ReDigi and the Path Forward for Digital Exhaustion

In December—more than a year after hearing oral arguments—the Second Circuit finally issued its opinion in Capitol Records v. ReDigi. The decision, written by Judge Pierre Leval, affirmed the district court’s holding that ReDigi infringed Capitol’s copyrights by operating a secondary market for used digital music files. Undoubtedly, the Second Circuit delivered a serious blow to the notion of digital exhaustion. ReDigi was, after all, the first U.S. test case for extending the exhaustion principle to digital goods. But in this post, I want to push back against a broad reading of ReDigi by first critiquing the Second Circuit’s analysis and second, by outlining a path forward hinted at by the decision’s conspicuously circumscribed reasoning. 

First, a refresher: Capitol Records sued ReDigi more than 5 years ago after ReDigi launched its online secondary market for used iTunes songs. ReDigi built its platform to recreate the mechanics of a traditional analog sale. Under its model, one party starts with a single copy of an iTunes track. At the end of the transaction, ownership of that track is transferred to another party who now owns a single copy. Crucially, the seller is left with no remaining copies. ReDigi designed its system—in its terms— to migrate files from the seller’s computer to its servers. It did this by deleting the original file, packet by packet, as it was uploaded. So at any given time, at most one full copy existed. After the district court granted summary judgment in Capitol’s favor, ReDigi filed for bankruptcy, significantly delaying its appeal to the Second Circuit. 

The Second Circuit’s opinion is an undeniable win for Capitol and other copyright holders concerned with potential competition from resale markets. The opinion is disappointing in a number of entirely predictable respects. But it also reflects an effort to tread lightly, opting to steer clear of some important legal questions and hewing its analysis closely to the facts of this particular dispute. As a result, it preserves a path for future litigants to navigate, even if it is neither wide nor clearly demarcated.

Capitol alleged that ReDigi both reproduced and distributed copies of its sound recordings. On appeal, ReDigi relied on two primary defenses—the first sale doctrine and fair use.

First Sale

Let’s start with first sale. When it came to reproduction, the Second Circuit agreed with the district court that ReDigi was in fact reproducing Capitol’s works, and that those acts of reproduction were not covered by § 109(a)—the statutory first sale doctrine—which is limited on its face to the distribution right.

Despite some caselaw suggesting that the transfer of a work to a new medium that simultaneously destroys the original copy is not a reproduction, the court’s conclusion that the copy stored on ReDigi’s servers is a new copy is consistent with the dominant view of reproduction. But from my perspective, that conclusion underscores the fact that a copyright system built around a careful accounting of copies is poorly positioned to track user experience or the economic value of works. From the user’s perspective, what matters at the end of the day is who has access to the work, not the precise mechanism by which copies are transferred. As the Supreme Court recently indicated in ABC v. Aereo, copyright law should law should be mindful not only of the precise language of the statute, but also the experience of the end users of technology.

Even if reproduction occurred, we argued as amici that the court should focus on the broader common law of exhaustion, not just its more limited statutory component. Both before and after the first sale doctrine was embodied in the Copyright Act of 1909, courts have applied exhaustion to permit reproduction when necessary to facilitate the transfer of a copy. But the Second Circuit viewed the enactment of § 109(a) and its predecessor provisions in 1909 Act as an effort to narrow the common law of exhaustion. Based on the doctrine’s development and the legislative history, we think the Second Circuit is mistaken.

As to distribution, ReDigi argued that the digital files it transferred were phonorecords subject to the first sale doctrine. But since the Second Circuit affirmed the district court’s reproduction analysis, it didn’t reach the distribution argument. The court likely understood its avoidance of this question as an exercise in judicial restraint, preserving the issue for future litigation. 

But avoiding the question has at least two negative consequences here. First, the court had the opportunity to clarify whether the transmission of digital files involves copies or phonorecords that violate the distribution right. London Sire and other cases suggest that such transmissions are copies for infringement purposes. But the ReDigi district court said they weren’t for first sale purposes. The court could have resolved that tension.

Second, the court created significant ambiguity when it comes to the ability of owners of digital copies to exercise their first sale rights. The court plainly stated that the owner of a digital copy is entitled to distribute it, more than hinting at some disagreement with the district court. And in two crucial footnotes, it suggested that making reproductions for space shifting and cloud storage is likely a fair use. The court even outlined what it saw as a lawful resale market for digital goods: “A secondary market can readily be imagined for first purchasers who cost‐effectively place 50 or 100 (or more) songs on an inexpensive device such as a thumb drive and sell it.”

But the Second Circuit doesn’t provide any clear basis for distinguishing the copies stored on this imagined thumb drive from those stored on ReDigi’s servers. Both are reproductions, but the court assumed the thumb drive copies are lawful while the server copies are infringing. Why? Is it the lack of an intermediary that makes the difference? Or perhaps the commercial intent of that intermediary? Is it the scale of the operation? Or the relatively frictionless nature of network-mediated transfer? The court leaves those questions unanswered.

Fair use

Turning to fair use, even under Judge Leval’s expansive view, the acts of reproduction here were not transformative. They don’t change the expressive content of the underlying works, and they don’t use those works for any new purpose. Perhaps not surprisingly for an opinion by Judge Leval, the court’s analysis is single-mindedly focused on transformation. 

But the proper inquiry into the purpose and character of the use does not begin and end with transformation. ReDigi’s acts of reproduction serve a different purpose; they are intended to facilitate otherwise lawful transfers of copies that consumers own. The court tells us you can sell your thumb drive, even though that entails making a copy. Once we’ve agreed, as the court seems to do, that making copies for transfer is fair, how do we distinguish the copies made here? The court’s reasoning is murky at best and self-contradictory at worst.

On the question of market harm under the fourth factor, the court tells us that ReDigi made reproductions for the purpose of competing with plaintiffs in the resale of their sound recordings. That is absolutely true. And it sounds like a damning fact, until you remember that is precisely the form of market competition that first sale has traditionally embraced. First sale is designed to create competition between new copies and used ones. In doing so, it increases access to valuable works, puts downward price pressure on new copies, and increases market efficiency. The court’s analysis under factor four is perhaps the weakest and most cursory of its arguments. It could have argued that the difference between physical and digital copies justifies a rule restricting digital transfers. Such an argument would overlook the impact of shifting file formats and degradation of digital data, but the court didn’t even bother to make it. Or the court may have suggested that the speed and efficiency of digital file transfer alters the economics of resale in some fundamental respect. Given the practical hurdles imposed by a resale market like ReDigi’s, that argument would be difficult to support as a factual matter. But the court didn’t even bother. But the perfunctory nature of the analysis may help to distinguish ReDigi in future cases.

What’s next?

Despite what may have been its effort to tread lightly, the Second Circuit has seemingly foreclosed on many of the most promising arguments for digital exhaustion. So where does this leave us? Is digital exhaustion dead?

In the Spotify era, the idea of reselling digital music may seem quaint, if not antiquated. But the music distribution market is likely to continue to undergo major changes in the coming decades, and the viability of digital resale will help shape those developments. Putting music aside, digital exhaustion is likely to have important implications in a range of markets for a variety works. So the ReDigi decision will have potentially widespread reverberations.

Although not a legally binding holding, the ReDigi opinion strongly suggests that resale, lending, or giving away digital goods is lawful at the individual level. If you want to sell your digital music collection to a neighbor by handing over your external hard drive, you are free to do so. Even assuming that’s true—and given the low likelihood of litigation over hand-to-hand resale, assumption will likely have to suffice—can lawful digital exhaustion operate at scale?

One strategy is to distinguish future platforms from ReDigi’s. The Second Circuit held open the possibility that other technologies, perhaps even ReDigi’s own version 2.0 software, might avoid the pitfalls of reproduction. If so, the first sale doctrine may yet apply to digital distribution. 

But even if reproduction is unavoidable, defendants could still distinguish themselves and their activities as a matter of fair use. What might the ideal defendant look like in a fair use case testing the bounds of digital transfer? A non-profit actor enabling non-commercial transactions defined by clear and enforceable practical limits would seem to be in the best position to take advantage of fair-use-enabled digital transfer at scale. The commercial nature of ReDigi’s activities seemed to color the Second Circuit’s view of its platform. And concerns about the impact of frictionless transfer of digital goods would inform a more thorough analysis under factor four.

Controlled digital lending platforms—like the one the Internet Archive operates and those UC Berkeley and other institutions are exploring—are distinguishable from ReDigi in precisely these respects. Under these programs, libraries scan their paper collections and offer the resulting digital copies for lending on the same terms as physical books, and in the same quantity. Unlike ReDigi, these non-profit institutions do not stand to profit from the digital books they lend. And by restricting the number of books borrowed and the terms of lending to those otherwise permitted by the libraries physical holdings, these programs impose meaningful limits on the speed and scope of digital transactions that guard against unforeseen disruption of the copyright holder’s market.

Beyond distinguishing ReDigi in future cases, the Second Circuit’s opinion suggests digital exhaustion is a question best-suited for Congress, given its ability to consider the impact of such a policy on a range of stakeholders. That may be true, but digital exhaustion is unlikely to be a priority for a Congress mired in shutdown negotiations and various Trump-related investigations for the foreseeable future. Somewhat more promisingly, Judge Leval opined at the ReDigi oral argument that the Supreme Court would likely decide ReDigi’s ultimate fate. Whether ReDigi is eager to continue this fight—and whether the Court would take the case—remains to be determined. But it is worth noting that the Court has heard an outsized number of exhaustion cases in recent years: Quanta v. LG, Costco v. Omega, Kirtsaeng v. John Wiley & Sons, Bowman v. Monsanto, and Impression Products v. Lexmark.

Finally, outside of the United States, the question of digital exhaustion is making its way through various European courts. Of particular note, the European Court of Justice will determine whether the resale of ebooks is permitted under the Copyright Directive in the Tom Kabinet case. Which, if any, of these paths to digital exhaustion is viable remains to be seen. But the fight over digital exhaustion is unlikely to be resolved anytime soon.

This post is part of Copyright Week

Lexmark and the future of sales

In Chapter 10 of our book—spoiler alert—we urged the Supreme Court to reverse the Federal Circuit's disaster of an opinion in Lexmark v. Impression Products. We certainly aren't making any causal claims here, but shortly thereafter the Court decided to grant cert, and earlier this week it issued its opinion.

The case involved the long-running effort by Lexmark to prevent the refilling of toner cartridges for its printers. As part of that strategy, Lexmark sold patented cartridges with restrictions attached: you can only use them once, and you can't transfer them to anyone else. Anyone who broke those rules, according to Lexmark, was a patent infringer. But those limits run counter to the doctrine of patent exhaustion, which ensures owners of patented goods the right to use and alienate them free from patent-holder control. The first question for the Court, then, was whether these sorts of post-sale restrictions are enforceable as a matter of patent law. The second question was whether exhaustion applies at all to products that were first sold outside of the United States. 

As to the first question, the Federal Circuit held that the scope of exhaustion is a matter of statutory interpretation. It saw exhaustion as a sort of implied license; buyers are granted the rights to use and alienate goods only so long as the patent holder does not indicate an intent to withhold them. The Federal Circuit tied that reading to the phrase "without authority" in the Patent Act's infringement provision.

The Supreme Court rejected the notion of exhaustion as a purely, or even primarily, statutory creature. Instead, it explained that exhaustion "marks the point where patent rights yield to the common law principle against restraints on alienation." The rights that flow from exhaustion are not the patent holder's to give or withhold. Instead, they are inherent limits on the scope of the patent. When a sale occurs, it "transfers the right to use, sell, or import because those are the rights that come along with ownership." In concluding that a patent holder cannot sell a product while retaining ongoing rights to control its use or disposition, the Supreme Court rejected twenty-five years of ill-considered Federal Circuit precedent, stretching back to Mallinckrodt v. Medipart.

As to the second question, Lexmark argued, and the Federal Circuit agreed, that sales made outside of the United States did not exhaust the patent holder's control over use and alienation. In 2001 in Jazz Photo, the Federal Circuit held that foreign sales did not exhaust U.S. patent rights. That decision relied largely on misrepresenting the facts and holding of the Supreme Court's only other international patent exhaustion case, Boesch v. Graff, from 1890. Crucially, that case did not involve a sale by the U.S. patent owner, rather the sales were made by a prior user entitled to make the product under German law. Boesch, the Court rightly noted, tells us nothing about whether a foreign sale made or authorized by the patent holder triggers exhaustion. Instead, as it did in Kirtsaeng v. John Wiley & Sons, the Court recognized that "the common law’s refusal to permit restraints on the alienation of chattels," the principle animating both patent and copyright exhaustion, "makes no geographical distinctions.”

As the Court succinctly put it, "restrictions and location are irrelevant; what matters is the patentee’s decision to make a sale." So in one opinion, the Supreme Court upended two pillars of Federal Circuit patent exhaustion case law. What does that mean going forward?

The immediate upshot of the decision is that consumers will have more freedom to use the products they buy in the way they see fit. They will also benefit from greater competition and lower prices as remanufactures like Impression Products enter the market without the risk of patent infringement suits.

Some worry that the loss of control patent holders have enjoyed over the past few decades will lead to changes that will instead harm consumers. Won't companies like Lexmark just raise their prices on new toner cartridges? Perhaps, but I think it is doubtful for a couple of reasons. First, Lexmark already sold unrestricted cartridges, for only about 20% more than the single-use variety. Second, raising the price on new cartridges may well steer more consumers to cheaper remanufactured alternatives, at least in the short term.

Another potential concern is that companies like Lexmark will stop selling products altogether and move to lease, rental, or subscription models that don't entail transfers of ownership to consumers. We've seen the software and digital media markets embrace these models in recent years. But aside from the legal benefits of avoiding sales, those models benefitted from the economics of digital distribution. At least until 3D printing becomes a widespread reality, the distribution of most patented goods will remain decidedly tangible. If so, leasing models won't be very appealing, especially for low-dollar-value goods. Patent holders would need to create and enforce systems for collecting ongoing payments and expired products from consumers, an expense likely not justified by the harm of secondary markets. Of course, exactly how much of an expense such a model would entail depends on how courts define a "sale" going forward.

The Court in Lexmark didn't directly tackle the question of what exactly constitutes a sale. In the copyright space, software companies have been able to convince some courts that simply dubbing a transaction a "license"—even when it is a one-time payment in exchange for perpetual possession—is enough to avoid a sale and exhaustion. If trivial changes to the wording of Lexmark's labels—essentially replacing "post-sale restriction" with "license"—were sufficient, the Court's opinion won't have much lasting effect. 

So what can we glean from the Court's opinion when it comes to sales? For one we know that a transaction still counts as a sale if the seller attempts to to attach restrictions. And we know it is still a sale if the seller explicitly reserves its patent rights. More precisely, we know that restrictions on alienation and use of the patented good are not enough to overcome a sale. That's important because the Ninth Circuit's deeply-flawed approach in Vernor v. Autodesk looks to three factors to determine whether a "license" or a sale has occurred. They are whether the copyright holder:

 (1) specifies that the user is granted a license;

(2) significantly restricts the user's ability to transfer the software; and

(3) imposes notable use restrictions.

Lexmark's restrictions applied to the use of the cartridges and to their transfer. In fact, the only thing that sets Lexmark's restrictions apart from Autodesk's is the use of the magic word "license." Unless the Supreme Court's full-throated embrace of patent exhaustion was nothing more than an exercise in empty formalism, Lexmark seems to suggest that the Vernor test is misguided. If that's right, Lexmark may actually help clarify the longstanding murkiness in copyright law over the license/sale distinction.

"Digital is cheaper" & other bogus arguments

A couple of weeks back, I took part in the USPTO & NTIA's Public Meeting on Consumer Messaging in Connection With Online Transactions Involving Copyrighted Works. The idea behind the event was to bring together academics, consumer advocates, and industry to discuss the problem of inconsistent consumer messaging when it comes to transactions for digital goods. I was there because my work with Chris Hoofnagle is the only published study of how consumers understand advertising language like the Buy Now button, which we show leads many consumers to incorrectly believe that "buying" digital goods gives them the same rights to lend, use, give away, and resell that they are accustomed to for tangible goods.

It was great to hear the range of perspectives from folks like BJ Ard, Lorrie Cranor, Bob Gomulkiewicz, and Florencia Marotta-Wurgler—among many others—who brought their expertise to bear on this issue. But the absence of Amazon and Apple, the two companies at the center of this problem, was notable. As the leading digital retailers, they are best positioned to improve consumer understanding of digital transactions. Despite what I'm sure were persistent efforts by the PTO and NTIA, those companies couldn't bother to send a single member of their legal, policy, or design teams. Nor did any major copyright holders take part. 

Instead, as the agenda reflects, industry was represented by the usual DC trade associations and lobbyists. Rather than contribute to a productive conversation about this issue, they spent the day denying there was anything to worry about and making their best efforts to scare the organizers off with vague and occasionally nonsensical allusions to antitrust and the first amendment.

There were also a few efforts to cast doubt on our Buy Now research. But thanks to some rather blatant filibustering, I wasn't given any meaningful opportunity to respond. Most of these arguments can be disposed of in a sentence or two, but one deserves a more detailed response.

The easy ones first:

  • A number of industry folks insisted that the licensing landscape is simply too complex to be reduced to our proposed short notice format. They pointed out that each publisher, studio, and label negotiates separately with each platform. There are thousands of different terms out there, they claimed, and that variety would overwhelm consumers, even in a more readable format. Even taking this assertion as true, it ignores the fact that no matter how many deals go on behind the scenes, each platform has a tiny handful of consumer-facing license. Apple has one license for US customers. Amazon has separate licenses for its Kindle, mp3, and Video stores. That's four. I think we could manage the complexity. When I asked the industry representatives to clarify this obvious discrepancy, they couldn't muster a single word of explanation. It was sort of embarrassing, honestly.
  • Others insisted that there can't really be a problem here because there are so few consumer complaints and so many repeat customers. It took me about 30 seconds of Googling to find multiple posts on Amazon's own forum where consumers are asking and complaining about the ability to resell and lend their digital goods. Of course, I don't have access to their internal emails. But the public record is enough to cast serious doubt on this claim. As for repeat customers, the answer is pretty simple. The absence of the rights to lend, resell, and leave digital goods in your will are latent defects. Consumers don't know about them until later, sometimes much later, and may happily spend hundreds or thousands of dollars before they discover that they can't use or transfer these goods in the ways they expected. Volkswagen sold 11 million diesel gate vehicles before anyone caught on, after all.
  • In a moment of what I can only explain as unintended forthrightness, a representative from the Digital Media Association argued that retailers shouldn't be forced to clearly disclose limits on digital goods because doing so might reduce sales. In other words, if consumers knew what they were really getting, they wouldn't want these products. Without realizing it, he spelled out the case for false advertising law. If telling the truth about your product prevents a sale, you don't deserve that sale.

But mostly I want to focus on an argument that crops up over and over again in this debate. It goes something like this: "Consumers understand that digital goods come with a more limited suite of rights because they are so much cheaper than their physical counterparts. This difference in price gives consumers a clear signal that they aren't receiving the more robust set of rights they get when they cough up the extra money for a CD, Bluray, of physical book." 

I'm tired of hearing this argument. For one, it demonstrates the same lazy set of self-serving assumptions that prompted our work on the Buy Now button in the first place. What consumers know, understand, and believe are empirical questions. If you want to use those terms and make those sort of claims, bring data to back them up, or stop expecting people to take you seriously. For another, it just isn't true. Digital isn't always cheaper. And it certainly isn't cheap enough in comparison to physical copies to communicate anything to consumers about their rights. How do I know? Fourth grade math. Try it at home, kids.

I looked at the top 100 best selling records, movies, and books on Amazon. And it turns out, the claim that digital cheaper is bogus. 

There were 77 CDs out of the top 100 that had an equivalent mp3 version for sale. That number excludes vinyl, cassette, and CD-only releases. The average price for a CD was $11.07. The average price for the corresponding mp3 was $11.38. That's right, it costs an extra 31 cents, on average, to buy a digital album—despite the fact that the CD can be resold, loaned to a friend, and given away in a will. What's even more absurd, 68 of the 77 CDs came with free AutoRip mp3 versions of the album. So for less money, you get both the CD and the mp3s.

In 40 of 77 instances, the digital copy cost more. In 36, the digital copy cost less. In 1, the CD and mp3 versions were the same. By the way, that was Jason Isbell's new record, which I've already preordered on vinyl. The distribution of prices—CDs in blue, mp3s in green, prices in dollars— looks like this:

There were 63 Bluray movies out of the top 100 that had an equivalent digital version. The average price for a Bluray was $17.54. DVDs were considerably cheaper, but I used the current generation format. The average price for the corresponding Amazon Video version was $17.63. Again, digital was more expensive, not less—despite what copyright holders and retailers want us to believe.

Out of the 63 titles, digital was more expensive in 24 instances. Blurays cost more for just over half of the titles, 34. And they were the same in 5 cases. Again, the distribution: 

Finally, let's look at book prices. Of the top 100 books on Amazon, a list that includes both hardcover and paperbacks, there were 88 titles with a corresponding Kindle book version. The average physical book price was $11.71. The average Kindle book price was $11.41. Digital readers saved a whopping 30 cents. Is that enough to communicate prohibitions on lending, resale, and giving away your books? Unlikely. Our research shows consumers place a considerably higher value on the rights associated with ownership.  Amusingly, George Orwell's 1984—the book Amazon infamously deleted from the Kindles of its customers—was among the titles that cost more on the Kindle than on paper. Does anyone really think consumers are paying more for the privilege of remote deletion?

Of the 88 books, the Kindle version was more expensive for 40 of them. They were the same for 1. And the physical books cost more for 47, again just over half. Here's the distribution:

So until someone comes up with a more complete dataset, which the trade associations and their members easily could, it's time to retire this bogus argument. But somehow, I doubt we've seen the last of it. For one side of this debate, facts don't seem to matter.

Truthiness on the Market

Last week, a couple of bloggers—one of whom spent a few years as an assistant law professor a decade or so back—wrote a laughably inaccurate critique of my paper with Chris Hoofnagle, What We Buy When We Buy Now, which is out now in the University of Pennsylvania Law Review. 

A response of commensurate seriousness and rigor probably would have taken the form of a single animated gif. Nonetheless, I can't resist the urge to outline their inept effort in more detail. But before getting into the specifics, some context might be helpful. What We Buy When We Buy Now reports the results of an empirical survey designed to measure the extent to which consumers understand the bundles of rights they acquire when they obtain digital media goods from companies like Amazon and Apple. In particular, we were interested in whether the phrase "Buy Now" accurately communicates to consumers the restrictions on use, possession, and alienability that are part of those bargains.

As it turns out, after constructing a simulated e-commerce site offering digital books, music, and movies and surveying nearly 1300 respondents, we found that a significant number of consumers are mistaken about what rights they get when they Buy Now. And that rate of confusion was considerably higher than the rate among those who purchase corresponding physical goods. What’s more, removing the Buy Now button and replacing it with a short notice outlining the things consumers can and can’t do with their purchases reduced confusion significantly. We also establish that these rights are material to consumers. That is, they affect whether consumers will buy these products and at what price. 

One final prefatory point on the timing here. The critique of our work comes on the heels of  the USPTO and NTIA announcing a Public Meeting on Consumer Messaging in Connection With Online Transactions Involving Copyrighted Works. The event grows out of the Commerce Department's White Paper on Remixes, First Sale, and Statutory Damages, which expressed some concerns about misleading terms like Buy Now. I was asked to present our work there and expect it to play a role in shaping the conversation. The critique does a great job of distracting from the issues on the agenda, issues that could have uncomfortable financial and public relations consequences for a handful of powerful companies.

Ok, so these two wrote a 2600 word response to our article. It was a real slog. I mean, are these guys paid by the word, or what? I expect long-winded posts from academics; we have nothing better to do with our time. But I'd expect a couple of free-marketeers to be out there contributing productively to the economy or something. So what flaws do they point out in our study? The following is as close as they come to actually addressing that question:

The authors seek to establish this deception through a poorly constructed survey regarding consumers’ understanding of the parameters of their property interests in digitally acquired copies. (The survey’s considerable limitations is [sic] a topic for another day….)

Cool story, bro.

In all that text, they don’t offer a single meaningful critique of our survey methodology, the reliability of our data, or our analysis of the data. Hell, they don’t even bother to summarize our findings, lest some stray reader make up their own mind. Instead, they argue the data don’t matter. And they offer their own set of assumptions about the state of the world that are inconsistent with the facts we present. Call me old fashioned, but I’m not ready to embrace the world of alternative facts just yet. If you want to critique empirical research, you have to engage with the evidence. These guys don't even try.

From there, they begin constructing their straw man, putting words in our mouths and mischaracterizing our conclusions in terms that are easier for them to dismiss. For example:

[T]he authors assert, because the common usage of the term “buy” indicates that there will be some conveyance of property that necessarily includes absolute rights such as alienability, descendibility, and excludability, and digital content doesn’t generally come with these attributes. 


We assert no such thing. In fact, we did the opposite. Rather than asserting, we asked consumers what rights they thought they obtained in digital transactions. The results were mixed. Some thought they obtained a right to lend their ebooks; others didn't. Some thought they obtained the right to leave their mp3 collections in their wills; others didn't. The key point, however, is that a significant percentage of consumers—far more than in the case of physical goods—misunderstood the nature of these admittedly and obviously non-absolute rights.

Our critics might not like the results that we uncovered, but they don't get to dismiss them by insisting that our conception of property requires some sort of absolute transfer of rights.

But they go on:

Getting to their conclusion that platforms are engaged in deceptive practices requires two leaps of faith: First, that property interests are absolute and that any restraint on the use of “property” is inconsistent with the notion of ownership; and second, that consumers’ stated expectations (even assuming that they were measured correctly) alone determine the appropriate contours of legal (and economic) property interests. Both leaps are meritless.

They are wrong on two fronts here. First, neither of these assumptions are necessary to establish our fairly modest claims. Second, we don't make either of these assumptions in the paper.

The claim we are making is that the Buy Now button inaccurately communicates something to consumers about the bundle of rights they acquire. The precise content of that bundle isn't all that important, so long as the rights at issue are material. It could be a big bundle or a small one, an absolute bundle or one subject to all sorts of restrictions and limitations imposed by public or private actors. Our goal was to assess the potential mismatch between the bundle consumers thought they were getting on the basis of the Buy Now label—whatever it happened to be—and the bundle as described by the license terms under which these goods were sold.

The relevance of that assessment has precisely nothing to do with "the appropriate contours of legal (and economic) property interests." This paper does not argue that we should reshape property law or expand copyright's first sale doctrine in response to these survey results. We certainly don't claim that consumer perceptions alone should define the scope of property interests. Our critics don't provide a quote or citation for that claim, because they can't.

Instead, what we argue is far more modest: that digital retailers should use language that informs consumers in a clear and effective way about the nature of these transactions. For the market to work well, consumers need accurate information about the products they are buying. And what bundle of rights is transferred in the context of a digital good is crucially important information. Today, consumers are expected to wade through 20,000 word license agreements in order to obtain it. Fixing that particular problem does not require a change to copyright or property law. 

Then things start to get comical, as our critics—in response to a study that reveals what consumers actually think based on, you know, like, evidence and stuff—make a number of speculative and wholly unsupported claims about how they assume consumers should think. Here are a handful of highlights:

When we buy digital goods, we probably care a great deal about a few terms. For a digital music file, for example, we care first and foremost about whether it will play on our device(s). 

In fact, given the price-to-value ratio, it is perhaps reasonable to think that consumers know full well (or at least suspect) that there might be some corresponding limitations on use — the inability to resell, for example — that would explain the discount. 

P&H want us to believe that consumers can’t distinguish between the physical and virtual worlds, and that their ability to use media doesn’t differentiate between these realms. But consumers do understand (to the extent that they care) that they are buying a different product, with different attributes.

And, furthermore, the notion that consumers better understood their rights — and the limitations on ownership — in the physical world and that they carried these well-informed expectations into the digital realm is fantasy. 

Here we have a set of claims about what consumers "probably" care about and what is "perhaps reasonable" to assume they "suspect." Upon what are these suppositions based? Apparently, our critics' superior instincts about consumer psychology. Because they sure aren't based on evidence. We went out and did the hard work of finding out what consumers actually believe. But our critics would rather trust their guts than accept the evidence we present. Part of what is on display here is the reluctance of a certain lineage of law & econ dude to admit that his convenient assumptions about the way the world works might not correspond all that well to reality.  

It turns out this unwillingness to confront reality carries over to the question of price as well. Our critics argue consumers should know they are getting a lesser set of rights because digital goods are so much cheaper than their physical counterparts. To support that claim, we get some back of the envelope calculations based on—and I'm not joking here—the price of CDs in 1982, which adjusted for inflation is $38 today. Putting aside the fact that we took price variation into account in our study, this is an astoundingly dumb argument. Let me get this straight: Today's consumers, many of whom weren't even born in 1982, should know they are getting a lesser bundle of rights in digital goods because CDs were really expensive 35 years ago? Back to the real world, the fact is digital and analog copies are often priced very similarly today, and in a surprising number of cases, digital copies are actually more expensive, not less. But hey, don't let the facts stand in the way of a good story, right?

Then they move back to this refrain:

P&H believe that digital copies of works are sufficiently similar to analog versions, that traditional doctrines of exhaustion (which would permit a lawful owner of a copy of a work to dispose of that copy as he or she deems appropriate) should apply equally to digital copies, and thus that the inability to alienate the copy as the consumer wants means that there is no ownership interest per se.

Citation needed. We do not make this claim in the article. That's not what this article is about. And any honest person who bothered to read what we wrote would understand that it is not advocating for a change in copyright law, the first sale doctrine, the law of sales, or the law of property. It is presenting data that reveal how consumers understand these transactions and then analyzing that data through the lens of false advertising law. 

At root, P&H are not truly concerned about consumer deception; they are concerned about what they view as unreasonable constraints on the “rights” of consumers imposed by copyright law in the digital realm. 

At root, Geoffrey Manne & Neil Turkewitz are not truly concerned with our scholarship; they are concerned about maintaining cozy and profitable relationships with companies like Amazon. See how easy that is? I have no idea if it’s true. And even if it is, it wouldn’t invalidate any accurate, substantive critique in their post—assuming you can find one. 

But I confess, through their unparalleled powers of reading between the lines and divining the secret motivations, these two have uncovered my deep dark secret. I admit to you all today: I support a digital first sale right

At times, it feels like these guys would rather be writing a hit piece on my book with Jason Schultz, The End of Ownership. And these concerns about digital first sale and the balance between IP and personal property would be legitimate grounds for disagreement there. But this paper, to repeat the refrain, isn't about that question, no matter how much they want it to be. Fellas, if you want to write a nasty blog post about the book, just let me know. I'd be happy to send you a couple of copies.

Then, without a hint of self awareness, they conclude with this gem:

When one starts an analysis with an already-identified conclusion, the path from hypothesis to result is unlikely to withstand scrutiny, and that is certainly the case here.

Finally, a point on which we can agree.

The Year Ownership Broke

We're taking part in Copyright Week, a series of actions and discussions supporting key principles that should guide copyright policy. Every day this week, various groups are taking on different elements of the law, and addressing what's at stake, and what we need to do to make sure that copyright promotes creativity and innovation.

One of our worries in writing The End of Ownership was that, given the pace of traditional print publishing, the questions we raise would be stale before the book made its way into the hands of readers. It turns out that fear, unlike many expressed in the book itself, was unfounded. Sure, new examples of the erosion of consumer property rights have continued to multiply like so many soaking mogwai in the months since we made our final manuscript edits.

But if anything, that demonstrates the ongoing relevance of the basic themes we identify: tangible copies continue to be displaced by digital distribution; license terms still routinely insist that sales should be treated as contingent grants of permission; embedded software and DRM, now more than ever, ensure that manufacturers and IP rights holders are the true masters of the devices we buy, preventing us from modifying, repairing, and even using the things we think we own; and despite evidence of deception, digital retailers still insist on marketing their goods using terms like "buy" and "own." But these profound shifts in the dynamics of ownership remain unknown, or at least poorly understood, by the average consumer attempting to navigate an increasingly complex marketplace—one that highlights the many benefits of licensing, subscribing, and "sharing," while simultaneously diverting attention from the security and autonomy afforded by old fashioned personal property rights.

2017 looks to be a watershed year for ownership. For better or worse, what happens over the next 12 months will go a long way towards deciding whether we will live in a world in which we own and can control the products we rely on. Or instead, 2017 could set us on a path towards a future that looks a lot like Philip K. Dick's Ubik, with each of us negotiating with Alexa over the fee to unlock our front doors.

In the market, the push for smart devices—burdened though they are by restrictive license terms, software locks, and code designed to constrain our behavior—persists. A cursory glance at the new wares on display at CES earlier this month reveals that the trend of software-enabled and network-connected devices is still going strong, all good sense be damned. Can I interest anyone in a wifi hairbrush? Or perhaps a smart mirror?

Given the state of market, developments in the courts, administrative agencies, and perhaps Congress are crucial in determining whether and to what extent consumers will be permitted to retain the rights associated with ownership in the things they buy. 

On the judicial front, the courts are poised to decide two cases that squarely address consumer ownership rights, one in the copyright space, the other concerning patents. The Second Circuit will be hearing the long-awaited appeal in Capitol Records v. ReDigi. There the court will be asked to decide whether consumers who purchase digital music enjoy the same rights to alienate their interests that analog consumers have exercised for decades at used record shops. ReDigi developed a marketplace for used digital music that carefully replicated the one-to-one exchange familiar from the analog world. But in an analytically-murky decision, the district court held that ReDigi's system infringed both the reproduction and distribution rights of copyright holders.

Meanwhile, the Supreme Court is set to hear Impression Products v. Lexmark, a case that challenges two flawed holdings of the Federal Circuit. Lexmark concerns the printer-maker's single-use restriction that forbids consumers from refilling ink cartridges they've purchased under pain of patent infringement. The Federal Circuit held, first, that such post-sale restrictions on the use of a chattel are enforceable as a matter of patent law, and second, that overseas sales authorized by the patent holder cannot trigger exhaustion—a holding in no small degree of tension with the Supreme Court's 2012 opinion in Kirtsaeng

On the administrative agency side of things—well legislative agency, but who's keeping score—the Copyright Office is fresh off a deeply disappointing report on Software-Enabled Consumer Products, in which it concluded after nearly 70 pages of analysis that software embedded in consumer devices poses no real problems. This, despite the fact that copyright holders have leveraged their ownership of software to interfere with resale, to stifle competition and user innovation, to prevent repair, and to disable devices altogether. The Copyright Office dismisses these concerns, in part, by arguing that the companies involved learned their lesson. It wrote for example that "market forces may discourage copyright owners from attempting to prevent independent repair activitie," citing the outcry when John Deere infamously asserted that farmers enjoy a mere license to operate the equipment they buy. But just weeks after the Copyright Office's handwaving, John Deere released a new license agreement that affirmatively denies farmers the right to repair their devices or even look at the software that makes them run. So much for the market.

As you may have heard, leadership changes are underway at the Copyright Office. The new Register of Copyrights will have many responsibilities, chief among many hope is, you know, registering copyrights. But to the extent the Copyright Office continues to pursue an active policy agenda, it needs to take the question of consumer ownership seriously. The Office still has an study of Section 1201 underway, which implicates many of the same concerns about ownership outlined above. Here's hoping the Office takes them more seriously the second time around.

With a new administration set to take over at the end of the week, other agencies will have new leadership as well, among them the Federal Trade Commission. The FTC has played an important role over the years guarding against the most egregious abuses in the digital marketplace. When Major League Baseball, Microsoft, and Walmart threatened to shut down servers providing consumers access to digital media purchases, the FTC stepped in. More recently, when Nest decided to brick thousands of Revolv home automation hubs without even the offer of a refund, the FTC again flexed its considerable muscle. It made sure that consumers were at least compensated for their now-worthless devices. With the Trump administration poised to choose to three of the FTC's five commissioners, the Commission may be less inclined to aggressively target deception and unfair practices.

Speaking of deception, spurred by the Commerce Department's White Paper on Remixes, First Sale, and Statutory Damages and some research by a couple of crackpot academics, the PTO and NTIA are planning public hearings on the "Buy Now" button and its potential for misleading consumers. Again, how high of a priority this issue will remain in a Trump administration is hard to predict.

Then of course, there's Congress. Copyright reform remains on the agenda. A meaningful overhaul of the 1976 Act should be a longterm project. The current proposal is short on details, but appears to be decidedly small-bore. We've also seen legislation that more directly targets the ownership question in recent years, like the Unlocking Technology Act and the You Own Devices Act. Although these bills haven't made much headway, they do suggest that some in Congress care about the threats to ownership facing consumers. And I expect we will see similar proposals in the coming term.

Legislation, regulation, and judicial opinions matter. But when it comes to the fight for ownership, we think nothing is more important than awareness. As we talk to people about our book, we see over and over again that once consumers recognize the threats to their rights that the current digital marketplace represents, they are alarmed. Some are shocked. Others are outraged. They see the choices they confront and their consequences differently. Some even start to behave differently. That's why we are telling this story to anyone who will listen, and why we hope you will do the same.

Why you can't buy our book from Apple

There's no small amount of irony in selling an ebook that highlights the downsides of shifting from tangible to digital goods. But many people, including us, are persuaded by the virtues of digital reading even once we are fully aware of the tradeoffs it involves. And ultimately, we favor consumer choice. So making our book widely available on as many platforms and in as many formats as possible was sort of a no-brainer.

That's why we were disappointed, if not altogether surprised, when Apple refused to carry The End of Ownership in its iBooks Store, one of the world's largest ebook marketplaces. Although the book is available on the Kindle and Nook , there isn't currently an iBook version. This post explains why.

Apple runs its iBooks Store in the same way it runs the rest of its business—with tight quality control measures and an innate skepticism of anything that interrupts its end-to-end control of the user experience. Much as it does on the App Store, Apple vets submitted iBooks before offering them to the public. Typically, this review filters out submissions with technical flaws as a way of maintaining high quality and promoting consumer confidence.

But sometimes Apple uses its veto power in questionable ways. It has rejected apps that criticize political figures and others that feature adult content. It has approved some cryptocurrency apps and refused others. It even rejected an overtly poltical Palestinian-produced game on the grounds that it was "not appropriate"—a position the company later reversed. Apple's history in reviewing books is likewise problematic. Books have been rejected for having bibliographies that linked to webpages selling hardcover books; others for merely discussing competitors like Amazon.

Booksellers are, and should remain, free to refuse to carry books. That editorial discretion is crucial to their own free speech interests. But how they exercise that discretion should give us cause for concern when the retailer controls a significant portion of the market. For the same reasons we should be troubled when Walmart refuses to sell certain books, we should worry about the implications of Apple's tight control over its marketplace. And since Apple is the only seller for iBooks, the problem is compounded.

So why did Apple decline to sell The End of Ownership? The book is openly critical of Apple in a number of respects. We critique its embrace of DRM, its crackdown on independent repair, its complex and unreadable EULAs, its use of the deceptive Buy Now language to sell digital goods, and its commitment to hermetically-sealed business models.

But that criticism is not the reason Apple cited. Instead, the company pointed to its trademark policies. Apple noted passages in the book where we used the term "iBook" to refer to ebooks sold by Apple. But according to its iBooks Store Formatting Guidelines, Apple prefers the terms "iBooks" or "iBooks Store" be used to refer to its software ecosystem, and discourages the use of "iBook" to refer to ebooks sold using that platform.

Apple's rules regarding "copyrighted terms," by which they mean "trademarked terms."

Apple is certainly free to develop its own internal style guide to determine how it uses its long and growing list of trademarks. But why should we care? The company would likely argue that it is fighting a battle against genericide—the loss of trademark rights that can occur when a brand name becomes the common name for a category of product. Xerox and Kleenex have been subjects of corrective advertising campaigns to remind consumers that they are particular products made by particular companies, not the general term for photocopiers and tissues. 

But we didn't use "iBook" as a generic term; we used it to refer specifically and exclusively to ebooks sold by Apple. So the argument is misplaced. Elsewhere, Apple repeats the common trademark mantra that a mark is an adjective, not a noun. But even casual perusal of Apple's website and marketing materials reveals widespread use of terms like "MacBook," "iPhone," and "iTunes" as nouns.

But there's another, more amusing, explanation for Apple's nitpicking about the use of this particular term. In 1999, long before it was seriously considering becoming a digital book retailer, Apple launched a consumer laptop meant to capitalize on the success of the iMac. At the time, it's professional laptops were called PowerBooks, so the new device was dubbed the iBook. Apple discontinued the product in 2006. But its legacy lives on. Ebooks sold on the iBooks Store are quizzically not called iBooks—because that name was already taken by Apple itself.

Regardless of the ultimate explanation, the fact is our book is currently unavailable to the millions of readers who buy iBooks—the books, not the laptops. We think some of them might benefit from the message our book offers. We could fight Apple on its rejection. After all, we are both IP professors with litigation experience. But Apple has too much power and too many resources for that fight to be worth the effort in the end.

So today, we are submitting three minor changes to the text of our book to appease the Style Guide Gods in Cupertino. To be as transparent as possible, we've included those edits below. We hope to see our iBook—again, the book, not the laptop—posted to the iBooks Store soon.

Page 93 - "Apple’s iBooks can only be read on Apple devices." changed to: "Ebooks acquired through Apple’s iBooks Store can only be read on Apple devices."
Page 177 - "We shouldn’t expect every ebook reader to figure out on their own how to make an iBook work on a Kindle" changed to: "We shouldn’t expect every reader to figure out on their own how to make an an ebook from Apple work on a Kindle."

183 - "Sometimes an owner will want to modify their copy for compatibility purposes—to make their iBooks work on a Kindle" changed to: "Sometimes an owner will want to modify their copy for compatibility purposes—to make their iBooks Store purchases work on a Kindle."

When your car owns you

When we try to illustrate the dangers of displacing consumer ownership with a model that relies on software licensing, it's easy to come up with troubling hypotheticals. Would you buy a car with DRM-enforced licensing terms that forced you to buy gas exclusively from BP? Or one that limited you to licensed replacement tires? What about a car that required you to pay an extra fee to carpool? 

To many, these sort of examples probably seem far-fetched. But we've already seen carmakers leverage their control over software to clamp down on independent repair shops. A combination of state-level legislation and industry groups coming to their senses helped address that particular problem. And Ferrari limits resale by insisting on a contractual right of first refusal on used cars. But this week, we got our clearest glimpse so far of the power software and license terms give manufacturers over your vehicle.

As Ars Technica reported, the new Tesla Models S & X can be equipped with "Full Self-Driving Capability" as a $3000 option. The software to actually enable autonomous driving isn't ready yet, but will be at a future date. But once it goes live, there will be one important string attached that could significantly reduce the feature's value to Tesla owners. As the company explained: "Using a self-driving Tesla for car sharing and ride hailing for friends and family is fine, but doing so for revenue purposes will only be permissible on the Tesla Network, details of which will be released next year."

In other words, if you want to make money by charging for rides in your self-driving Model S, Tesla is going to get a cut; no competing ride-hailing services for you. You'd think for $70,000, Tesla owners would get the right to use their cars as they see fit. But Tesla doesn't think so. 

Tesla may have good technological reasons to worry about the integration of its software with other car sharing networks. But it also has strong anti-competitive motivations for this sort of restriction. If Tesla owners can't opt for Uber or Lyft, Tesla will be in a much stronger position to dictate pricing and terms for its ride-sharing service and may be at an advantage when it launches its own fleet of on-demad self-driving vehicles. But regardless of the company's motivations, car owners should be troubled by the claim that a manufacturer can dictate how and for what purpose they drive their vehicles.

Downloads and consumer choice

A few months back, reports surfaced that Apple—the pioneer of digital music sales—would discontinue downloads in favor of an exclusive subscription streaming model within the next two years. Although Apple has sold tens of billions of tracks since launching the iTunes store more than a decade ago, digital downloads have slumped as streaming services have seen massive growth. So Apple's aggressive moves into the streaming market took no one by surprise. The news of the end of downloads, however, raised eyebrows in some corners. But Apple was quick to deny reports that it planned to kill off downloads.

Recent developments cast some doubts on those disavowals. When Apple Music launched in China, it offered monthly subscription plans, but didn't include the ability to buy individual downloads from the iTunes store, although movie and book downloads were available. Similarly, á la carte downloads were conspicuously absent from this week's Apple Music debut in South Korea.

One of the recurring themes in our book is the importance of consumer choice. Streaming  models are not inherently bad. Far from it. In fact, they are often preferable from the perspective of individual consumers. But not everyone wants to stream. Some consumers prefer to own their stuff, both physical and digital. And the option to buy, to possess, to control a local copy does valuable work in the market. Most importantly, it limits the power of service providers to control price and availability. 

We've consistently expressed concern about a future that focuses on inherently contingent forms of access like subscription streaming services to the exclusion of more permanent and durable means of acquisition. When we described this possible future, we were worried that these concerns would be labeled "alarmist." But more and more, they look like the world in which we all live.

Just Blizzard Being Blizzard

Blizzard Entertainment, the company that develops hit video game franchises and licenses forgettable movies that squander directional talent, is back to its favorite pastime—filing and threatening gratuitous copyright lawsuits. Before detailing the latest litigation, it's worth revisiting the company's greatest hits. They include:

  • Suing the developers of bnetd, an open source matchmaking server in 2002. Although bnetd didn't copy any Blizzard code, the developers reverse engineered Blizzard's service to discover and implement the protocols that allowed games like Starcraft, Warcraft, and Diablo to interoperate with Blizzard's servers. In 2005, the Eighth Circuit affirmed a district court opinion holding that bnetd violated the DMCA's anticircumvention provisions.
  • Using the DMCA to shut down private World of Warcraft servers in 2008. 
  • Suing the developers of Glider, a program that allowed gamers to automate some of the monotonous drudgery required to level up in Blizzard games. Here, Blizzard wasn't content to offer one damaging misreading of copyright law; it had two. First, Blizzard—like Autodesk, Amazon, and John Deere—argued that its customers don't actually own their copies of World of Warcraft, they merely license them. And applying its disastrous precedent from Vernor v. Autodesk, the Ninth Circuit agreed in 2010. Second, Blizzard argued that by using the game in ways it didn't like, users were infringing the game's copyright. Specifically, Blizzard pointed to its terms of use, which prohibited the use of "cheats, bots, mods, and/or hacks." Despite that prohibition, the court correctly held that breaking a rule is not the same thing as copyright infringement, even if you don't own the copy in question.
  • Apparently unwilling to give up on the "cheating is infringement" argument, suing ten anonymous developers—one of whom was almost certainly Angelina Jolie—for creating the ValiantChaos MapHack, which gave players the advantage of seeing obscured areas of the game map and tracking the movements of opponents.
  • Shutting down the popular Nostalrius server, which gave 150,000 users the chance to play a classic version of World of Warcraft earlier this year. 

But Blizzard has another hit on its hands, Overwatch. Which apparently means more litigation is in the offing. This time Blizzard is suing a German company that sells Overwatch cheats. Again, Blizzard argues that gamers who use these cheats are committing copyright infringement, and Bossland, the company offering them, is contributing to that infringement.

To see the absurdity of this position, consider an analog analog. Some players of Hasbro's classic Monopoly often find the standard set of rules that accompany the game lacking. Maybe you think they are needlessly complex and lead to interminable gameplay. Maybe you are a lunatic and don't think they are complex enough as is. In either case, users come up with their own variations on the rules. There are two approaches to respond to these house rules. One approach is the one Hasbro has adopted. Embrace them. The other is the approach Blizzard has pursued. Insist your rules are the best rules. The only rules. And that anyone who modifies, alters, or breaks them is not only a cheater, but a copyright infringer. Imagine if Hasbro went around suing players for deciding you shouldn't receive rent while your game piece is in jail. Of course, you own your Monopoly set, so you can do what you damn well please with it. As a matter of copyright law, the same should be true of software-based games.