Point-Counterpoint: On Digital Book DRM

There is increased interest among trade publishers in pursuing some sort of “interoperable digital rights management” (DRM) for digital ebooks. There are many unlikely allies, who think that achieving a little DRM encourages publishers to move into digital spheres, and gives them breathing room. I think this is a really bad idea, and I wanted to publicly detail a few reasons.

What I’ve compiled is largely a list of counter-arguments; there are many affirmative defenses for unencumbered content that could be promoted. I’ve also numbered these paragraphs; on re-reading, they more often than not meld and intertwine as a potlatch of thoughts, and have not taken to my weak organization very well.

In a separate post, my friend and colleague Bill McCoy from Adobe will attempt to establish his own conclusions about whether an ebook DRM standard is a useful compromise, or a fool’s errand. (Note 11/24/08: Bill’s post is now available here.)

Why digital book DRM is Bad bad bad:

  1. The embrace of content restrictions reduces sales, lowers price points, and reduces revenues. No one has ever made this argument better than Tim O’Reilly, and he said it nearly six years ago: piracy is progressive taxation. The take home: “Give the wookie what he wants.”

  2. DRM antagonizes customers. DRM solutions inevitably frustrate the ability of consumers to interact with content they purchase (or license — this being an important distinction and an issue addressed separately). I recently heard one senior trade publishing executive indicate that DRM will be particularly necessary for libraries, presumably because books are more widely in circulation within libraries, and one wouldn’t want the students sharing books too widely, would one? Suspending for a moment the fact that any respectable engineering student would see ebook DRM as fruit ripe for picking, let us instead turn for comparison to the market for licensed journal content, where in book publishing theory there should be articles zipping around P2P networks faster than hotwired bits, driving Elsevier into the fiscal ground. Ahh, nope.

  3. First sale. Increasingly, content industries are persisting to embrace forms of either DRM-protected or licensed content acquisition, as opposed to more permissive ownership. I find many things disquieting with this approach, not the least of which it weakens the bond that consumers have with their culture. While that might sound academic, I would suggest that there is an important and fundamental difference in my behavior when I know that I own something that provides me with insight and beauty, versus knowing that my engagement is brokered by a corporation. In May of this year, I wrote a blog post called “On Owning Books” in which I suggested that the implicit prohibition of the first sale doctrine – the ability that I have to gift books that I own, and for libraries to lend them – would be a serious and gross mistake by publishers. I hold my faith in the truth of those words.

  4. DRM restricts the future. Even if DRM can be made relatively inconspicuous and step on the user with a relatively soft footprint, our expectations of media use and functionality will change more rapidly than the bodies governing DRM specifications can adjust to them. DRM can only be built for the world at hand, along with glimmers of the world just now peeking over the horizon; the greater promise of our inventiveness and creativity will inevitably be prohibited by any DRM system. DRM that is a requisite ticket to play in a market that is inherently anti-competitive and frustrates system-breaking innovation. We have no idea what the book of the future will look like: how it will be combined with other media, integrated with other content, and premised on user interaction. Publishers think they will be selling “books” but the books they will be selling are not the books of today. As Sara Lloyd of Pan Macmillan UK has noted in A Book Publisher’s Manifesto:

    One of the key perception shifts that publishers need to make, then, is about the book as ‘product’. Whilst the book continues to be viewed as a definable object within covers, as a singular ‘unit’, publishers will continue to limit their role in its production and distribution, and this is a sure fire way for publishers to write themselves out of the future of content creation and dissemination. …

    Publishers need to provide the tools of interaction and communication around book content and to be active within the digital spaces in which readers can discuss and interact with their content. It will no doubt become standard for digital texts to provide messaging and commenting functions alongside the core text, to enable readers to connect with other readers of the same text and to open up a dialogue with them. Readers are already connecting with each other …

  5. Technically infeasible. All DRM, technically, is reliant on a “closed system” — in other words, a system where there are tightly controlled and protected inputs and outputs. Apple, for example, appears to be complying with the desires of content owners and embedding support for High Bandwidth Digital Content Protection (HDCP) in its laptops. This standard frustrates copying content as it moves across display connectors. The classic problem with closed system design is that the system architect has to permit enough permeability to allow the user to actually interact with the system; additionally, operating systems themselves are rarely truly closed, and there are myriad ways of utilizing lower level services to frustrate permissions expectations for higher level software and hardware systems.

  6. Technically fragile. DRM systems do not age gracefully. Because they are tightly coupled systems — a series of protections that must all be aligned in order to permit proper functioning — DRM is particularly sensitive to technical obsolescence. Thus the value of content secured by DRM is inherently reduced compared to its unrestricted use state, because its anticipated future value is less. In turn, that reduces the ability of a DRM protected system to maximize revenue generation over the content’s lifecycle — and for books, the potential value should last more than my lifetime, but be able to be handed down to my daughter, now 7 years old, and generations beyond that. Libraries are not just about access, but preservation for the ages. DRM breaks libraries.

  7. Sociologically unrealistic. There is a weird veil of unreality in DRM discussions because it presumes that the publishing market is coherent, stable, and interoperable enough to not only permit but encourage creation and compliance with a standard DRM specification. This is not the case. Just to name one obvious example: Amazon has few incentives to relax its own proprietary approach to the distribution of content through the Kindle. Arguably, DRM can only be envisioned as a market-wide system only if those sectors of the market likely to opt-out are not financially relevant. However, the organizational “field” of publishing — the complete set of firms involved in the publishing enterprise — is heterogeneous, and likely to be more so over time.

  8. Strategically unsound. DRM discussions are feasible when there is an identified digital asset that can be protected. That might hold true, as a characteristic, for something like a song, or a movie. Books are texts, and DRM will only work when they are downloadable objects possessing binary wrappers. However, as the Google Book Search system might suggest, increasingly people may well turn to networked access for books, in which reading a book becomes a set of HTTP-brokered requests. There are tremendous advantages to network based reading, and as networks become more pervasive and more mobile, it is hazardous to presume that they will not evolve into the most prevalent form of book “ownership.”

    Google will vend access to online books to consumers and institutions if the proposed settlement with the AAP and the Authors Guild is approved by the U.S. SDNY court, but they will be gating book access via account names and institutional IP address ranges — authentication and authorization — a very different approach from restricting behaviors over a book “object” with DRM. Google has reserved the option of selling downloaded content (for now, most often mentioned in a PDF format), but whether they would want to participate in a publishing-centric DRM scheme of download content is subject to some considerable debate. And, for what it’s worth, not mentioned as part of the settlement.

  9. DRM is unfair. DRM requires technically sophisticated systems. This would limit access to digital books by those who could most benefit from it, in the developing world. At the bottom of the pyramid, prospective readers are least likely to have robust network connections, up to date software stacks, and modern devices. DRM disenfranchises the most important part of the world — the portion of the world that most seeks change, on one hand, and is most likely to be the source of world-changing transformation, on the other. What works in Africa will work everywhere, and what will be created for Africa, by Africans, will be the light of the future for more of the world than we might imagine. Publishers, above everyone, have an obligation to embrace the world, supporting access to the world’s knowledge and understanding.

  10. A final bit of history: DVDs. Some of my publishing colleagues point to DVDs as a positive case for DRM, positing that the (weak) DRM on DVDs preserved a market and enhanced revenue. There are many problems with this argument, just one being that DVDs didn’t strictly embrace DRM, but rather combined a weak encryption scheme (CSS) with embedded region encodings that prevented playback on the most prevalent consumer equipment.

    Let’s remember some critical characteristics of DVDs:

    1. They are likely the last digital media to be produced successfully in a physical format (with Blu-ray expected to be nearly dead on arrival). DVDs have been successful because they are convenient to use when network bandwidth is limited, and because they are owned objects, not because DRM encouraged greater sales. In many ways, DVDs are more like (print) books than ebooks. The growth sector is now Netflix/Roku, and Hulu, not DVD.
    2. Region codes actually reduced the size of potential markets. Regional, geographic restriction is a concept unfortunately familiar to book publishers, who are just now beginning to see their geographic sale of rights beginning to erode. In practice, DVD region codes sharply reduced the potential world market. A personal example: my household is Japanese and English, and yet these are two different region codes (“2” and “1” respectively). With some effort, I obtained a dual-region player, but the region codes helped keep Japanese films absurdly expensive in DVD format within the U.S., and sharply limited my household’s consumption of movies that we otherwise would have been happy to acquire.
    3. The DVD encryption scheme (CSS) was ripped off almost immediately, and has achieved a lore and fame singular in the Internet’s intersection with media industries. It’s trivially easy to move DVD content to one’s computer, if one is so inclined.

In sum, I think DRM for digital books is a bad idea for publishers, authors, and readers, not just today, but even more importantly, for the future.

tags: , , , , , , ,