Digital Music: Who Benefits?

Continuing on the thought-stream I started last week...

The problem is one of scarcity thinking. It seems to me (and quite a few other smart people) that the whole landscape of economics is shifting, from one based on scarcity, to one of abundance. Chris Anderson writes:

... the classic definition of economics is "the science of choice under scarcity". That's a warning sign right there. From Adam Smith on, economics has focused almost exclusively on behavior within constraints. My college textbook, Gregory Mankiw's otherwise excellent Principles of Economics, doesn't mention the word abundance. And for good reason: if you let the scarcity term in most economic equations go to nothing, you get all sorts of divide-by-zero problems. They basically blow up.

On the MIT Forum panel, Dave Dederer and Robert Acker both seemed to get this. Dederer pointed out that however the business model shakes out, it was all rosy for musicians who own their copyrights--those who haven't sold them to RIAA-associated record labels. Acker brought up Anderson's Long Tail concept several times, pointing out the benefits to consumers of having all this previously unavailable music available.

The other two panelists (wish the moderator had taken some time to express his views) seemed to be stuck in the old world view, a little bewildered at the amount of music "piracy" going on. David Weinberg, the Universal VP, seemed to love the ability to profit from their entire catalog of golden oldies. Put the old hits online, keep them under copyright, let anyone buy them from you, but sue the crap out of them should they share them without you getting compensated. And Bill Valenti of Melodio provides a great distribution network--distribute it via cell phone. People could still share their music, and get nickeled and dimed for the privilege. Because the whole platform is "trusted," whenever somebody shares a song with somebody else, the content kings in the sky know about it and can charge the recipient accordingly. And we thought we'd gotten past 1984 without Big Brother.

Bruce Schneier, a noted computer security expert, provides the security community's definition of trusted:

A "trusted" computer does not mean a computer that is trustworthy. The DoD's definition of a trusted system is one that can break your security policy; i.e., a system that you are forced to trust because you have no choice.

In other words, if you have a trusted device, the question to ask is who trusts it? Can you trust your cell phone to do what you want, without telling the rest of the world about it? Obviously, in this case, not. It's Melodio who trusts your cell phone, not you--trusts it to tell them when you share your music.

Let's take a look at the music industry through a different view: who is involved? We have several different players. Two stand out as the most important ones: musician, and listener. The others are just middlemen trying to make a profit by controlling the distribution of music between the musician and the listener, taking advantage of natural bottlenecks that create scarcity, and now trying to impose an artificial scarcity as those bottlenecks disappear.

Let's look at these roles closer, and assess their value in the economics of music.

1. Musicians. Most musicians who make a living from music, do so through concerts and shows. Very few make a living off of CD sales, even the most popular of them. In most cases, CD sales basically prime the audiences and provide a large number of people willing to buy tickets to hear the band live--and the band gets most of the proceeds of their shows, unlike of their CDs.
2. Filters. There are millions of musicians. In the age of record albums, CDs, and distribution of music on physical media, it's simply not possible to make a profit by recording all the musicians and putting them in record stores. Somebody needs to pick musicians to record, promote, and turn into stars. This has been the recording companies.
3. Producers. These are the people who record the music, making it available. These are the owners of the studios. It used to be that a good studio and high-quality audio equipment cost hundreds of thousands of dollars, putting them well out of the range of the amateur. No longer is that the case--as the cost of high quality recording equipment has come down, and as commodity PCs have learned how to mix music, anyone with a few thousand dollars and some time on their hands can put together a studio better than what was available at any cost a few short decades ago. Until recently, the recording companies owned the studios. Now anyone can.
4. Marketers. These are the people who promote musicians, turn them into stars. The recording industry is practically a star making machine. Who becomes a star? Who knows? It's not often the best musician, the best songs, the most presentable, or anything--it's just whoever is good enough who happens to be in the right place at the right time to catch the attention of the people with the marketing power--again, the recording industry.
5. Distributors. This is the network of distributors and music stores that present albums, singles, tracks, whatever, to the audience. There are several different distribution chains--radio stations, MTV, record stores, and now digital stores like Amazon, Apple iTunes, Rhapsody, and Yahoo!. Ultimately it's the distributor who collects the money from the consumer and delivers the goods, feeding the entire chain back up to the musician, at which point there's very little revenue left. The recording companies themselves have owned part of this distribution chain, with their CD music clubs and the distribution networks that deliver to the retail outlets.
6. The listener. Finally we're down to the people who pay for it all--you and me.

Notice that the middle part of this chain has been completely dominated by the recording industry--mainly by a handful of huge record labels. They've completely owned the filters, producers, and marketers, and a large part of the distribution chain. But times are changing.

Filtering is essential in an economy of abundance. When there's an overwhelming array of similar items, how do you choose? I used to sell skis for a living, at the REI store in Anchorage. We had dozens of different skis, in both cross-country and downhill categories. Our selection of cross-country skis were easy to sell, because there are enough distinctly different ways of cross-country skiing that we only had one or two options for a category. You could get a ski that was good for classic touring, or skating, or telemarking, or backcountry. You could get something that was okay all around, but not particularly good at any individual aspect. It was easy to place each ski model on a scale and define its strengths and weaknesses, because the scale was very clear, at least to me, and there were at most two choices for any position on that scale. I could easily talk to a customer and filter down to the right ski for their desires.

Not so in the downhill ski section. Yes, there were skis designed for cruising, and others for doing lots of sharp turns. (This was before the days of the short fat skis that dominate the racks now...). Some skis were stiffer than others, making them better for heavier skiers, or more aggressive ones. But for every point on this turns vs. stiffness landscape, we had four or five different models of skis. What's the difference? They're all slightly different, but which one is best boils down to taste, and frankly doesn't matter that much. I could not effectively filter down to an individual model for a customer, because there were too many choices. The best I could do was get them to the right category, see which color they responded to the best, and reinforce their decision that that set of skis would be great for them. It was essentially an emotional decision on the part of the customer--it didn't really matter which one they got.

Taking this to music, the recording companies have provided filters by creating specific genres, attempting to classify each artist into one (and only one) genre, based on what the music sounded like. They hire a few people as "experts" in these artificial genres to pick artists they think will sell, based on the arbitrary genres they just set up. And they cross their fingers, hoping that some music critic or teenager will love the artist enough to spark mass sales. If not, they roll the dice on somebody else.

It's not a very effective filtering system, but it's the best we could do, when the filtering had to come before the production, because of the scarcity built into physical production.

With production growing abundant, this is no longer necessary--we can produce the music first, and filter later. Recommendation systems like Amazon provide a mechanism for people recommending titles they particularly enjoy. Anyone can be a music critic, and recommend their favorites to their friends. Word-of-mouth has long been recognized as the most effective marketing strategy, and with an abundant music landscape, it has become more important than ever. People find music based on recommendations from people they trust. To truly influence people to buy a particular band, you need to be trustworthy--you need to have a good reputation. This is the essence of branding.

In this environment, if you were a musician, would you choose to hire a small marketing/public relations firm to promote your music, or a large record company who doesn't care much about your particular band, as long as they have enough stars in their stable of musicians? As I see it, the recording industry has virtually had a monopoly on production and distribution, which gave them a lot of power and control over filtering, marketing. As they lose their production monopoly, and people find other distribution means, they lose relevance in filtering and marketing.

Distribution, of course, is the big battleground right now. Millions of people have discovered the Peer-to-peer file sharing services, which completely bypass the recording industry's traditional locked-in distribution channels. Pundits keep saying that if the recording industry could come up with a new distribution model that gave people what they wanted, most people wouldn't resort to "piracy." I'm not sure that's true--the recording industry is fighting for relevance in all of its former roles, and the distribution monopoly is the only one they have left. If they can't convince both musicians and audiences that their distribution channels are the only way to legitimately get music, their whole business model collapses.

Okay. For the rest of this post, let's pretend that all music has to be distributed through sanctioned, RIAA-approved distribution channels. What distribution models are out there, what makes them work, and what's in it for the artists and listeners?

1. Physical distribution. That's what we have now. This model is becoming obsolete--digital distribution is becoming much easier, cheaper, and interesting.
   • For the recording industry, this model no longer works because they can't control what you do with the music once you've bought it--you can copy it to any digital device, share it with your friends, play it anywhere, and the recording industry doesn't get another dime.
   • For the artist, this model works for those chosen by the recording industry, but not for anyone else--independent labels of new musicians find it virtually impossible to get wide distribution in record stores.
   • For listeners, the main drawback is cost--it's expensive to buy a full CD of music if you only like one track. It takes a little effort to copy a CD to your computer to get it into your MP3 player or digital jukebox, and you can get sued if you share it with others--potentially even if you make a mixed tape to give to your love interest.
2. Apple iTunes model. This is the initial runaway success. Time will tell whether it will last for the long run. In this model, people buy a track at a time, for unlimited use on a limited number of devices.
   • For the recording industry, this system is great--until someone cracks the copy protection on the individual tracks. They get revenue from all the downloads of music, and if the user wants to distribute it widely, they get additional revenue. They potentially get paid for the same music from the same listener multiple times.
   • For the artist, this is also good, even if they aren't with a big label. According to Dederer, some 25% of their revenue already comes from the Apple iTunes store, because they were able to negotiate a deal directly with Apple. For artists with some marketing clout who own the copyrights on their music, this is fantastic. For artists working with RIAA labels, nothing much has changed. For new or unknown artists, if they get some good marketing or PR, there are more opportunities here than there was with physical distribution, because they can skip the big labels and get straight into the distribution channels.
   • For the listener, there are pluses and minuses. On the plus side, it doesn't cost much to buy a single track compared to buying a full CD. They can sample all kinds of music without spending more money. They can use others as filters to recommend good tracks buried far down the long tail of old hits from the RIAA catalogs. The big minus is that there are technological restrictions to putting the music on a bunch of devices. No more mixed tapes or CDs. No giving the music to your friend--you'd have to buy it for them. And, worse, sooner or later you might lose your entire music collection due to a hard-drive failure, theft, or any other technological calamity. What many people buying music from iTunes have overlooked is that they're allowed to copy their music to a very limited number of devices (somewhere around 3 to 5, not sure exactly how many). Once you've copied it to the limited number of devices, you're done--you can't move it to another device without buying the music again. This goes back to the "trusted" computing we mentioned earlier--you can only play this music on devices Apple trusts, and when you've reached their arbitrary limit, they've cut you off. Upgrade your iPod enough times and you've got to replace your entire music collection.

     When people start hitting this limit, the entire iTunes model may blow up. Yes, there are ways of cracking this Digital Rights Management (DRM) scheme, but doing so in the United States makes you a criminal, thanks to the Digital Millennium Copyright Act (DMCA). This isn't going to make people very happy.

   Melodio is using exactly the same business model, only they're adding more features to attract listeners to using their system, and using a system that's going to be harder to crack, meaning they can technically enforce their arbitrary rules more effectively than Apple. This benefits only the recording industry.

3. Subscription model. This is what Rhapsody and Yahoo are developing. In this model, subscribers have unlimited access to all the music in the distributer's catalog, during the time they're subscribed. The services can track who's downloading what, and pay the license owners based on the proportion of downloads for the music actually listened to.
   • For the recording companies, this is another good model--they essentially get a smaller payment for each song, but get it every time the song is played.
   • Again, for the artist, this arrangement is quite similar to the track download model. They get paid in an equivalent way, they can bypass the record companies and work with the distributors directly if they own their own copyrights, basically this is only a difference in distribution.
   • For the listener, it's "All you can eat" versus "you can't take it with you." As long as you're at the buffet, you can sample whatever you'd like. For heavy music listeners this is great, until you leave the restaurant--give up your subscription and you have nothing left over.

     Like the previous model, this solution depends heavily on DRM and trusted platforms. If you have a trusted music player, you'll be able to download music to it--but for it to be trusted, it'll need to quit playing that music when your subscription expires. This adds all sorts of technical complexity and sophistication to these devices, adding to their cost, reducing their reliability, and limiting your freedom to do what you want to do with it. Basically, with either of these schemes, you do what the distributors say you can do with the music, or don't listen to it at all. Or download cracking tools and become a criminal. There are slight variations in the rules you must obey, but you can pick and choose among these variations to get the plan that suits you best.

4. Media tax. The final "legitimate" music distribution model out there is what radio stations use, and it's also in place in Canada. Radio stations pay a fee to an industry association called ASCAP, for the privilege of playing music by ASCAP artists. ASCAP is then supposed to distribute the proceeds to the copyright holders (often the recording companies, who don't necessarily pass these proceeds back to the artists). In Canada, the federal government puts a surcharge on CD blanks, cassette tapes, and similar media people often use to copy and share music. Again, these funds are paid to industry associations who are made responsible for divvying up the proceeds as appropriate.

   In the age of digital devices and small physical media sales, you could still charge a tax on MP3 players, CD burners, ISP connections, and the various other means of transporting music.

   Do these schemes work? I really don't know much about them, beyond what I've just said. So here's more speculation:

   • For the recording industry, who are big copyright holders, this is a fair solution, though it makes selling their services to musicians more difficult. At least they get paid for the music they currently own.
   • For musician, the big question is, how do you divide these proceeds appropriately? With no real ability to collect usage information, there's no easy way to tell who is listening to what, or where. You might have to do something akin to the Neilsen television ratings to get some sample of the overall population to figure out how many people are listening to your songs, and get a piece of the pie. Under this type of system, there will be some disproportionate winners, and some disproportionate losers.
   • For listeners, this is the best possible solution. We get full freedom to do whatever we want with our music--copy it freely, use p2p applications to download and share it, mix it in all sorts of cool and wonderful ways, and give it away. On the other hand, people who don't listen to music as much, but buy the same services and equipment, still have to pay the taxes. This ends up being socialism for the music industry--everybody pays so that a group of people (musicians and the recording industry) benefit.

   Is this good or bad? It's great for music lovers, it's okay for record companies and musicians with recognized hits. For everyone else, there's a slight sense of unfairness about the scheme.

And that's about it. I don't know of any other music distribution schemes that provide any benefit to the entrenched interests of the recording industry. That's not to say there are no other models. In my next economic musing, we'll take a look at a completely different model, applying an open source model to music distribution.

I welcome your comments, and would greatly appreciate any thoughts you have on this subject. Please leave a comment or trackback on this post, and I'll get it out there as soon as I can. Unfortunately, I'm getting flooded with comment spam about online poker and home mortgages, so I've had to turn on moderation. If you're not spamming, your comment will be posted promptly!