Imposing technological solutions for Digital Rights Management (DRM) is, in theory, a viable way to stop the piracy of online content. In theory. In practice DRM presents a host of implementation problems and customer service headaches because legitimate content owners are punished or disadvantaged alongside thieves.
Because of the gap between current theory and practice, proponents and opponents of DRM attempt to dominate the DRM debate with apocalyptic rhetoric, political gamesmanship, and the kind of righteous indignation that is both an intellectual guilt trip and calculated lie at the same time. Neither side is really interested in what is practical or effective, or even in learning how pirated content is actually consumed by the end user.
The reason they are not interested is because they cannot sell the products they want to sell in a calm atmosphere. It takes fear of nuclear annihilation to sell both bomb shelters and bombs, so both sides in the DRM debate stoke the rhetorical fires and present their hard-line solutions as the truth, the light and the only way. All of which I’ve talked about here, here and here.
Still, it seems to me that there is one DRM question worth asking. Here’s how I referenced the issue in one of those previous posts:
Elsewhere in the clip, Mr. Doctorow makes a good point when he says that books will be copied and scanned regardless of the DRM that publishers employ. I agree. But that’s not necessarily saying something important. Scanned versions of books are almost inevitably going to be less clear than licensed e-books or even licensed digital copies. Yet even assuming someone cracks a DRM-protected book and makes it available for free, the book business has a kind of built-in protection against wanton piracy simply by nature of its content.
Music comes in small files. Movies are dumb — you just stare at them. Books, however, are big, and require active engagement over a long period of time in order to be consumed.
Maybe someone somewhere is downloading two hundred cracked e-books at a whack, then reading the first sentence of each in order to find a great read, but I think it’s unlikely. In fact, it’s unlikely that most pirated books are ever completely read, precisely because a book is relatively hard to digest. This means the ratio of thefts by end-users who intend to enjoy the content they steal to thefts by pirates who intend to profit from the content they steal may be lower in the book publishing industry than in any other medium. Which means the book publishing industry has more to gain by going after traffickers and less to gain by going after end-users than any other industry.
Before efforts are made to strengthen legal protections and prohibitions, or great sums are shoveled at the development of technological DRM solutions, it seems to me that each content industry would profit greatly by first asking the following question:
What are consumers actually doing with the content they pirate?
Not only do I think this question is important for each content industry, I think it gives insight into how the various content industries differ with regard to the net effects of piracy. For example, because songs are usually short — say, three minutes on average — and because music can be enjoyed passively while doing something else, it’s probably fair for the music industry to assume that pirated songs are consumed by the pirating public. In fact, it’s hard to image that a consumer would take the time to download a song that they wouldn’t at least listen to once.
But is that necessarily true of all content industries and products? I would argue that it is not, and that the publishing industry in particular is making a mistake in presuming (or being led to conclude) that illegally downloaded books are either fully or substantively consumed by the thieving public. Assuming we’re not talking about books-on-tape — which are essentially massive audio files — pirated publishing content simply cannot be digested in a distracted manner. You cannot read a book as you would listen to a pirated MP3 file or even watch a pirated DVD: the workload while reading is exponentially more involved, and there is no multitasking you can effectively do while reading for comprehension or enjoyment.
Too, if we’re talking about pirated fiction, there’s the additional requirement of creating and sustaining suspension of disbelief for the reader, and the open question of whether reading from a screen provides the same emotional experience as reading from a printed page. In fact, even as the market floods with e-readers from all quarters, everyone seems to agree that all of them — including the Kindle and the new iPad — are substandard reading experiences. (Currently, as both a test and an interest, I am reading Paul Carr’s freely-distributed .pdf of his book, Bringing Nothing To The Party, and I’m finding it almost unbearable.)
Given all these obstacles, are we really to conclude that downloaded books are consumed and experienced at the same rate as downloaded music? Books have miserable technology supporting them, while music is delivered with cutting-edge simplicity (if not cutting-edge fidelity). Books require dedicated concentration, while music is often so much background noise. Books demand language skills, reasoning skills, emotional sensitivity and a base level of cultural awareness, while music requires only functioning ears.
Several weeks ago, while reading some of the twits from the recent DigitalBookWorld (DBW) conference, I became convinced that this aspect of the DRM question deserves a closer look by the publishing industry. I tried to twit the question to people inside the DBW conference, but it was either ignored or lost in the storm. Fortuitously, however, I ran across the following twit the next day:
@ljndawson Stockfield: Even when we make the entire book available for free, people only view about 12 pages of it. #DBW
Imagine that you are publisher. You know that 10,000 copies of your latest bestseller have been downloaded illegally, but you also know that 95% of those pirated copies will never be read. Suddenly your effective losses shrink from 10,000 stolen copies to 500. Are you going to shovel the same amount of time and money at technological DRM solutions if your exposure is actually 5% of total pirated downloads? I think not.
As a practical matter, and repeating what I said in an earlier post, what’s stolen when content is pirated is an experience — whether that experience be emotional, visceral or rational. The digital file format that any e-content comes in is simply the delivery mechanism: is it is not the content itself.* If consumers are downloading the delivery mechanism — the file that contains the content — but they are not actually digesting the content, there is, as a practical matter, no negative effect on the creator of the content. From the point of view of protecting content, a stolen digital file that is never looked at is the same as a file that does not exist.
Again, imagine you write a book that ten thousand people download illegally, but no one ever reads past page five. Was something stolen from you? Legally, yes, but as a practical matter I would argue that there would be no advantage in trying to prevent these crimes from taking place. By the same token, if the publishing industry is worried about protecting its content, it needs to make a distinction between the ease of acquiring unauthorized electronic files and the difficulty of digesting the contents of those files.
Note, too, that the objective here is not zero losses, but acceptable/recoverable losses. Physical objects are also stolen, and anyone who owns a bookstore must dedicate a certain amount of time and energy to defeating in-store theft. Likewise, publishers are concerned when a crate of books goes missing, and put procedures in place to prevent such losses. Both the bookstore owner and the publisher know they can never cut their property losses to zero, but they do what they can to deter wholesale pillaging, then pass the cost of the remaining losses along to the consumer. The same approach — judging the cost of what is lost, and balancing the cost of defeating that loss with the cost of charging the consumer for other people’s thievery — should also be used in responding to the pirating of digital content, but currently that doesn’t seem to be the case.
Everyone agrees that it’s impossible to convert pirated copies to lost sales in any industry because the majority of people who pirate content would not necessarily buy that content if prevented from acquiring it free. I agree with that assumption. What both sides in the DRM debate also assume is that the total number of downloads equals the total number of consumer experiences of that content, and I particularly think that’s a faulty assumption for the publishing industry to make.
I also think that looking at the number of people who actually use downloaded content might provide a metric by which the number of lost sales could be determined, or at least more closely approximated. In effect, the more likely a person is to dedicate themselves to consuming an entire pirated book, the more likely they might be to buy that book if otherwise prevented from acquiring it for free. (Their determination to overcome the inherent obstacles of the content implies interest/demand.)
Not all books types are the same, of course. Pirated textbooks might be more prone to full use because students are required to complete coursework in return for a passing grade — which, at the college level, they (or someone else) are paying for. Similarly, some reference works might be said to be consumed even if only a small part of the content is digested. How-to books comes to mind, where only one section of a book might be useful to any given consumer because it meets that consumer’s specific needs. Fiction, on the other hand, usually needs to be fully read in order to be understood, and many nonfiction works (biographies) are only comprehensible when fully digested.
So what should the publishing industry do? Well, here’s what I would try if I ran a publishing house:
- Gather data on the use of pirated texts. What are end users doing with the content they steal? The goal here is to come up with percentages for various book types. E.g., 12% of stolen fiction is read to completion; 82% of stolen how-to books are consumed by end-users in a way that answers their questions about the subject matter.)
- Define losses only as content that is consumed by end-users. Equate a percentage of those losses to lost sales. E.g., if 19% of a given pirated book type is deemed consumed by the pirating public, a working assumption might be that 20% of that number (slightly less than 4% of the total) might have bought those books if the contents were not available free. Vary the estimate to see what the range of exposure is to those losses.
- Use these projected sales losses to quantify the need for DRM relative to physical losses which are already known. Calculate the cost not of eliminating either type of loss, but of bringing digital losses in line with physical losses. What is the best plan for doing so?
- Calculate the cost and ROI of preventing or prosecuting end-user theft vs. the cost and ROI of preventing or prosecuting third-party trafficking. Invest in the most cost-effective solutions.
At the end of the day, if pirated publishing content is not generally being consumed by the pirating public, it means that comprehensive DRM solutions for the problem are of minimal utility. If people aren’t using the products they download then you don’t need robust protection from consumer theft. Instead, you can afford to target the people who host, sell, or otherwise attempt to financially profit from trafficking in your stolen content.
* A footnote here to clarify the critical difference between digital content and physical content. If you steal a book or a CD you have stolen both the content and the costs associated with making that physical object. When you steal digital content there is no object to steal because it effectively costs nothing to create a digital copy of an existing file. People who say that content theft is it not theft tend to ascribe value only to the object: to the physical book. Because digital files are not objects they leap to the conclusion that nothing is stolen when a digital file is copied, but this is demonstrably false.
Stealing or revealing state secrets is punishable not because the state wants to discourage the theft of electronic files, but because the state wants to protect the contents of electronic files. For example, having information about troop movements is valuable apart from any electronic medium that might contain that information — or even if that information is never written down or recorded anywhere and giving or selling that information to the wrong person could be a crime. (All information has a market value, even if that value is most often zero.) For the record: you can steal objects, you can steal information, and you can steal objects that contain information, but it’s all still stealing.
— Mark Barrett