This past March, the Supreme Court considered in Kirtsaeng v. John Wiley & Sons “whether the words ‘lawfully made under this title’ restrict the scope of §109(a)’s ‘first sale’ doctrine geographically.” It ultimately held that they do not — copyright’s first sale doctrine has no geographic limitation. But since then, many have looked beyond Kirtsaeng’s narrow and fact-specific holding to find what they believe to be endorsement of broad, and potentially troubling to creators, expansions to the first sale doctrine.

Digital First Sale

Almost immediately, some speculated that Kirtsaeng would have an impact on “digital first sale”, especially given the fact that a decision in Capitol Records v. ReDigi dealing with that very issue was pending in the Southern District Court of New York (ReDigi operated a service allowing users to buy and sell “used” mp3s).1 For example, Joe Wikert, of TOC, wrote the same day that “the Kirtsaeng ruling can only help ReDigi’s case and that’s a good thing for anyone who wants the ability to resell their digital goods,” though he was far from the only one to share this view.2 But the court in ReDigi didn’t see anything in Kirtsaeng‘s general remarks about first sale that rendered the provisions at issue ambiguous, and it would hold that ReDigi did not have a valid first sale defense — the first sale doctrine only applies to the distribution of particular, lawfully made, copies of a work, while ReDigi was distributing new, unlawfully made, reproductions of copies.

Sales vs. Licensing

More recently, a software reseller has sought solace in Kirtsaeng‘s first sale language for a different issue in Adobe v. Kornrumpf, currently on appeal to the Ninth Circuit.

Adobe was in the practice of licensing OEM versions of its software — such as Photoshop Elements 8 — to be sold exclusively in conjunction with hardware equipment. According to court records, Kornrumpf acquired copies of these OEM products and resold them separately. Adobe sued, alleging infringement of its exclusive distribution rights since Kornrumpf had not lawfully acquired title to the software.

The court ruled in favor of Adobe, finding that the software was licensed, not sold. The decision was a fairly straightforward application of the test developed by the Ninth Circuit in Vernor v. Autodesk: “[A] software user is a licensee rather than an owner of a copy where the copyright owner (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.” Since the evidence demonstrated that Adobe clearly met each of these elements, the first sale doctrine did not apply, and Kornrumpf was liable for infringing Adobe’s exclusive distribution right.

But Kornrumpf appealed, with a novel argument: the Supreme Court’s decision in Kirtsaeng limits the scope of Vernor. Specifically, he argues that the Court held that “market division is not a valid aim of copyright”, that it held “courts should not be burdened with enforcing restrictions on the alienability of chattels”, and that it reaffirmed “the venerable common-law aversion to restraints on alienability in the copyright context.”

As in ReDigi, these are essentially policy arguments, as they are not supported by the legal conclusions of Kirtsaeng.

Reaffirming Traditional First Sale Doctrine

In fact, if anything, Kirtsaeng reaffirmed the point that the first sale doctrine is inapplicable when copies have been licensed rather than sold. In responding to the argument that its interpretation of the first sale doctrine would render copyright’s importation provisions superfluous, for example, the Court says that, to the contrary, the provisions “would still forbid importing … copies lawfully made abroad … where … any … licensee, consignee, or bailee sought to send them to the United States.” [Emphasis added.]

What’s more, if there is anything that is venerable, it is copyright law’s tradition of recognizing a distinction between transferring title of a copy of a work through a sale and licensing a work. Going back to what is considered the origin of the first sale doctrine in U.S. copyright law, the 1908 Supreme Court case Bobbs-Merrill v. Straus, one can see this distinction already established. As the Supreme Court said there:

The precise question, therefore, in this case is, does the sole right to vend (named in § 4952) secure to the owner of the copyright the right, after a sale of the book to a purchaser, to restrict future sales of the book at retail, to the right to sell it at a certain price per copy, because of a notice in the book that a sale at a different price will be treated as an infringement, which notice has been brought home to one undertaking to sell for less than the named sum? We do not think the statute can be given such a construction, and it is to be remembered that this is purely a question of statutory construction. There is no claim in this case of contract limitation, nor license agreement controlling the subsequent sales of the book. [Emphasis added].

So it would seem that Kornrumpf, like ReDigi, will have trouble convincing courts to accept its first sale arguments. Nevertheless, these two cases exemplify a recent push, intensified by the Kirtsaeng decision, to inflate the first sale doctrine beyond its due boundaries. No doubt its proponents will raise the issue as Congress reviews the Copyright Act. Why is this troubling?

The Benefits of Traditional First Sale

For starters, a meaningful ability to license works benefits both consumers and creators. Many have observed that there is a general shift from ownership models to access models, especially with music and films.3 Licensing facilitates these new models, by allowing a more granular grant of rights than sales can provide. This gives consumers a diversity of options to read, watch, or listen, at a range of price points.

Put another way, think of it in terms of ordinary goods. For example, you can buy a car, enter into a long-term lease, rent a car for a short period, or take a cab. No one would argue that a framework that only allows outright sale of cars is preferable to this system. The same is true with real property: you can buy a house, lease an apartment, or book a hotel room. It would be an incredibly inefficient system if only the first of these was possible.

As far as digital first sale goes, there are issues with treating digital works like physical goods. There really isn’t such a thing as a “used” digital file — the file is absolutely identical to a “new” file and will remain identical indefinitely. This means “used” digital files are exact substitutes for “new” digital files and would compete directly with them.4 Obviously, there are limited issues regarding transferability and alienability that may arise, such as providing mechanisms for passing along digital files to heirs. But forcing digital works to act more like physical objects is clearly not an ideal solution.

The “First Sale License”

Which seems to be the thinking underlying much of the push for expanding first sale in the digital world: we should ignore the beneficial characteristics of the internet and arbitrarily force it to operate like a brick and mortar store. In many ways, first sale is the square peg in the future’s round hole.

Either of these — preference of sales over licensing or digital first sale — would be detrimental to creators and consumers. But the negative effects are multiplied when taken together. Most of today’s popular and innovative services — Spotify, Netflix, Audible, to name just a few — would be unlikely to exist under a system where any transfer of digital works are presumptively sales, and once transmitted they can be resold. Even copyleft systems like Creative Commons and open source software would be in trouble since they are built on copyright licensing.

In there place would be services that would operate under a “first sale license”,5 setting up convoluted systems that would allow them to provide access to songs, movies, ebooks, and other digital works without permission from or payment to creators. Imagine a streaming service that provides on-demand access to the universe of recorded music. Identical to, say, Spotify, but this service is premised on the conceit that a user “buys” a “used” song when it begins playing and then “sells” it back automatically when the song is finished. In this fashion, the service can claim it does not need permission from creators nor does it need to pay any applicable performance royalties. I don’t think such a service is outside the realm of possibilities — and if you don’t think some company will go through the effort to build a Rube Goldberg-like contrivance to take advantage of perceived loopholes in copyright law, well, I’ve got a thousand tiny antennas in Brooklyn to sell you.


  1. Kilpatrick, Townsend & Stockton LLP, “No first-sale doctrine for digital music files“, Lexology (April 8, 2013): “ReDigi had nothing to do with foreign manufacture/sale, and importation of copyrighted goods, but because the Supreme Court’s decision changed the meaning of the first-sale doctrine in the eyes of many, there was speculation that the decision might cause the Southern District to apply a comparably broad reading of the first-sale doctrine”. []
  2. See also Maria Scheid, “The First Sale Doctrine and the Sale of Digital Goods in Light of Kirtsaeng and ReDigi“, Copyright Corner (April 23, 2013): “the Supreme Court’s decision in Kirtsaeng is particularly important because it reaffirms the basic notion that one has physical ownership of the things that you buy. A court may hold this broad rationale to be equally applicable to digital goods, meaning consumers should be able to resell their digital goods under the protection of the first sale doctrine”; Daniel Gervais, “Digital Kirtsaeng: The first-sale doctrine and online content“, Blouin Beat (April 3, 2013), “ReDigi and others, relying on Kirtsaeng, will say that there is no valid reason to distinguish physical copies from online copies. When someone is done with a copy they should have the right to transfer it to someone else. They have copyright policy (as explained by a majority of the Supreme Court in Kirtsaeng) on their side.” []
  3. See, for example, Steve Jobs Was Wrong — Consumers Want To Rent Their Music, Not Own ItFuture of Music Biz: Ownership vs. Access with Ted Cohen, Founder of TAG StrategicSpotify CEO: Music access, not ownership, is the future. []
  4. Accord Register of Copyrights, DMCA Section 104 Report, U.S. Copyright Office, pp. 82-83 (August 2001):

    Physical copies of works degrade with time and use, making used copies less desirable than new ones. Digital information does not degrade, and can be reproduced perfectly on a recipient’s computer. The “used” copy is just as desirable as (in fact, is indistinguishable from) a new copy of the same work. Time, space, effort and cost no longer act as barriers to the movement of copies, since digital copies can be transmitted nearly instantaneously anywhere in the world with minimal effort and negligible cost. The need to transport physical copies of works, which acts as a natural brake on the effect of resales on the copyright owner’s market, no longer exists in the realm of digital transmissions. The ability of such “used” copies to compete for market share with new copies is thus far greater in the digital world.


  5. Analogous to the “DMCA license” that some online service providers claim allows them to build business models around copyrighted works without permission, so long as the works are uploaded by users and the service provider responds to DMCA takedown notices. []


  1. As a fairly normal consumer, I like giving books as gifts. As e-books continue to grow in popularity, I might like to give e-books as gifts.

    Libraries seem to be in the business of lending content such as books. This is generally seen to be a public good (promotes literacy, creates new readers, etc.) and beneficial to all sectors of book publishing. It seems reasonable to believe they ought to continue to lend books, even e-books.

    So your statement that “Either of these — preference of sales over licensing or digital first sale — would be detrimental to creators and consumers” seems a bit limited in its conception of the benefits of first sale. There are benefits for all stakeholders in first sale and there needs to be a place for it in the digital market. The alternative is that publishers, intermediaries, and consumers alike all will lose something important.

    • Devlin Hartline

      Libraries seem to be in the business of lending content such as books. This is generally seen to be a public good (promotes literacy, creates new readers, etc.) and beneficial to all sectors of book publishing. It seems reasonable to believe they ought to continue to lend books, even e-books.

      I’m not sure I get your point. My local public library has a liberal e-book lending policy:

      Is that not your experience too?

    • There can be a conflict of interest between libraries and authors/publishers. In some countries, including the UK, authors have obtained a ‘public lending right’ which gives them a kind of royalty on each loan of a book. Publishers also often attach conditions of sale which are intended to ensure that when the binding of a book wears out, the library buys a new book rather than rebinding it (though I suspect that these conditions are widely ignored). It isn’t obvious where the overall balance of advantage lies for publishers and authors. Libraries are major buyers of books, but lending must also to some extent cannibalise potential retail sales. Fairly early in the history of lending libraries (free public libraries or subscription libraries like Mudies) publishers seem to have decided that on balance libraries were beneficial to sales, and they produced special ‘library editions’ in a format convenient for libraries and borrowers (e.g. long Victorian novels were divided into three small volumes, so that people could borrow and return them one at a time). At the same time, more luxurious editions were produced for the wealthy classes to buy for permanent retention in their own ‘library’. But times have changed, and it is possible that the balance of advantage has swung the other way. It probably varies for different types of work.

    • Exactly. Throwing away the first sale doctrine because of some incredibly contrived examples of potential abuse seems like a horrible idea.

  2. First Sale does not exist because rightsholders and lawmakers sat down and agreed it would be a good thing.

    It exists because:

    A) in the time when First Sale was developed as a concept, it was very difficult (labor- and equipment-intensive) to disintermediate media rights from a physical token. That is, if you wanted to read a work, you had to have a physical book; there was no way for you to reproduce the text of the work independent of the paper-and-cardboard-and-ink of the printed book. So it was pretty straightforward to ensure that unauthorized copies were not coming into existence as a result of First Sale activity. (Note that bootleg recordings, copied music albums and books, and other such activity *was* unlawful and was punished when discovered.)

    B) it is very difficult to enforce strict transfer-of-rights standards on physical objects. You can take steps to make copying more difficult, but there was no way to tie the physical token to a specific owner. The communications bandwidth of the time simply wouldn’t have allowed a system where the media player had to phone home to a central authenticator and verify ownership. (Not that people didn’t try!)

    …and all of this applied historically, in the past, until now. It’s increasingly easy to separate media from physical objects; some media, now, is never associated with a specific physical object at all. It’s increasingly possible that a media player could contact an authentication server, either periodically or on every attempt to view a work. Meaning that the historical situations that resulted in First Sale are going away.