being intrinsically tied to their own identities, often invoking specifics of race, gender, sexuality, and class.  Oliar and Sprigman, in an argument adapted by the authors here, argue that comics have achieved a sort of self-help propertization of their jokes by moving to identity-specific routines that are not as easy to appropriate.  But not only are the enforcement mechanisms tribal and more effective for men willing to threaten physical violence, but this definition of comedy depends on valuing certain kinds of comedy and preventing certain kinds of copying—as with chefs, comics don’t seem to mind if individuals retell jokes to each other, like my husband retelling a certain hilarious Louis C.K. line.  Norms tend to be directed at other group members and uninterested in the behavior of nonmembers, not just because of the inability to control nonmembers but because a nonmember’s activity doesn’t implicate the same kind of issues of respect and acknowledgement.

Even where norms govern instead of law, there is always a question of whose claims count and whose claims, though an alien might think them similar, are not even recognized as such.  The norm comedians articulate is that it’s not permissible to deliver material that isn’t theirs.  But that’s not, in point of fact, true.  (I’m not questioning their research nor the sincerity of their sources: the point is that their sources believed something that was self-evidently untrue.  This is always a signal that something very interesting—usually something about status—is going on.)  In fact, comedians often provide material for other comedians.

The authors give us two common situations of this type.  First, comedians may help others create jokes.  There is no joint authorship norm.  People who provide parts of jokes to comics have given a gift.  And as Carol Rose and Lewis Hyde among others have noted, a gift can also be a burden, because it creates an obligation for the recipient.  The anxiety generated by a gift can be managed by reciprocal norms: the contributor will perhaps someday receive assistance in creating his own jokes, and also be able to claim complete ownership of those jokes.  The contributor’s creativity and labor when she helps out, however, are not her own.  They disappear into the primary comedian’s joke.

Second, comedians may use jokes created by other people, who submit them and get paid if their jokes are used.  Like the singer/songwriter model in music, the stand-up character can be as constructed as any boy band.  Now, the authors maintain that the practice of using writers is much less common than it used to be for comedians, and relatedly that comedians today invest relatively less in performative aspects (costume, movement, props) and relatively more in individualized schtick.  One reason comedians invest more in writing their own jokes is, they suggest, that it is naturally harder to write for another person with a unique persona than it is to write generic jokes.  This extra difficulty can be expected to raise the relative price of buying modern, identity-specific jokes.

Yet buying jokes is apparently a perfectly legitimate means of becoming their exclusive owner even among stand-up comedians.  Writers can’t claim the jokes they wrote even in seeking to prove their comic credentials. Some creative work counts in producing property claims; other creative work doesn’t.  It’s about power.

There’s more: all this depended on a particular definition of “comedian.”  There’s actually a gap between what The Knockoff Economy calls “rival creators” and the group known as “consumers.”  Expand the frame, and a property norm may only apply to a subset of broadly similar activities—a truly local ownership norm.  This matters, among other things, because copying may work very differently in the other subsets, most obviously in advertising where becoming a political slogan (“Where’s the Beef?”) can be confirmation of fame and value.  Sitcoms have extensive writing staffs devoted to producing jokes for hire, as do The Daily Show and The Colbert Report, customized to their public personae.  I therefore have doubts about the proposition that writing for another specific person is more difficult than writing generic jokes, and thus relatively more expensive.  Even accepting that this is harder to do than to write for oneself, the supply of aspiring writers outstrips demand so greatly that it seems unlikely that writers can demand much of a premium for writing in a different voice.  Indeed, the assumption that writing for a specific character is especially difficult may itself be an effect of the ideology of romantic authorship in which genius is individual, unique and nontransferable. 

I greatly enjoyed the book and its examples; my comments are designed to point out that when we look at creative situations, it never hurts to keep in mind Lenin’s question about the meaning of history, translated as “who does what to whom?”

Fagundes: maybe slowing the fashion cycle would get us better clothes.  Scorecasting: football has it all wrong, and there’s a better strategy to be adopted.  But people haven’t done it because if you tried this and lost even once you’d be fired: fear/other forces constrain and affect innovation.  The team that did the most innovating in pro football was the worst, the Cincinnati Bengals—being bad forced them to innovate.

Nicole Stelle Garnett: is this innovation or churning?  As a land use person, her question was whether we were getting more or better?  Not just laws that prevent copying but those that require certain aesthetic taste—not uncommon in real property, where codes often require old-looking houses because architects have decided that there was a certain perfect point in time.  Having nice houses will make us better people!  Miami, Denver, El Paso have adopted these New Urbanist codes focused on aesthetic choices, freezing things in time.  In art, they favor crazy/new aesthetics, but in land use they may require copying!

McKenna: we will always get innovation; the inescapable question is what kind. The legal rules can shape/force a path. There’s no such thing as neutrality in the legal rule, and it’s better to make that choice on the surface.  This also matters to rights v. utilitarianism.  We are picking winners: downloading has been bad for some players in the music industry and good for others; we have to decide what matters.

Sprigman: Defense of more v. better: deepest fountainhead of innovation is natural selection. Just doesn’t know what beauty might look like in 20 years.

Przybyszewski: Technological fabric innovations: Gore-tex etc.—these are clearly better than what preceded them within their segment, not just different.

Yi Qian: Theory: competitors lead innovators to innovate more in searchable qualities—surface materials of shoes, workmanship; experiential/functional attributes that take more time to discover would get less innovation—durability—even if quality of those attributes would stay the same.  Confirmed by Chinese data.  Pharma: see innovation more on shape and color. Music: innovation on packaging.  Welfare implications may differ by industry. In fashion, where people derive benefit from appearance, this may be all good, but maybe not so much in pharma.

Samuelson: agrees with Beebe about the slippage between creativity and innovation.  Is there also slippage between innovation and improvement?  Is innovation just newness?  Patent = novelty.  But new and inventive may not be better; patent is agnostic about improvement.  Though we expect that people won’t take the time to patent things that aren’t improvement, if you actually look at patents that’s not true at all.  Improvement innovation can be measured on the tech side only sometimes.  Rube Goldberg patents.  Copyright side: we can judge whether a movie is good, but it’s just much harder.

Sprigman: patents may be improvements or not depending on the circumstances: a water pump good for an undeveloped area might be a terrible idea for South Bend.  Fashion: some years are good and some aren’t; he thinks we tend to undervalue change.

Raustiala: some people push back on fashion churn: isn’t it bad for social/environmental reasons?  He thinks it is kind of bad.  Book takes a catholic view of saying that we don’t know what counts as good; we aren’t social planners so we said more is better and elided innovation and creativity.

Michael Heller: the dichotomy real/fake also has a spectrum—there’s a world of fake Rolex connoisseurs.  What’s the implication of the specturm of fakes?  Reinforces Gosline’s work.

Also, thinks Przybyszewski was too pessimistic about Hawes’ message.  Her reputation allowed her to make money.  And most of us don’t make money on our books! They’re collateral effects of other things we do.

Beebe: Evolution as a model of innovation is provocative.  Darwinian view: there was no progress in evolution—that’s what was so shocking about it!  If we’re just meandering along, what’s the point?

Sprigman: survival.

Lea Shaver: acting and music are like fashion: just as hard to make it there even with copyright protection!  (I wonder what the relevance of the “cost disease” is here.)  As for Gosline: are these insights only applicable to showoff goods?  Recounts her experience of starting with cheap smartphone and converting up to the iPhone after learning more about smartphones/getting sold on the features. 

Innovation: can we say a field has advanced in 20 years?  Easy to say in computing.  Can we say that comedy is improved from 20 years ago?  Football seems to have improved in sophistication.  What about fashion?  [What about teaching?]  If you looked quantitatively and qualitatively in people’s clients, they can afford more of what they like—a big part of innovation is making stuff cheaper.  Thus inclined to the “more” is better thesis.

Gosline: “gateway drug” thesis has come up before as a phrase people use to describe their experiences.  More likely in the case of signalling goods because the social reaction helps to make the comparative aspect more fertile. Even with nonsymbolic, functional goods, it can still occur if your experience with the cheap version is suboptimal.  Even things that aren’t seen can make people feel a certain way about themselves—buy Tide to show they care about their families.

In terms of what’s real or fake, the binary is problematic, which is why her research looks at that—but she emphasizes that people who’ve paid for the real thing also get called fakes.  Some of the judgments on the FB page were wrong; people were deemed fakes even if they owned the real thing.

Christopher Buccafusco: Surprised at willingness to accept anything goes, given that preferences are (Sprigman has argued!) shaped.  Systems are embedded in systems.  Football: internally, wins and losses are equal/zero-sum in football so it’s tough to talk about progress.  Individually, progress is possible.  Externally, you can expand the size of football’s pie.  In baseball, steroid use expanded the audience for baseball, which was good for baseball!  Within and across fields there will be different kinds of welfare.

Sprigman: true, taking preferences as given here.  Consistency is overrated (unlike change).  His experiments look at preferences structured by law; preferences structured by nonlaw sources may be resilient. Not terribly interested in debiasing here though confident that their preferences are not exogenous.

David Opderbeck: we’re all circling around teleology/progress.  Darwin upset the watchmaker universe and showed that evolution produces waste.  Using that metaphor might be helpful.

Avishalom Tor: in competition, we don’t question that consumer demand drives it all.  If we give up that idea, our structures of analysis become unstable (which might be the right result).

Pom part two: remedies, First Amendment, concurrences

Pom argued that the First Amendment protected its ads.  But Central Hudson gives commercial speech constitutional protection only when it’s not misleading.  To get around this, Pom argued that the FTC couldn’t determine that ads are “actually misleading” without extrinsic or empirical evidence of deception.  And, Pom continued, it couldn’t find ads “inherently misleading” if they stated accurate and verifiable facts.  Thus, at most, the FTC could require limited disclaimers as to “potentially misleading” ads, because Pom and the FTC just disagreed about the meaning of scientific evidence (or, you know, its presence).  (Despite the dumbness of Pom’s contention that literal truth can’t be inherently misleading, the doctrine here is indeed a mare’s nest, since the Supreme Court has tossed around actually/inherently/potentially misleading without saying much useful about what those terms ought to mean.)

No dice.  Precedent establishes that empirical or extrinsic evidence isn’t necessary for the Commission to determine that ads are actually misleading, based on its own expertise.  The ads claimed that Pom products had been proven to prevent/treat diseases.  This is not true.  That’s deceptive and actually misleading, as is failure to substantiate claims. Zauderer didn’t require empirical evidence; Kraft didn’t; and the FTC isn’t going to either, even when a claim is made by implication, as long as the claim is reasonably clear.  Misleading commercial speech is not protected by the First Amendment, full stop, no further Central Hudson analysis required.

Even so, the Commission addressed Pom’s argument that “an advertisement cannot be inherently misleading on its face when it states objectively accurate and verifiable facts.”  Not so.  Literally true statements can have misleading implications.  Language games with “inherently” don’t change that (or we’d just talk about “false” claims, not “misleading” ones as well).  Pom wanted the Commission to evaluate each element of an ad in isolation and find that no individual statement was itself false/misleading, but that wasn’t the law.  The cases Pom cited involved bans on simple truthful statements in professional advertising, such as bans on listing jurisdictions where an attorney was licensed or certifications s/he held.  But in those cases, the Court found, for example, that consumers could easily confirm the licensing or certification.  Pom argued that their claims were objectively accurate and verifiable: their products are high in antioxidants; they cited their studies; the results were published in peer-reviewed journals.

However, though many of the individual facts in the ads were verifiable, there were many omissions that consumers couldn’t verify independently.  As an example, consumers couldn’t verify that one of the five studies cited in the ads was rejected by the American Heart Association and the Journal of the American Medical Association because of research shortcomings, and was only accepted for publication without peer review. Nor could consumers verify that the results of a much larger, well-designed, well-controlled study were inconsistent with Pom’s arterial plaque claims because Pom suppressed publication of the negative results for a significant period.  Many of these representatiosn were qualitatively different from the verifiable statements in the professional advertising cases.

Pom’s fallback argument was that its ads were only potentially misleading under the terrible Pearson v. Shalala, 164 F.3d 650 (D.C. Cir. 1999), and couldn’t be “banned on the basis of a genuine dispute about the level or meaning of scientific evidence.”  The Commission didn’t read Pearson to prevent a finding that Pom’s claims were misleading for want of substantiation, even if that were based on a finding that the ads were potentially misleading.  Pearson involved a prospective ban on certain supplement claims; the DC Circuit (well-known for its expertise in marketing and consumer perception) concluded that sufficiently prominent disclaimers could avoid misleadingness.  By contrast, the Commission reviewed the ads that Pom actually ran.  Its disclosures and disclaimers were much less than “what the D.C. Circuit hypothesized would be sufficient to prevent health claims with disputed scientific support from being misleading.”

Commissioner Ohlhausen disagreed with regard to some ads that she thought had sufficiently qualifying language.  She would have required extrinsic evidence before finding them misleading.  If Pom had used Pearson disclaimers such as “the evidence in support of this claim is inconclusive,” the Commission would have considered the ad as a whole, including those disclaimers.  But without disclaimers, the ads were deceptive and misleading.  And a case-by-case approach differs from regulations that prohibit categories of speech or preapprove language.  Pom was simply being required to have adequate substantiation for the claims id did make.

Pom brought a new Fifth Amendment due process challenge, claiming that the Commission was retroactively requiring a new evidentiary standard.  But the general rule on substantiation is well established, and how that applies to health claims has also been detailed before.  Given that what counts as substantiation is always fact-based depending on what’s competent and reliable scientific evidence for the claims at issue, there was no due process problem.

The Commission decided to ignore the four media interviews challenged by Complaint Counsel: appearances by Mrs. Resnick on The Martha Stewart Show and The Early Show, sharing recipes and marketing ideas related in part to POM; a magazine interview with Mrs. Resnick in Newsweek,in part promoting the sale of her book about the POM business; and a television interview with Mr. Tupper on FOX Business discussing the current relevance of the pomegranate and pomegranate juice.  The ALJ found that they weren’t ads, but the Commission didn’t adopt that ruling.  The FTC’s jurisdiction covers deceptive commercial speech, not just ads—the FTC’s not NAD.  Instead, there just wasn’t enough evidence of the circumstances surrounding the interviews to determine whether they were commercial speech (which would require an analysis of the content, including whether it referred to a specific product or product attributes; the means used to publish the speech, including whether it was paid for; and the speaker’s economic motivation/whether the interviewer or the interviewee controlled the questions).  Given the numerous other deceptive claims made by Pom, the Commission exercised its discretion to avoid basing liability on those media interviews.

Pom then argued that no cease and desist order was warranted because it stopped running the ads at issue, and because it adopted a new review process in 2006 that got rid of most of the offending claims.  The Commission disagreed.  Relevant considerations are “(1) the seriousness and deliberateness of the violation; (2) the ease with which the violative claim may be transferred to other products; and (3) whether the respondent has a history of prior violations.”  The strength of one factor can outweigh weaknesses in others.  Here, though Pom had no history of violations, they were serious and deliberate, and the type of health claims could be easily transferred to the Pom entities’ other products.  Pom made numerous unsupported claims about serious and even life-threatening diseases.  It did so with gusto and awareness of the weakness of its claims, including inconsistency between its claims and the results of some of its later (better) studies.  And the Pom entities “could use similar marketing techniques to make disease claims about other food products, including the other food products Respondents currently sell.”

Given the ready transferability of the technique, and the fact that the Pom entities had already sponsored research with an eye toward making health claims for other products they sell such as Wonderful Pistachios and FIJI Water, the Commission imposed a fencing-in provision, applying its order to the challenged Pom products along with any other food, drug, or dietary supplements sold by Pom and the related entities identified by the FTC. 

Moreover, the order required at least two RCTs before Pom could make “any representation regarding a product’s effectiveness in the diagnosis, treatment, or prevention of any disease.”  The Commission specified that two RCTs would be required, consistent with FTC precedent and with expert testimony about the need for consistent results in independently replicated studies.  Even with a RCT, results from any one study may be due to chance or may not be generalizable due to the uniqueness of the study sample.  Separately, Pom had a demonstrated propensity to misrepresent research in its favor.  Pom engaged in “a deliberate and consistent course of conduct – no mere isolated incident or mistake.”  A two RCT requirement would constrain its ability to repeat this misrepresentation.  The Commission declined to require FDA preapproval; the need for a clear standard to which to hold Pom was adequately met by the two-RCT requirement.  In a footnote, the Commission stated that it didn’t foreclose FDA preapproval as an appropriate remedy in another case.

The two-RCT requirement only applied to disease prevention, risk reduction, and treatment claims. Other health-related claims would require competent and reliable scientific evidence, sufficient in quality and quantity when considered in the light of the entire body of relevant and reliable scientific evidence, to substantiate them.

Commissioner Ohlhausen would only have required one RCT, on the theory that requiring two wasn’t reasonably related to the violations at issue “because a second study would not cure any particular statistical or methodological problems” [have some xkcd!] and because the requirement could chill consumer access to potentially useful information about an admittedly safe product, raising First Amendment concerns.  Following Pearson, commercial speech doctrine prefers disclosure over suppression.

Commissioner Rosch concurred, though he thought the majority took the “safest” route in its discussion of the expert testimony and the Pfizer factors rather than the most sensible one.  It was unnecessary for the Commission to rely so heavily on the opinions of paid experts.  He would have decided that the net impression from the ads included claims about the level of substantiation Pom possessed, conveyed both expressly and by implication. These claims were false.  The Commission’s own common sense and expertise can determine whether false substantiation claims are conveyed, without extrinsic evidence.  “[W]hen an advertiser represents in its ad that there is a particular level of support for a claim, the absence of that support makes the claim false.”

Commissioner Rosch agreed that if Pom’s ads just made health claims, standing alone, they couldn’t be challenged as false or deceptive.  But they didn’t stand alone.  Sometimes they were linked to treatment/prevention claims for heart disease or prostate cancer, either explicitly or implicitly, creating a false net impression that “the highest possible level of substantiation exists for the POM product being advertised.”  Implicit claims were made by “liberally quoting physicians, by equating POM with POMx (which is depicted as a prescription drug), or by depicting POM itself as a medicine,” as well as by “describing POM as a life insurance supplement or a healthcare plan.”  The Commissioner didn’t consider erectile dysfunction to be as serious as heart disease or prostate cancer, so linking Pom with treatment/prevention of ED standing alone didn’t create a net impression that these claims were supported by the highest level of substantiation.  But there was no error in requiring such substantiation, since establishment claims are binding on advertisers who make them, and Pom did make such claims.

Commissioner Ohlhausen’s concurrence disagreed with the majority about certain ads, which she read as making claims only about the “general effects of the POM products on the continued healthy functioning of the body,” not about diseases.  The majority found disease claims implied without extrinsic evidence.  She worried that too easily finding implied disease or establishment claims, withouth extrinsic evidnece, would “undermine an important balance that is struck in the regulation of food, supplement, and drug advertising under the FTC Act and other federal laws”—that is, the distinction between structure/function claims and disease claims. 

She also thought that the majority conflated disease treatment claims with prevention/risk reduction claims, including by finding implied prevention claims for ads describing studies of subjects already suffering from prostate cancer or ED. “Because it seems unlikely that a consumer would assume that any food or food product that lowers the risk of disease is also a viable treatment for that disease, I disagree with the majority’s conclusions that such claims are facially present in certain exhibits.”  Likewise, she didn’t think that consumers would assume that treatments for existing cancer or heart disease would prevent the onset of those conditions.  Also, she thought the use of qualifiers such as “small study,” “initial scientific research,” and “promising,” “hopeful” or “encouraging” results didn’t make establishment claims.  Sure, Pom referred to millions of dollars spent on research, but it was referring mostly to how much it spent showing the amount of antioxidants in the Pom products and the general effects of antioxidants on the human body; extrinsic evidence was needed to show that consumers would think it meant more than that. (I think this all really ignores the forest for the trees, given these ads.) 

The majority’s ruling blurred the boundaries between structure/function and disease claims, creating an FTC/FDA inconsistency.  “[I]t is difficult to imagine any structure/function claims that POM could associate with its products in the marketplace without such claims being interpreted, under the FTC precedent set in this case, as disease-related claims…. [W]e should be careful not to interpret claims so broadly that we undermine distinctions between types of claims, and the substantiation appropriate to them, that Congress and our sister agency have found important to the public’s health and wellbeing.”  [The FDA has not actually so concluded; it’s just been told by Congress and the DC Circuit to pretend that there’s a difference to consumers for its regulatory purposes.  I applaud the majority of the Commission for focusing on its consumer protection mission; if consumers aren’t in fact distinguishing so-called structure/function claims focusing on the absence of specific diseases from disease prevention/treatment claims, as the studies I’ve seen show they aren’t, then they can be deceived by supposed structure/function claims that implicitly or by necessary implication serve as disease claims.]