X

The end of software ownership--and why to smile

A recent ruling, while called "a major blow" to consumer rights, reflects an increasingly popular media market shift from physical to digital and from owning to renting, argues Stanford Law fellow Larry Downes.

Larry Downes
Larry Downes is an author and project director at the Georgetown Center for Business and Public Policy. His new book, with Paul Nunes, is “Big Bang Disruption: Strategy in the Age of Devastating Innovation.” Previous books include the best-selling “Unleashing the Killer App: Digital Strategies for Market Dominance.”
Larry Downes
8 min read

Editors' note: This is a guest column. See Larry Downes' bio below.

Consumer advocates are up in arms over a recent ruling by a federal court of appeals in Seattle. The decision, Vernor v. Autodesk (PDF), held that the terms of an end-user licensing agreement, or EULA, can change the sale of commercial software into a mere license, in this case a license that prohibits users from reselling their copy of the software.

The case involved an eBay seller named Timothy Vernor. Vernor bought several outdated copies of Autodesk's AutoCAD program from a business that had originally purchased the software from Autodesk. Vernor then resold the copies well below list price on eBay, despite efforts by Autodesk to have the listings canceled.

The three-judge appellate court, reversing a lower-court decision, ruled that the original owner was bound by the software license not to resell the copies. As a result, Vernor obtained no rights by purchasing the discs, and his subsequent sales violated Autodesk's copyright.

The Electronic Frontier Foundation, which filed an amicus brief supporting Vernor, decried the ruling as "a major blow to user rights."

Beyond the ruling, the opinion announced a simple test to determine when a software transaction constitutes a licensed use of the product rather than the outright sale of a copy. The transaction is a license, the court said bluntly, any time the seller specifies in a EULA or similar document that "the user is granted a license." The document "significantly restricts the user's ability to transfer the software" and otherwise "imposes notable use restrictions."

That test would treat nearly every EULA I've read as a license, not a sale. And given such helpful guidance by the court, it's safe to assume that any EULAs that don't meet these minimal requirements will be quickly rewritten. Under the Vernor test, most software transactions will be treated as licenses.

And the difference between a sale and a license can be significant, as the Vernor case demonstrates. Under a longstanding principle of copyright law known as the "first-sale doctrine," for example, the lawful owner of a copy has the right to resell it without interference from the copyright holder. The first-sale doctrine makes possible the secondary markets of used-book, -music, and -game stores, among others.

Indeed, the primary goal of sellers seeking to characterize transactions as licenses is often just to avoid the first-sale doctrine and foreclose the creation of those resale markets.

Beyond software, the appellate decision has the potential to impact resale markets for a wide range of information products that come bundled with license agreements. This includes other forms of digital content, including electronic books, music, video games, and movies.

Information owners were already a dying breed
Despite the ruling's potentially long reach, however, I don't share the doomsday view of those who believe that it will dramatically reconfigure the software industry or other information businesses.

That's largely because the reconfiguration has already happened. And the process had very little to do with law.

As recently as 10 years ago, the only way a commercial software vendor could distribute its product was to copy it onto some physical medium and sell the copies, perhaps through retailers, just as book, music, and film producers did. But the creation and distribution of physical copies adds a significant cost. And unlike books and other consumer-oriented media, media copies of software are simply loaded onto the user's computer and stored as backups. The physical media added no value to the software itself.

In fact, the use of physical media makes software considerably less useful. For once users load the physical media onto their computer, the program is fixed in time and function. Updates, bug fixes, upgrades, and new versions can be implemented only by issuing more discs, leading to relatively long gaps between releases of most application and system software products.

So software vendors have been quick to shift from physical copies to downloadable content distributed electronically. Thanks to the Internet, most commercial software today is sold in electronic form, either directly from the manufacturer or through discount online retailers, including those of Amazon.com, Costco, and Wal-Mart Stores. Similar shifts from physical to electronic formats are happening for other content, including news, books, music, and movies.

The result has dramatically changed the nature of information products and services. Most important, new approaches to information retailing, along with a diminishing reliance on physical media, means that we no longer need to possess physical copies to enjoy information. (Younger users have almost no experience with physical copies.)

Though the Vernor court made no mention of these trends, in some sense, the decision simply acknowledges a fundamental change in how users now interact with software and other information products. Without the need to deliver a physical copy to the user, information transactions increasingly look less like the sale of a copy and more like a license to use a service. And the shift has been driven not by changes in the law so much as the availability of more efficient technologies.

The joy of renting
Still, the Vernor decision signals a dramatic change in the law. As EFF's Corynne McSherry concludes, the ruling "will go a long way toward ensuring that software buyers will rarely be software owners." Instead, vendors who simply utter the right "magic words" can "turn unsuspecting buyers into renters."

But what's so bad about renting? Those most anxious about the move from owning to renting information appear to be stuck in a mind-set that measures value solely by what one possesses. They see the technological, as well as the legal, changes away from ownership as an elaborate conspiracy aimed at depriving users of the full value of what they pay for information.

On this view, vendors are intent on gaining full control over access, use, and resale of information products and services. Once secured, they can charge whatever they feel like, change terms at will, and withdraw access when doing so is to their benefit. Treating information transactions as licenses gives them the legal cover, it is argued, to do that and more.

There's some merit to that argument. It's true, for example, that renters can't resell their copies or buy cheaper copies from the used market. And it's also true that hard-fought consumer protections embedded in the law of sales don't always have equivalents in the law of licenses.

But in practice, these differences haven't proven material. That's especially so in the software market, where used copies rarely substitute for an up-to-date version. Given the relatively low cost and rapid obsolescence of both the media (diskettes) and the content (Windows 98) due to improvements in hardware and functionality, the used-software market has never been as robust as it is for more stable forms of information, such as books.

(It's noteworthy that the Vernor case involved AutoCAD, a complex application that retails for about $4,000. Given that price, a secondary market even for outdated versions is more likely to thrive than one for, say, copies of last year's TurboTax.)

More important, whatever losses consumers face in the shift from buying to renting are greatly outweighed by the benefits that come from disconnecting information products from inflexible and rapidly obsoleted physical media.

In addition to eliminating the unnecessary cost of producing, transporting, and retailing the physical copies, electronic distribution enables software vendors to update and repair their products quickly and cheaply, and issue patches, upgrades, and new versions on much shorter cycles. Game producers can extend the life and usefulness of their products by offering downloadable new levels or host users in virtual multiplayer modes.

The elimination of media is having an even more dramatic and beneficial impact on the structure of the software industry. As the Internet continues to improve in speed, cost, and reliability, software vendors are evolving into service providers, offering a wide range of hosted services (many of which are supported solely by advertising, including Gmail and other Google services) that require little or no software running on the user's computer.

As this cloud model of computing continues to develop, the concept of software as a "product" is fading, along with the property rights mentality associated with "copies" and "ownership."

Beyond software, other information products are going through a similar metamorphosis. Apple alone has sold more than 10 billion song downloads through its iTunes Store. Amazon recently reported that it is now selling more "books" for the Kindle than hardcover copies. And on-demand downloads of movies and TV programs through cable and telephone service providers, or through third parties including TiVo and Netflix, are having a similar effect on the entertainment industry.

The liberation of information
And why not? Digital music can be played on any device. Electronic books can go anywhere the user goes. And streaming movies on demand is far more convenient than mailing discs back and forth.

Indeed, as consumers overcome their emotional attachment to ownership and embrace the rental model, subscription and rental services for all kinds of content are becoming increasingly popular.

That's because, as Kevin Kelly wrote in a 2009 essay, "Access is better than ownership." For one thing, ownership imposes ongoing costs to the user. Renters don't have to worry about replacing outdated media (LPs to tapes to CDs, e.g.). Subscribers need not bother with mundane tasks such as "backing up, sorting, cataloging, cleaning, or storage."

Renting, on the other hand, lowers distribution costs and enables universal access across the full spectrum of consumer electronics devices. And a licensing regime unleashes new forms of rental, subscription, collaboration, and other use-based models. Content creators are only just beginning to experiment with many of these new approaches, but the potential is there for rental plans that are tailored to the specific needs and preferences of individual users.

Even licenses that deny users the right to resell their copies can be beneficial. In the absence of resale markets, vendors can safely introduce a variety of new pricing models, including per-seat pricing for large business applications and the reduced prices students pay for the same products businesses pay full price to use.

At a more fundamental level, as Kelly and others have pointed out, the unique economic features of information made it a poor fit for physical copies and ownership all along. Free of physical media, for example, information can be universally accessed by an unlimited number of users, added to and improved on without the need and cost of releasing new versions. Until now, physical media was just a necessary evil.

Information ownership is not only unnecessary, in other words, but also an oxymoron. Media copies are dumb. Licensed uses can be and often are smart.

This is not to say there won't be some hiccups in the inevitable transition from information ownership to information rental. Some software vendors and other information service providers will do stupid things that conflict with their own long-term interests. Mistakes--including Amazon's clumsy handling last year of an unauthorized edition of George Orwell's "1984" it simply removed from Kindle users' libraries--will be made.

But here, as is often the case at the accident-prone intersection of innovation and law, it is the rapid pace of technological change, not the courts, that is driving disruptive changes. The law is simply trying not to fall farther behind.

So whether you think that information renting is a disease or its cure, the Vernor case will, in the long run, prove to be little more than a symptom.