Vernor v. Autodesk and the End of the First Sale Doctrine

The 9th Circuit’s Vernor v. Autodesk test demolishes the first sale doctrine by making its application contingent solely on the licensing agreement written by the copyright holder. Though the Vernor case centers on the distribution of software, there is no limiting principle that prevents the Vernor test from being applied broadly to all copyrighted works. Thus, the Vernor test, if upheld, it could mean the end of all markets for used copyrighted works.

The First Sale Doctrine

The first sale doctrine was established by the Supreme Court in Bobbs-Merrill Co. v. Straus where a book publisher printed the following note on its copyright page: “The price of this book at retail is $1 net. No dealer is licensed to sell it at a less [sic] price and a sale at a less [sic] price will be treated as an infringement of the copyright.” The Supreme Court held that, under existing copyright law, copyright holders have the exclusive right of distribution over the “first sale” of their works, but further distributions are outside of their control.  The first sale doctrine was later codified in the Copyright Act.

As a legal principle, first sale strikes a balance between the rights of copyright holder and the rights of the owner of a copy of said material. The doctrine also embodies the general principle in property law that unreasonable constraints on alienation (gifting, selling, etc.) are void.

First sale also promotes the value of free access to information by making out of print copyrighted works widely available and lowering prices through the existence of secondary markets.

Many copyright owners justifiably dislike the first sale doctrine because it prevents them from maintaining a monopoly on their copyrighted works, and it enables secondary markets which tend to drive down prices.

The Vernor Decision and the Sale/License Distinction

The essential facts of Vernor are straightforward. Vernor purchased used software at a garage sale and attempted to sell it on eBay. The copyright holder, Autodesk, filed several DMCA take-down notices with eBay. After some back and forth, Vernor brought a declaratory action in Federal District Court to establish that his resale was protected by first sale doctrine.

The primary legal issue was whether the transfer of Autodesk’s software to the customer who had sold it to Vernor constituted a sale or a licensing. This is the legal hook: if all that was transferred was license, the “first sale” has not occurred and the doctrine does not apply.

The licensee/owner distinction was not clear law prior to Vernor. The District Court in Vernor, determining that there were conflicting precedents on point, applied the 9th Circuit case United States v. Wise and found that the critical factor in the sale/license distinction was whether the purchaser had a right to possess the copyrighted work perpetually or whether he was required to return it to the copyright holder. The court found the right to perpetual possession and thus held that Vernor was covered by the first-sale doctrine.

On appeal, the 9th Circuit held that the license/ownership distinction depended on only three factors: (1) whether the copyright owner specifies that a user is granted a license (2) whether the copyright owner significantly restricts the user’s ability to transfer the software (3) whether the copyright owner imposes notable use restrictions. Applying this test to Autodesk, the court found that the transfer in question was a mere transfer of license and, thus, that Vernor was not protected by the first sale doctrine.

Criticism of the Vernor Test

The problem with the Vernor test is clear: a copyright holder can completely avoid the first sale doctrine by using the term “license” coupled with the “significant” transfer and use restrictions. Thus, under Vernor the application of first sale doctrine depends solely on the discretion of the copyright holder and what “magic words” he chooses to place in the license agreement.

The Vernor test completely undermines the first sale doctrine and all of its underlying policies. First sale is meant to balance between the rights of copyright owners and the rights of owners of copies; Vernor undermines that balance by making its applicability contingent on a copyright holder’s preference. Under the Vernor test, if Bobbs-Merrill Co. had written its note slightly differently, referring to the purchaser as a licensee, and imposing more use and transfer restrictions, the case would have come out the other way—an absolutely preposterous result.

And finally, Vernor effectively negates the principle against unreasonable restrictions on alienation as they apply to copyrightable works. A copyright holder only has to write the “magic words” in a licensing agreement to prevent resale. Thus, there is little standing in the way of copyright holders from unilaterally destroying secondary markets—the used software, book, DVD, CD and videogame market—and maintaining a monopoly on its work.

Currently, the plaintiffs in Vernor are petitioning for an en banc rehearing in the 9th Circuit. If denied, they will likely take it to the Supreme Court. Vernor as it stands today has the potential to fundamentally change not only the legal relevance of the first sale doctrine, but the entire economic, social and legal landscape for copyrightable works.

6 Replies to “Vernor v. Autodesk and the End of the First Sale Doctrine”

  1. You've done a good job pointing out how the Vernor test undermines the policy concerns behind the first sale doctrine. In addition to that problem with the test, I see at least three other specific problems with it.

    First, it is logically flawed because the basic premise of the test is that the terms in the license agreement are valid and binding on the user. However, this premise itself depends entirely on the very question the court is attempting to answer: is the user an “owner of a copy”? If we assume the user *is* an owner of a copy, then terms in the license agreement that purport to restrict a copy-owner's rights are simply invalid. Yet the test plainly assumes that the terms are valid, meaningful, and therefore useful for determining copy ownership. This is known as “begging the question” and is a well-known logical fallacy.

    Second, the test fails to recognize that an owner of a copy might also happen to be a licensee. This is quite obvious, but the test doesn't provide for this possibility. Under the test, a user is either an owner or a licensee — never both.

    Third, the test conflates the act of licensing intangible rights with transfer of ownership of personal property. It treats them as one and the same, but in reality most software transactions involve two parts: a transfer of possession (and possibly ownership!) of personal property, *and* a grant of a license. To determine the ownership status of a *copy*, which is by definition a piece of personal property, the court must examine the former part. The test must address whether or not the transaction that resulted in the transfer of possession was a sale or other transfer of ownership. The terms of the license agreement are completely irrelevant to this determination, because they are (usually) not part of the transaction in which the user obtained possession. Consider software sold at retail: the retail sale is a transaction entirely separate from the grant of license; the copyright owner isn't even a party to the retail transaction so it seems absurd to assert that their license agreement controls this transaction.

  2. I agree with this analysis of what the case will do, and hope the Ninth Circuit re-hears en banc. Check out this analysis, at the link below, that argues that the issue the Ninth Circuit should have decided is whether, when a software user “upgrades” to a newer version, thus “keeping the software,” this is a “first sale,” at all. Seems to me that the answer is, “Of course not. The original owner still has the software.” Preventing software users from strategic behavior by buying the upgrade and selling the original version is appropriate, but the Ninth Circuit has unnecessarily expanded the power of licensors to write transfer-of-title restrictions into the law of personal property. Interestingly, in the usual case, since those licenses aren't effective until after the user opens the software and clicks “yes” to the EULA, something that happens after the purchase has been made, the sale can't be said to be conditional upon the license terms. Which is of course exactly what Dan has said in his third comment.

  3. Pingback: La déclaration des droits de l’utilisateur de livre numérique – 2ième version « Bibliomancienne

  4. Pingback: The Indelible Bonobo Experience

  5. Pingback: E-Book Customers Don’t “Own” Their Books « Christian Book Shop Talk

  6. Pingback: Understanding the Redbox/Disney War - Plagiarism Today

Leave a Reply