digital rights management

This comes up often and I grit my teeth every time.

You can’t have a discussion of any length about ebook sales and pricing and DRM in any sized group of digital publishing observers before you hear that it is somehow wrong or unfair that a “purchaser” can’t do everything with an ebook they’ve bought that they do with a print book they’ve bought.

That is: various “controls”, sometimes deliberate (DRM) and sometimes circumstantial (tech doesn’t always work smoothly) make it hard or impossible to lend, give, or re-sell an ebook in the same way that you do a printed book. Have enough of these conversations and you will become educated about “first sale” rights, which are enshrined in law, which basically say that when you buy something you own it and can lend, give, or re-sell it.

So the way the complaint often goes is that those damn publishers are putting this damn DRM on my ebooks so I can’t do all the things with them I can do with my print books.

This has always struck me as highly questionable on its face. First sale rights make complete sense with something physical. They make no sense with something digital. When you lend, give, or re-sell a print book, you don’t have it anymore. When you lend, give, or re-sell a digital file, you still have it and you could lend, give, or re-sell it again and again without limit. Surely, that’s a distinction that justifies a departure from the physical world paradigm.

The complaint that first sale rights are being abused — often delivered as a complaint about publishers — proceeds from a fundamental misunderstanding that publishers themselves are entirely responsible for creating. You don’t actually “buy” an ebook the same way you buy a physical book. What you actually buy is a license to access a digital file, which — in the developing world of the cloud — you may or may not ultimately “possess” in any machine or device you own. (Of course, you can own the machine or device, which is physical. If you lend, give, or re-sell it, you won’t have it anymore.)

Publishers promulgated this misunderstanding. From the beginning, publishers analogized ebook distribution to print book distribution. They started out using about the same retail price and about the same discount structure to intermediaries as they did with print books. Some, at the very beginning, even tried to make the royalties the same (in the neighborhood of 5 to 15 percent of the retail price.) It seemed simple and it seemed logical. It has turned out to be neither.

There is a core reason why publishers promote this nomenclature of misunderstanding. Publishing contracts vary widely, but one thing is pretty common among all of them and has been for a very long time. They enumerate the splits between publishers and authors on rights sale revenue for a long list of possible transactions: first serial, second serial, book clubs, paperbacks, cheap hardcover editions, foreign editions in English, foreign editions in foreign languages, and others.

And then they almost all say — almost forever have said — that all rights transactions not enumerated will see revenue divided between authors and publishers 50-50. In fact, according to some agents, even in contracts where an ebook royalty is specified, the sale of electronic book rights are almost always specifically designated as a 50-50 split.

So if publishers called their ebook transactions what I believe they really are — rights licenses — they’d have what looks to me (but I’m not a lawyer) like a contractual obligation to pay authors half the revenue. Since that is double what many publishers, and all the big publishers, think is “fair” and commercially viable, there’s no motivation to move the conversation back in that direction, even if it would make the consumer interaction, and the restrictions policed by DRM, sensible.

Of course, smart agents have been thinking about this question too. They see very clearly that ebook sales are different from print book sales. First of all, ebook sales are — almost without exception — governed by a contract between the publisher and the consumer’s source. That’s not true (with very rare exceptions) for relationships between publishers and print retailers or wholesalers. But it is true for the relationship between publishers and book clubs. In fact, the book club paradigm has much more in common with the ebook marketplace than the publisher-bookseller relationship does. Book club deals are covered by licenses. Book clubs “print” their own editions, just as ebook resellers deliver the books in their own proprietary format or DRM.

It is worth emphasizing here that the publisher is (in today’s world) very seldom delivering the file directly to the end consumer. The fact that the publisher gives the intermediary a clean digital file, which the intermediary then manipulates and copies (or, we could say, “prints” in its own proprietary edition) to deliver to its customers underscores that there is activity betwixt publisher and consumer that falls under a license. And it is a license that is spelled out in a contractual relationship.

But agents have apparently chosen, at least for now, not to fight the royalty battle with publishers on these terms. For any agent to do so would be employing a sort-of “nuclear option”; they might be right and they might even win in court (eventually), but they’d effectively deal themselves out of the game from the moment they attempted to enforce this position.

This is symmetrical with the publishers’ restraint on the non-compete clause. From the publishers’ perspective, it is transparent and obvious that an ebook edition competes with a print book edition of the same book. All book contracts have non-compete language. But no publisher has yet used that particular argument to strongarm an author who wants to self-publish an ebook when their print contract didn’t contemplate ebooks. Both sides — despite the flare-up that occurred last year when Andrew  Wylie appeared to go toe-to-toe with publishers for a little while before he apparently backed down — want to continue doing business and prefer to negotiate solutions rather than attempt to impose them, even if they have a very strong position.

And agents are aware that they and their authors might also benefit from the misunderstanding about whether the ebook transaction is a sale or a license. If it is a license that doesn’t explicitly grant first sale privileges (and, by the way, it actually often is already: check your Kindle agreement with Amazon!), then consumers might insist on paying less for it than they do for print. At least, that has been a component of their restraint. Now that ebook prices, most dramatically for hardcovers, have been coming down in relation to the print prices (first through Amazon’s deep-discounting initiative, and then made permanent by publishers lowering their established prices when they switched to agency), consumers are getting a dollars-off deal as compared to print much (though not all) of the time.

Even though authors don’t sell their copyrights to publishers (they license their use) and publishers don’t sell inventory or even production masters to ebook resellers (they license them to replicate and distribute the publishers’ ebook files), the fiction that Kindle or Nook or Kobo or Google or iBookstore is selling the book to you or me will persist. If we had truth in labeling here, it would make the restrictions comprehensible. It would even make consumers understand why Amazon was within its rights (and upholding its responsibilities) when it chose to “cancel” the licenses it granted erroneously for an edition of “1984″ a couple of years ago. We can all recall the high dudgeon among many observers when they infamously reached into people’s Kindles and erased a file they were given by somebody who did not have the rights to grant those licenses to it. But truth in labeling would also eliminate an ambiguity that works in favor of publishers’ margins today.

What would worry me if I were a publisher is that someday somebody who is not an agent trying to keep things sweet with publisher customers will file a lawsuit to make the case that ebook sales are licenses already covered in just about every publishing contract. That would suggest a potential liability equal to half the ebook revenue minus what has been paid so far on every ebook ever sold under any contract where that kind of rights split language still governs. Publishers have perhaps mitigated their exposure by putting new ebook agreements in place with many authors, but they still wouldn’t want a court poking its nose into this particular problem.

On the other hand, it would certainly make things a lot clearer and stop a lot of silly conversations if we all understood that ebook access is granted by license, not sale.

Looking forward to lots of hellos at Tools of Change this week. I’m sure, as always, it will be a jam-packed and stimulating couple of days.

Share this post

There is a lot of disagreement about piracy and DRM (digital rights management) among thinkers in the publishing space. This post will express a few thoughts about both but, mainly, this post is a plea not to conflate the two into the same discussion. In fact, whether they are part of the same discussion appears to be as contentious a point as whether piracy is a threat to publishers and whether DRM should be employed at all.

First, let’s define some terms. I make a distinction (which is not universally accepted) between piracy — which I would define as making a copyrighted file available for free access to anybody who comes along — and “casual sharing”. Casual sharing takes place between people who know each other; piracy takes place among strangers.

It has been observed by many for a long time that DRM does very little to prevent piracy, which is usually executed through web sites that host unprotected versions of content. It has been frequently demonstrated that DRM can readily be “broken” (I have two friends who routinely break it for sport: one in the US who isn’t in the publishing business and one in Brazil who is. Neither of them ever sell or transfer the jailbroken files, but they peel off the protection just to prove they can. And they say they always can.) In fact, books which had never existed in digital editions, like the Harry Potter series, are served up on pirate web sites.

You can scan a printed book and create a digital file pretty readily. There’s recently been a gadget introduced that provides a little automation for that capability. But you can buy content conversion commercially that will give you an ebook file from a printed book for low hundreds of dollars per title. So I would emphatically agree that DRM would do little or nothing to deter a pirate who has a minimum of determination to deliver a pirated ebook file, whether there was DRM or not; whether there was an ebook at all or not!

But casual sharing is another matter, or so it seems to me. People share published material all the time through email, usually by forwarding a link to something they want somebody else to see but sometimes by attaching a file or embedding text or images in the body of an email. Some people (my wife among them) maintain mailing lists of people whom they alert about one thing or another. This kind of person-to-person curation is the new automation-assisted word-of-mouth, and it is a critical component of modern communication.

So here’s what I think. I have no idea whether piracy helps sales or hurts them but, whatever it does, I can’t see how DRM prevents it. But I do think DRM prevents “casual sharing” (it sure stops me; and I think most people are more like me than they are like my friends who break DRM for sport) and I believe — based on faith, not on data — that enabling casual sharing would do real damage to ebook sales with the greatest damage to the biggest books.

Big general publishers survive based on the performance of their biggest books. Agents survive based on the sales of their biggest authors. So the biggest publishers and the biggest agents, if they see it the way I do, would be in favor of DRM even if does nothing at all to prevent the kind of piracy they attempt to cure with take-down notices.

There are a lot of good reasons to dislike DRM. It can make purchasing or consuming something harder. It is apparently responsible for the lion’s share of customer service costs for all ebook vendors. It can foil legitimate use by a legitimate purchaser. And it costs money and adds complications. In general, the more comfortable you are with technology, the more likely you are to be annoyed by DRM.

But it drives me a bit nuts when people attribute the belief that DRM protects against piracy to everybody who accepts the sense of using it.

So with this as background, I picked up a link earlier this week to an interview on O’Reilly Radar with my office-mate (but a man who very much runs his own business) Brian O’Leary entitled “What’s the current impact of piracy on the book publishing industry?” Brian has been trying for almost three years to measure the real effect of pirated editions (not casual sharing) on sales. His method is to watch the pirate sites for the appearance of books and then to measure the sales for the weeks before the pirated edition appears and the weeks after. If piracy is cannibalizing sales, one would expect to see a decline following the appearance of the pirate edition. If piracy is stimulating sales through additional word of mouth, one would expect to see sales rise.

Of course, the data to do this analysis can only come from the publishers and publishers, despite their often-professed concern about piracy (and their apparent willingness to spend a lot of money to track and combat it), have mostly not been willing to participate in Brian’s efforts to measure its impact. But what Brian did see (mostly through O’Reilly data, and O’Reilly is a DRM-free publisher) suggested that piracy might lift sales more often than it hurts them.

In the interview, Brian makes some very good points but then I get to this:

“I’m pretty adamant on DRM: It has no impact whatsoever on piracy. Any good pirate can strip DRM in a matter of seconds to minutes. A pirate can scan a print copy easily as well.” (I agree about the “good pirates”, but is the “no impact” statement data-driven? I doubt it.) But then:

“DRM is really only useful for keeping people who otherwise might have shared a copy of a book from doing so.” So, he’s against DRM even though he agrees it prevents casual sharing. And I’m not aware that anybody, including Brian, has ever attempted to measure the impact of casual sharing.

This is interesting, because he and I have exactly the same opinion about what DRM can and can’t do, but we don’t have the same opinion about whether it should be applied or not!

The point that Brian makes which I take to heart, though, is about trying to base opinions on data whenever possible rather than on conjecture. Many of his colleagues-in-arms against DRM attribute its continuance with ignorant and wrong-headed thinking: publishers and agents who somehow are deluded into thinking that by using DRM they restrain piracy. At the same time, concern about casual sharing is either ignored or elided.

And while gathering data about the true effect of piracy is difficult and gathering data about the potential true effect of unfettered sharing of commercial books is impossible, I am in a good position to gather data about what senior publishing executives and powerful agents believe about piracy, casual sharing, and DRM. So I created an informal survey to find out.

I asked three questions.

1. Do you think DRM is necessary to protect the sales of ebooks for popular titles?

2. Do you think DRM is an effective check against piracy?

3. Do you think the main benefit of DRM is that it prevents casual sharing?

I asked top executives in major houses and agents who handle major authors.

Nine executives and four agents (more than half the number I asked) were kind enough to come back to me with answers (so far). I’ll report on the findings in my next post.

Share this post

The one subject I didn’t touch in last week’s series of posts on ebooks was DRM: digital rights management, the software that controls what you can do with an ebook (or any other) file. This topic is so fraught with emotion and misplaced certainty that it has “third rail” aspects to it. So we tackle it today with the knowledge that we’re going to annoy many people: there’s no way to avoid it.

I hold two conflicting notions about DRM over time:

1. In the not-so-long run (5 to 10 years), we will be holding very little content in our devices or hard drives. We will access files — those we create and those we obtain — from the cloud. We will see only what we have license to see (as managed by our passwords, our iris scans, our fingerprints…) When that time comes, everything is, effectively, DRMed and, because we will all have our own private stuff up there, we’ll be damn glad it is and damn glad it works. Large elements of today’s DRM concerns will disappear (such as whether you, the purchaser, can access content on multiple devices); some other objections to DRM expressed today will become fights about the license, but not about DRM itself (lending your content or giving it to a friend.)

2. Also in the not-so-long-run, just about all of us will be in social networks that make file sharing (to the extent that we still have the files) with multiple users very efficient and very simple. When we’re all on Facebook and an unprotected file is posted, how many degrees of separation will there be between you and your friends and the entire world? Is it hard to imagine that every digital book would be available free on Facebook? Or through Facebook?

Both of these futures are within sight; very few people would say that either is impossible within a relatively short time. And both are very different from the world we have been living in for the past 15 years or so as the digital revolution has gotten started.

There is definitely a school of thought, which seems most widespread in the library community and among aspiring authors and aspiring publishers (those which are not, or not yet, making tons of money from selling content), that we should live in a DRM-free world. There are, broadly speaking, four lines of argument against DRM:

1. That it is commercially stupid, because it stops sharing and viral spreading of the word about content that will only increase sales. This is the “obscurity is a greater enemy than piracy” school of thought. Evaluating the scanty evidence about the effects of piracy for books so far would suggest that file sharing boosts sales more than it cannibalizes them. “So far” are important operative words.

2. That it violates the “first sale” doctrine, by which when somebody buys a copyrighted physical something, they can then do what they want with it, including lend it or sell it on to somebody else.  This argument is often couched in moral terms suggesting that the sellers of ebooks who put on digital controls are not just being unwise but also unjust (even though in the physical world “copying” is not something you’re permitted to do without paying for permission.)

3. That because of DRM, abuses occur such as people losing the use of files they bought (because they get a new device or computer and it won’t transfer or because the seller of the file, who was storing the backup copy, goes out of business or because, as happened last week, Amazon reaches into your Kindle and erases a book that they just found out they didn’t have the right to sell you.)

4. That it is futile because all DRM can be “hacked”. (Of course, more to the point, DRM can only raise the cost of getting an unlocked file: anybody can create one by re-keying or scanning and OCR-ing a text, the more expensive and cumbersome version of “ripping” a music CD.)

Let’s deal with these in reverse order.

Of course, all DRM can be hacked. The clearest evidence of this is that pirate sites carry books that didn’t ever have a digital file because somebody went to the trouble to scan or re-key them. There is pretty widespread agreement that DRM is like a lock on the door to keep an honest person out, not a security guard that will stop any interloper or thief.

I have been a longtime believer in what is called “social DRM”; the watermarking of information tying the file to its purchaser (or licensor). It is often said that those watermarks can be hacked off as well. True, but if the lock is to remind the honest person not to open the door, it would seem like social DRM should do it. Would you like a file with your name on it (let alone your phone number or your credit card number) on a pirate download site?

Using social DRM would make it easier (although not necessarily easy: interoperability problems are not all due to DRM) for you to share a book with your mother or your spouse, whom you could presumably trust not to spread your branded file far and wide. It would serve as a real deterrent to having the file end up on Facebook.

When Amazon erased 1984 and Animal Farm from their customers’ Kindles, it sparked widespread outrage. It properly raised the spectre of what a malevolent government could do in a connected world. That’s a big problem, but, in my opinion, not primarily an ebook problem.

We are headed for a world where our files are in the cloud and we need to be tethered to access them across our devices. The advantage to that is that we’ll have access to all our files in the cloud all the time on any device wherever we are. The drawback is that the cloud also will have access to our devices and that our files could be made inaccessible at any time. That’s a big problem that requires legal protection, but focusing on ebooks would really miss a much bigger point.

As for inaccessibility that results from device changes or people going out of business, I wonder where people making that argument have been for the past 40 years. Can you play a record on your cassette player? Can you load the program you bought on 5.5″ floppies twenty years ago on your new computer? We have been living with format changes that render our content or software impossible to use for the lifetime of most people living. Why should ebooks be exempt from a problem that existed even before the digital age?

It is absolutely true that ebooks reduce “first sale” flexibility. It is reasonable to say that an ebook “purchase” is not a purchase in the way we used to understand the word: it is a license with real limitations. And DRM is the tool by which the file creator and seller enforces those limits: enabling or disabling print or copying capability; allowing or forbidding some number of pass-alongs or use on multiple computers or devices.

But it is also true that digital files don’t “wear out” and books do. And books aren’t infinitely replicable for free (quite aside from any licensing cost), and unprotected digital files are. And the copying and printing you can’t do with a DRMed ebook file, you also can’t do with a book.

The argument that ebook pricing should reflect reduced useability is a reasonable one, although pricing is really decided by supply and demand, not by reason or rectitude. (History suggests that all new formats — from CDs to VHS tapes to DVDs — arrive at a premium price and it is ratcheted down over time.) The argument that ebook ownership and rights need to replicate the world of the print book is just that: an argument. And I don’t think it is an argument that would move me as a content owner if I believed that enabling that replication might also result in many potential purchasers of the IP just securing it for free.

From my perspective, the “commercial stupidity” argument against DRM is the strongest one of all. But I believe the evidence that supports the idea that it is stupid is about to become dated. Most of our ebook experience so far has been in what we called the “vision” stage of adoption: a time when very few people read ebooks. We have only recently moved into the “establishment” stage, largely enabled by the Kindle and the iPhone. The Kindle and iPhone are devices for the affluent and the Kindle, particularly, appeals to an older demographic. I can’t prove it, but I’d say the more affluent and the older are less likely to steal content than the population at large. (I don’t know an adult that downloads free and illegal music; I don’t know a millennial who doesn’t!)

So we have evidence from a world where, a) very few people read books on screens at all and b) those who do skew older and more affluent, that pushing out free copies — and indeed, the effect of piracy as well — tends to increase sales of a printed book. With evidence of what is really happening sketchy (although many people, I among them, believe the “obscurity is more damaging to sales than piracy” argument has held true so far), trying to attribute reasons for it is a pretty speculative exercise. But I would speculate that people are buying books of things they get free digital files of because most people don’t want to read digital files.As ebook uptake grows and, according to our paridigm of adoption becomes damn near universal over the next ten years or so, that will change. In an ebook-consuming world, a free ebook will satisfy the potential purchaser, not spur them to a sale.

There are ebooks available without DRM. Many publishers, including O’Reilly, Harlequin, and Baen, sell them from their websites. There are some non-DRMed files available from Kindle (according to my best source), but it isn’t easy to figure out which ones they are. Fictionwise once reported that as many as 50% of the ebooks they sold were without DRM (publisher’s choice), but we don’t know how that experience will port over to BN.com, Fictionwise’s new owner. Smashwords, the new open-source ebook developer and retailer (you send them your .doc file; they’ll put your ebook on sale at the price you want to charge) has no DRM option and they say they never will. But at least so far, Smashwords is for self-published content, not for big publishers or big authors.

My hunch is that the biggest authors will continue to insist on DRM and that they are sensible to do that. And that lesser authors will often be comfortable without DRM, and they are probably sensible to do that as well. But as the establishment stage of ebook adoption continues, I’d also expect that the “viral effect” of non-DRMed titles will stop being healthy for sales. This is an argument that still has a long time to run.

Share this post