As a consultant in the Digital Rights Management (DRM) space, I have worked with organizations that look at DRM from different points of view: media companies, technology vendors, licensing collectives, standards bodies, and attorneys. These various factions have agendas that are often difficult to reconcile, yet they must be reconciled if music and other media are to have a viable future online.

DRM-related technologies could theoretically evolve to the point at which they satisfy most of the concerns of media companies, digital media hardware and software vendors, and consumers. There will always be disagreements over such things as the definition of "fair use" or "fair dealing," but technology can get better and better at enabling uses of content that meet consumers' reasonable expectations without denying creators of content their due compensation. Hackers will always be following closely behind the developments in technology, resulting in threats to legitimate offerings that help ensure that media companies give customers reasonable value for their money.

Yet rather than outline a Utopian vision of the future of DRM technology, this paper will focus on the one factor that can really make or break the future for music companies as well as other segments of the media industry. This issue is one that receives scant attention in the press compared to stories about the latest free file sharing service, piracy arrest, or DRM technology announcement. The issue is: who will pay for DRM?

No grand vision of the DRM-enabled future stands a chance of becoming reality if no one will pay for it, yet that is exactly what is happening now. This paper will explain why this is a central problem and how the status quo came to be, and it will offer some suggestions for how the problem can be solved for the long-term health of the music industry.

Why Media Companies Won't Pay

There are many aspects of DRM that cost money, including:

• Invention and development of appropriate technologies.

• Development and promulgation of standards (whether open or de facto) that promote interoperability among different content distribution and playback components.

• Design of products based on good technologies and standards.

• Unit costs of playback devices or software that contain DRM technology.

There are also four general types of entities that could possibly pay for some or all of the above:

• Content owners.

• Content distributors, including internet service providers.

• Vendors of playback technology, whether software or electronic devices.

• Users.

The above are listed in descending order of incentive that they have to pay for DRM. Content owners have by far the most incentive, but even they have expressed reluctance to pay. Many DRM vendors have come and gone since the field came into existence in the mid-1990s. They came to media companies with solutions that may not have worked very well, but even if they did, the media companies did not adopt them (beyond small pilot projects) for several reasons.

First, the technology was often "raw" and required lots of customization and integration to work properly, and -- especially in the U.S. -- media companies did not have the internal resources to do the integration, nor did they want to spend the money to hire outside firms to do it. Some DRM vendors tried getting around this problem by acting as service providers -- that is, by running the server software that does the content packaging, delivery, and payment processing for the media company. The cost structure for operating this service is high, and many of these service providers tried offsetting it by charging media firms a significant percentage of transaction revenue. Media companies generally refused to give a vendor a portion of their profit margins, so that model didn't work either.

Another reason why media companies didn't adopt these early DRM solutions was that usability was a problem, and usability concerns usually took priority over technical ones, except in extreme cases of highly valuable and time-critical content. Of course, this attitude began to shift in the entertainment industry as piracy became more of a problem. But the final reason why entertainment companies refuse to pay for DRM is a feeling that various executives have expressed, that vendors should be required to protect against piracy and that charging media companies to do it is tantamount to blackmail.

There are really two reasons why media companies should be interested in DRM. Piracy deterrence is one, of course, but the other is to enable new ways for the public to consume entertainment and information, giving media companies new ways to make revenue and content creators new ways to get exposure for their work. Lately, the music industry in particular has been focusing on the former and giving short shrift to the latter.

Paying for Anti-Piracy

As far as piracy and DRM are concerned, media companies should look at them from a purely financial standpoint. They should be asking themselves how much money they really lose to piracy, and how much, in comparison, it would cost to subsidize the spread of DRM so that piracy is sufficiently reduced. That is the only financial calculation that makes sense, yet there is little evidence that music companies are doing it.

Current estimates of revenue loss due to piracy need improvement. IFPI's figure of US $4.3 Billion (EUR 3.7 Billion) per year has little credibility outside of the industry, there are no good estimates available of revenue loss due specifically to Internet piracy, and on the other side of the issue, there are no credible figures backing up the occasional claims that online file-sharing leads to increased legitimate purchases. While it is true that such estimates are hard to calculate, the entertainment industry trade associations need to work harder to produce better figures and publish them fearlessly, regardless of their magnitude. Only then can music companies get a good idea of how much money is worth spending to combat piracy -- through technological as well as prosecutorial or legislative means.

The truth is that music companies are also already spending money on technological solutions to piracy, whether they admit it or not. They spend money on the DRM technology that supports such record-company-owned online music services as MusicNet. In addition, they are paying vendors to develop technology that detects pirates' activities online, and they pay for other vendors to hamper Kazaa and other file-sharing services by putting decoy versions of popular music tracks up on them.

Music companies' money would be better spent on research and development of DRM-based music distribution technology that actually gives users a reasonable experience for their money and makes them want to consume music online. Music companies don't do this; instead, most are content to let R&D be done by technology vendors who have little incentive to build the right solutions and less than ideal knowledge of content owners' needs. We need look no further than CSS encryption for DVDs as an example of what happens when R&D is left in device makers' hands. CSS was invented by Toshiba and Matsushita, two device makers. Its chief virtue for them was low unit cost; it was poorly designed and easily hacked.

Alternatively, media companies lobby governments to create laws mandating or restricting technology, like the legislation introduced in the U.S. in 2002 at the behest of Disney and News Corp. (the Consumer Broadband and Digital Television Promotion Act), which fortunately did not pass. These laws are either too specific, meaning that they become quickly outdated, or too broad, meaning that they unfairly hobble technology companies and stifle innovation; plus, they are very hard to administer fairly.

Music and other entertainment companies should form a consortium that does the R&D and makes it available in the form of open standards and other freely licensable intellectual property, where possible. I am not aware of the laws regarding the formation of such entities in Europe, but U.S. antitrust law contains specific carveouts allowing such consortia to form. The best-known example in America is CableLabs, the research consortium for the cable television industry. The phrase "where possible" refers to the fact that a lot of core DRM intellectual property is covered by patents held by companies such as Sony, Philips, IBM, Macrovision, and ContentGuard, the latter being a company that spun off from Xerox in 2000 and is part-owned by Microsoft.

An R&D consortium could address a number of technical hurdles to the adoption of DRM, such as:

• Network identity: consumers, when they purchase digital content, reasonably expect to be able to play it on all devices they own and through all services to which they subscribe. Giving each consumer a single ID that works with all of his or her devices and services would help solve this problem, but consumers want to ensure that no one entity controls all such IDs, for privacy reasons. The Liberty Alliance, led by Sun Microsystems, is defining a solution to a more general form of this problem that is very difficult to solve in a practical manner. An R&D consortium could help solve this problem in ways that are specific to music companies' needs and more practical to implement.
  

• Personal networks: how does one define "all the devices they own" in the above, when cars can be driven by multiple people and the stereo is in the family room of the house? How can devices share identities and content use licenses? The industry needs models to help device vendors sort this out. There is some promising technology from IBM's Almaden Research Lab (xCP, or eXtensible Content Protection), but it is years away from practical application. Without coordinated research, the industry will get de facto standards from the likes of Sony and Microsoft.
  

• Rights descriptions: devices for playing and copying content should be able to read descriptions of content rights that include the rights on offer, the types of users to whom they are being offered, and the consideration (e.g., money) required to obtain those rights, all expressed in a standard format. There are a number of so-called rights expression languages (RELs) in existence, many of which are based on ContentGuard's XrML (which in turn is derived from research done at Xerox PARC in the mid-90s), but standards aren't really in place. A consortium could help ensure that such standards are developed and enforced for music (and other media) applications. Organizations representing sub-segments of the media and telecommunications industries, such as the Open EBook Forum, Open Mobile Alliance, and MPEG, are engaging in this type of activity already, but not in coordination with each other.
  

• License fee distribution: one of the major obstacles to getting music and other content online in legitimate ways is the enormous complexity of administering financial compensation from various types of licenses, such as performance, mechanical, and publishing. These licenses aren't going away, but some of the complexity of administration could be minimized through automation. A research consortium could help develop solutions that facilitate proper compensation while minimizing transaction costs and making the processes opaque to consumers. A handful of software vendors have been developing solutions to this huge problem, but they are addressed to individual media companies, not industry segments, which is where they belong; furthermore, licensing collectives do not have the funding or skills to solve these problems on their own. In Canada, an organization called the Rights Clearing House is attempting to build a solution to part of this problem.

Even if R&D is subsidized, device vendors still have to pay unit costs to build DRM technology into their devices; there is no such thing as effective DRM with zero unit cost. Yet there will always be non-copyrighted material for devices to play, and there will always be musical artists who want their music to be freely available so they can get exposure; therefore there will always be a legitimate market for playback devices that don't incorporate DRM. This will create a situation where devices with DRM cost more to build than those without DRM. Device vendors will want to pass the additional costs of DRM on to consumers, and the result will (from consumers' perspective) will be two tiers of devices, of which one does less but costs more than the other. This makes little sense.

Should media companies also subsidize unit costs? Here the answer seems to be no. Although there is some precedent for this -- early record companies selling record players cheaply in order to sell more records -- it doesn't seem practical, especially since media companies don't make playback devices nowadays (one notable exception, Sony, notwithstanding). Device makers do have some incentive to incorporate DRM: highly desirable content made available only in protected formats will drive demand for DRM-enabled devices.

Device Levies

The other choice is to level the playing field by raising prices of non-DRM-enabled playback devices through levies, the proceeds of which would be distributed to content rights holders. Device levies are highly controversial, but they may be the most reasonable alternatives. To ensure fairness, levies should only apply to devices that don't include DRM, as was suggested by the parties in the recent negotiation of the levy on PCs sold in Germany. Device levies act as proxies for appropriate distribution of proper payments for presumed uses copyrighted material on devices that have no ways of metering usage, which DRM-enabled devices presumably would. In this way, device levies would serve a purpose similar to blanket licensing deals enforced by collectives like Stemra for music reproduction and Stichting Reprorecht for text in the Netherlands, or their U.S. analogs, the Harry Fox Agency and the Copyright Clearance Center.

The big problem with device levies, of course, is how to distribute the proceeds. Some countries with levies, such as Canada and Greece, have had major problems with undistributed or allegedly misdistributed funds. This is a problem that will never have a good solution, but on the other hand, it gives media companies incentive to see to it that DRM is included in devices, because of its potential to track usage precisely and therefore compensate rights holders accurately.

Assuming that levy proceeds could be distributed fairly, how much would they actually help achieve parity between DRM-enabled and non-DRM-enabled device costs? A look at the recently introduced German levy on PCs shows that the answer is mixed. This past February, the German Patent and Trade Mark Office (Deutsches Patent- und Markenamt) decided on a levy of EUR 12 on all complete PC systems sold in Germany.

There are two ways of evaluating the appropriateness of this figure. One is to compare it to the actual unit cost of effective DRM, which knowledgeable observers would agree is well below EUR 12: so far, so good. The other way is to compare it to the cost of consuming media through legitimate means through one's PC. If one were to subscribe to an online music service like Tiscali Music and download the occasional movie and eBook, one could easily exceed EUR 12 in a matter of months -- making the non-DRM-enabled PC the bigger bargain, assuming of course that most of the user's desired content items were available through illegitimate means.

That brings us to the other reason why media companies, and especially music companies, should be interested in DRM: because it enables the development of new business models that include new ways for the public to hear music.

New Business Models

Before the Internet age, music fans had three choices: buy records, listen to the radio, or go to concerts. Each of those choices locks one in to a set of features, including how much choice one has in the music one hears, where and how one hears it, and how one pays for it. The Internet and other digital distribution networks allow music companies to explode those choices. Webcasting, subscription services, and even free file-sharing networks are examples of new sets of options for listening to music, and all are exciting.

The problem is that the public is very much used to the traditional ways. They can't be expected to jump to these new ways of listening to music without them being easy to use and without being told, over and over again, what value they have. The new types of services appeal to early adopters, those who are curious about new technology and willing to put up with the inconveniences associated with immature products.

As a technically minded person, media industry professional, and lifelong music fan, I am one of those early adopters. I have accumulated over 2000 albums at home, but the new online music services have helped change my listening habits to the point at which the idea of "owning" music has little appeal to me. I hardly ever listen to any of my albums anymore. My tastes are eclectic, but I know what I want to hear when I want to hear it. Commercial radio in New York City is horrible. So I pay for three different subscription services: Listen.com's Rhapsody in the office, XM satellite radio in the car, and Music Choice on digital cable television at home. I would love to pay for one service I could access in all three places, but that's not possible yet, at least not in America. I have never burned a single CD in my life -- not because I don't know how, but because I don't want to be bothered handling physical media anymore. I have looked seriously at the new generation of internet-enabled audio components, but I find them all over-engineered; I am not interested in spending my time organizing a music collection on a hard disk drive -- or anywhere else, for that matter. To be quite specific, I am willing to spend money not to have to do that.

Meanwhile, the early successes among new online music services are the ones that do little more than emulate traditional ways of consuming music. Apple's iTunes music service, which only allows simple downloads of tracks and some entire albums, is really an online version of a record store. Satellite radio in the U.S. is another example: it merely emulates traditional radio, though with better music selection and fewer commercials.

One could reasonably say that it is too early for mass marketing the subscription services, because their feature sets and prices have not been field-tested long enough. Yet music companies have failed by letting critics in the press control the dialog, with their pejorative talk of "renting" vs. "owning" music. It is not about renting vs. owning, it is about paying for a service that gives people good value for the money by supporting their preferred ways of listening. Selection, sound quality, service reliability, recommendations, and artist information are all worth paying for, and none are found on the free file-sharing services. It will take the industry time to figure out what the right price is, but in general, music companies should be spending money on marketing the value of these services, and they are not doing so.

Digital rights management is one of several enabling technologies for the new breed of online music offerings. It facilitates delivery of music to authenticated users, and it does so in a way that guards against piracy and helps ensure that artists are compensated. The point is that part of paying for DRM is paying to market new types of music offerings, so that the public is interested in paying for them. It is a much bigger change than moving to a different type of physical media for recorded music, such as from LPs to CDs or from 8-tracks to cassettes.

Conclusion

Music companies have to change their attitudes about paying for DRM in order for it to become mainstream technology. Free file-sharing services may be stifled by prosecution or technology, but they will always exist in some form. DRM solutions designed by and for consumer device makers will never meet media companies' needs, but that's what the media industry will end up with if it does not control the research and development process itself. Piracy will prevail, and both music consumption models and new artist development will stagnate, if music companies do not offer alternatives that are desirable for consumers and protect the interests of artists and record labels.

Music companies must control the design and proliferation of technology that meets their needs. It is the control that costs money, not just the technology. The amount that sufficient control costs is the real measure of the impact of digital technology on the music industry's bottom line. It is time to start calculating that cost and then paying it.