Or, Why Your Kindle Isn’t Your Nightstand

Eighth Annual eCommerce Best Practices Conference
Stanford University, June 24, 2011

The second break-out session was wide-ranging, in part because it took a broad view of “virtual goods.”  A narrow definition of the term might be limited to “goods” that exist only in a virtual world, such as Second Life or World of Warcraft.  These “goods” are assuredly and obviously “virtual.”  But the essential insight here is that there is a whole category of “goods” that consumers think of as real but aren’t.  An obvious example of this is a book you buy for your Kindle.  Amazon works very hard to make it function like a book–an enhanced book, to be sure–but in many respects feels like a book.  If you don’t believe me, compare the experience of reading a book on Kindle with reading the same thing as a PDF.  It’s hard for me to put my finger on it, but when you read a PDF, you know you’re having a “computer experience.”  When you read a Kindle ebook, it’s easy to forget that.

The problem is that, if consumers are fooled into thinking a “virtual good” is a “real” good, they will assume they have rights that they really don’t have.  This creates the sort of conflict that Judge Fogel discussed in his keynote:  the disconnect between the technologists and consumers.  Oddly, in many cases, the more technologically sophisticated a consumer product is, the greater the disconnect.  This is because a goal of digital consumer products is to make the digital product feel like a familiar “real” product.  The machine hides its machinations.

This is where the First-Sale Doctrine comes in.  Legally speaking, the doctrine provides that once a product is sold, the intellectual-property rights* in the product are exhausted.  In other words, once the product is sold, the licensors of the IP in the product have gotten all they are going to get from that product:  it is out of their control.  Until that point, of course, the IP rights holders have tremendous control over the product:  who is allowed to sell it, to buy it, to distribute it, to import it, and so forth.

Actually, it’s not quite so clear-cut for trademarks, but there some functional equivalents in trademark law.

Consumers, I suspect, see this as a pretty fundamental property right:  the right to do with your stuff what you wish.  If they purchase a book under copyright, they can re-sell it, lend it, burn it, use it as a door stop, or whatever.  When they purchase, say, a blender under patent, they can re-sell it, lend it, destroy it, use it as a door stop, or whatever.  The copyright holder’s and patent holder’s rights are exhausted.

But when you buy an eBook from Amazon (or Apple or whoever), you don’t have this right.  That’s because you don’t own the book.  You merely license it, and you are subject to the terms of the license, which are unlikely to replicate your first-sale rights.  As Prof. Goldman bluntly put it:  What are the first-sale rights in virtual goods?  The short answer:  none.  Even if there were, the licensor could easily defeat those terms in the terms of the license.

(This wasn’t raised by the panel, but what if consumers found a way to easily lend their eBooks to friends?  Would the rights-holders sue them, the way the RIAA and MPAA sued file-sharers?  It is probably safe to say that consumers by and large have had little sympathy for file-sharers–witness the huge civil penalties being meted out by juries in file-share cases.  But the eBook situation seems essentially different.  Consumers have always known that copying music was legally dodgy, even as they engaged in it on a small scale in the pre-Internet days.  Sharing an eBook, by contrast, involves an activity that, heretofor, has always been legal.  Indeed, lending books is an expected and sociable activity:  “I read this great book.  Here, let me lend it to you.”  Considering that the prices for real books and eBooks on Amazon are pretty close, consumers can be forgiven for wanting to treat their eBooks like real books.)

But there’s more at stake than just the consumer-technology disconnect.  The loss of first-sale rights alters the market for virtual goods.  The First-Sale Doctrine puts two kinds of pressures on prices–one upward, one downward, but which do not necessarily cancel each other out.  On the one hand, the existence of a secondary re-sale market encourages purchases of goods because buyers know they can recover part of the purchase price of a item by reselling it.  On the other hand, the market for used goods provides an alternative for consumers should the new-good price get too high.  In theory, then, consumers should be more reluctant to purchase virtual goods because, if the consumer regrets the purchase, it’s money down the drain.  This ought to lower the price of virtual goods.  At the same time, virtual goods have no secondary market to compete with, which should prop up those prices.  It is not clear, however, whether even  sophisticated sellers of virtual goods have thought through how these two pressures interact.

The panel’s discussion was bookended by discussion of the remarkable decision, Costco v. Omega, and the more directly relevant decision, Vernor v. Autodesk.  (The Costco decision is fascinating and complex–see below for a capsule.**)  In Vernor, an individual tried to re-sell legitimately-obtained copies of Autodesk’s AutoCAD software.  Had Vernor been selling books, there wouldn’t have been anything the copyright owner could do.  But Vernor was selling software, and the strange thing about software is that you need to make a copy of it just to use it.  (It has to be copied from the storage media–hard drive or whatever–to the computer’s RAM for it to work.)  This means that it’s not enough to own the media on which the software resides (and even this is an old-fashioned way of looking at it, now that we more typically download software rather than purchase discs containing the software).  You must also own a license to keep making copies of the software.  Unless that license accompanied the sale of the software, the software was (legally) useless.  Vernor argued that the First-Sale Doctrine applied because he “bought” the software.  From a consumer’s point of view, he was the owner of each copy of the software.  Autodesk argued that there hadn’t really been a sale, but only a licensing of the software.  The Ninth Circuit held that resolution of the issue depended on the terms of the license, particularly whether and how the license restricts the buyer’s ability to transfer or dispose of the copy.  To no one’s surprise, Autodesk’s sophisticated license placed a number of such restrictions, and the Ninth Circuit ruled in Autodesk’s favor.

Combining Costco/Omega and Vernor, one of the panelists posed this fascinating hypothetical.  What if your car’s on-board computer used software that was subject to a highly restrictive license?  Could you sell the car?***  Everyone agreed that this was an absurd result,  But no one could really explain how Costco/Omega and Vernor wouldn’t prevent the sale.  The crown symbol on the watches in Costco/Omega were just a tiny (and functionally insignificant) portion of the watches, yet they affected the entire watch.  Vernor held that the First-Sale Doctrine does not apply to restrictive licenses of software

**Omega is very careful about how it distributes its luxury watches.  Like many manufacturers, it sells its watches at different prices in different countries, depending on a variety of factors–but in general, Omega is trying to set a price point that will result in maximum profits for that country.  For some countries, that price point is necessarily lower than for others.  Thus, the same watch might be sold in one country for much less than it is sold in the United States.  Costco, the warehouse retailer, purchased watches from an American company that, in turn, purchased the watches from foreign third parties who had purchased the watches from authorized Omega retailers in countries where the prices are low.  Thus, from Costco’s point of view, it had purchased the watches legally from an American company, and re-sold them at prices far less than Omega’s authorized American retailers could.

Under trademark law, there wasn’t much that Omega could do.  But Omega was very clever.  It imprinted on the back of its watches a tiny crown symbol.  Even though it was tiny, it was completely copyrightable.  Thus, Omega sued Costco not for trademark infringement but for copyright infringement–for distributing the copyrighted crown symbol without authorization.  Costco responded that it was protected by the First-Sale Doctrine, arguing that once Omega sold the watches, it lost its right to control the crown symbol imprinted on those particular watches.  The Ninth Circuit held for Omega, on grounds that the first sale had occurred out of the country (i.e., the initial sale to the authorized foreign retailer), and the First-Sale Doctrine (and the Copyright Act generally) didn’t apply outside of the country.

***  I think you’d need to add another assumption for the hypothetical to work.  I suspect that on-board automotive computer programs are firmware, which means that they live and operate in the same ROM chip.  Thus, no copying is necessary and copyright isn’t implicated.  But couldn’t a car manufacturer artificially invoke copyright law by placing the programming on some kind of high-speed storage, like flash storage, and quickly copying the programming to RAM?  In addition, how long until cars’ programming is so extensive that firmware becomes too expensive for it all?