Setting the Wayback Machine to 1995: "Cheap Speech and What It Will Do": Music: The New System

[I thought it would be fun to serialize my 1995 Yale Law Journal article "Cheap Speech and What It Will Do," written for a symposium called "Emerging Media Technology and the First Amendment.) Thirty years later, I thought I'd serialize the piece here, to see what I may have gotten right—and what I got wrong.]

[a.] What It Will Look Like

I want to start by discussing how the new technologies will change popular music. These changes may be less politically momentous than the similar changes that I think will happen in print and video. But the music industry will probably be the one that changes most quickly; and in any case, many of the things I say in this Section-about cost savings, increased choice, information overload, and so on-will apply equally well to the others.

The reasons for the changes will be very simple: There's lots of money in them. The existing music distribution system is inefficient, both for consumers and for musicians. For consumers, in particular, it has three problems:

Cost: Music costs more than it could. Consumers must pay about $8 to $15 for a new album, though musicians generally see less than ten percent of this in royalties.

Choice: Consumers get a smaller selection than they could-many titles, especially ones that are relatively old or that appeal to fairly small markets, aren't available in most places.

Convenience: To buy music, a consumer has to take the time and trouble to go to the store.

And these problems translate into problems for musicians. High cost, low availability, and the inconvenience of buying lead to fewer sales.

These inefficiencies aren't the result of some sinister plot or even of market irrationality. They are an inevitable consequence of the existing distribution system. People today must buy music on some tangible medium, such as tape or CD. This means they generally have to go to the music store (inconvenient), which has limited shelf space (lowering the choice). And the tangible medium has to be created, imprinted, distributed, and sold (costly).

The infobahn, once it brings high-speed two-way communications to private homes, is a far superior way of delivering music to the consumer. It will work something like this:

(1) Using your computer-or perhaps your TV set, with a keyboard, a touch screen, a mouse, or even voice activation-you access an electronic music database. This database (actually, there'll probably be several competing databases) will contain virtually all the music that's available in electronic form.

(2) You choose the music you want, by album name, by song title, by artist, by composer or songwriter, or by genre. You might even ask the computer for suggestions, based on the artists or albums you tell it you like. (The suggestions will be derived from judgments entered into the computer by reviewers.) You can also browse in some way, perhaps looking only at music of a particular kind, or music that has gotten good reviews. You can then play the music, to make sure you really want to buy it.

(3) Once you decide you like it, you download the album to a digital recorder connected to your computer. Your bank account gets debited automatically.

This would mean:

Cost: Once the music is recorded-which even now can cost fairly little -the only significant other expenses will be advertising costs, royalties, the cost of electronic distribution, and the cost of the recording medium (which will be supplied by the customer). There'll be no need to spend money to create tangible recordings, ship them, and sell them. Assuming cheap electronic transmission (an assumption I'll try to support shortly), a CD-quality album may well cost as little as $3 to $5—a $1 royalty,  plus amortization of the recording costs and advertising costs, plus the $1 or $2 that the customer will have to pay for the recording medium. An artist who's willing to pocket less money to get more customers might be able to charge $3 or less.

Choice: You'll have close to the whole music library of the world at your disposal. Copyright owners will be able to sell to any infobahn-connected consumers, not just to the ones who have access to a store that's willing to stock the work. Because there'll be no shelf-space limitation-computer storage is cheap and getting cheaper-it won't matter how esoteric your tastes are; there'll be room for nearly everything.

Convenience: You'll no longer have to drive to the music store or wait in line. You'll also be able to select what you want more conveniently, because you'll easily be able to pre-listen to what you're buying,  and because you'll have readily available reviews. The copyright owners will benefit from this system too, because whenever consumers read a good review or like a song they hear on the radio, they'll be able to buy the music instantly, or at most have to wait until they get home.

[b.] Why It Will Look Like This

Music Database Operators: There's a lot of money to be made here. In the United States alone, there were over 475 million albums sold in the first six months of 1994, at an average suggested list price of over $10 each, including both the more expensive CDs and the cheaper, lower sound-quality tapes.  The sales volume should increase as costs go down, and the convenience of buying the product from home should raise volume even more. Skimming even, say, ten cents per transaction would mean, at today's rates, almost $100 million yearly.

Mail-order CD catalogs—including computerized ones, such as cdconnection.com, which is accessible either directly through your modem or through the Internet —are already the first step toward the system I describe. They attract customers by offering a large selection,  slightly lower prices,  and the convenience of home shopping (partly countered by the inconvenience of having to wait for the CDs to arrive by mail ). {Cdconnection.com, which lets you order CDs from your computer but delivers them to you through normal mail, boasts over 100,000 CDs, covering the entire catalogs of all U.S. major labels and more than 2000 independent labels, plus several thousand import CDs. I had looked for years for one album—Leonard Cohen's New Skin for the Old Ceremony—and finally found it there.} Direct downloading should provide even greater cost savings, selection, and convenience.

Setting up a music database shouldn't be much harder than starting a mail-order CD business today; and it should be much easier than starting a chain of music stores. Like the U.S. Postal Service, the telephone system, or the Internet, the infobahn should let any business be accessible through it.  The database operators will have to buy computer equipment and design some software, but this shouldn't cost much. Even a database operator who gets only, say, 1% of the total market can make almost $10 million yearly by charging a $1 markup (which will still save consumers a lot of money). Ten million transactions yearly—thirty thousand daily—can easily be handled even today with fairly cheap equipment. {Storage of the music shouldn't be a problem, either. While the central databases will need a lot of storage, disk space today costs less than $1 per megabyte, down from a bit under $2 per megabyte a year before.} And the low cost of setting up a database should keep competition high and consumer prices low.

Copyright Owners: There's also profit here for copyright owners. The new system will let copyright owners exploit markets that are closed to them now: people who would pay, say, $5 for electronic delivery of an album but not $10 for the album in the store (cost);  people who don't have access to stores that stock the album (choice); and people who otherwise wouldn't take the trouble to go to the record store, or who may want to buy an album they hear on the radio but forget about it by the time they get to the store (convenience). And the electronic database operators would easily be able to pay the copyright owners royalties as high as what the owners get from music store sales, if that's what it takes to get the owners to sign up.

This will become especially true when, as some copyright owners join, others will find themselves pushed by competitive pressures to do the same. Once even a few albums become available for $5 rather than $10, albums that sell for $10 will be at a significant disadvantage. Though music isn't fungible—loyal fans of New Kids on the Block might not think Tom Waits an adequate substitute—some product substitution will doubtless occur.

{Some people to whom I've described this theory have suggested that copyright owners won't license their music for electronic distribution because they fear copyright infringement. Once music is electronically available, the argument goes, people could buy it once and then upload it to some computer bulletin board, or sell bootleg copies via e-mail. Copyright owners will therefore be reluctant to allow electronic music distribution.

But while electronic copying is indeed a serious threat to music copyright owners, this threat will exist whether or not the music is available in some central database. Even without an electronic music database, a pirate could easily go to the store, buy a CD, and make many copies of it on his DCC or Minidisc recorder. The electronic databases might make commercial piracy more tempting, because they will increase the number of people with digital players, who are the pirates' potential customers. On the other hand, some of this effect may be counteracted by the lower costs of legitimate buying through the databases. Pirates will thrive more when their rip-offs are competing with $10 albums than with $5 albums.

The music industry's reaction to DAT (Digital Audiotape) recorders, the first digital-quality home recording medium, is worth noting. When it became likely that home digital recording could let people easily make CD-quality copies of music, the music industry and the DAT industry agreed to a compromise, embodied in the Audio Home Recording Act of 1992. The Act (1) legalized home noncommercial copying; (2) taxed DAT recorders and blank DATs, with the proceeds going to a fund to compensate copyright owners for the expected losses due to home copying; and (3) required DAT recorder manufacturers to design their recorders in a way that limits the possibility of large-scale copying.

A similar solution might be set up for the electronic music databases. The music industry may be able to push through a law that would compensate for possible copying losses by: (1) requiring an extra royalty payment on each electronic sale, and (2) requiring, say, designers of e-mail systems or bulletin board systems to put in checks that would make unauthorized copying harder. This is hardly a foolproof solution, but it might provide some protection for copyright holders while still allowing them to exploit a powerful new sales tool.}

Consumers: As I mentioned before, consumers can also benefit greatly from the new system. True, any change-especially one involving computerization-risks alienating customers, but the new system can be made very user-friendly.  The system needn't be any harder to use than an ATM; and, as with the ATM, which has probably saved billions of person-lunch-hours per year, the new system's benefits should be substantial enough that people will learn to use it.

Moreover, the physical advantages of music store layout-the ability to browse, and the possibility of stumbling over something good that one hadn't even thought of buying-could be made available on the home computer, too. The software could easily have a general "Browse" (or "Browse The Kind Of Music I Like") feature, if this is what users want. The software could also have other useful features—such as a convenient pre-listen mode, or cross-references to reviews—that many music stores don't have. {Cdconnection.com's main information screen asserts that its system contains 50,000 ratings from fellow customers, and 26,000 professional ratings from the All-Music Guide. You can look up the ratings easily as you browse through the CDs.} And if people really need human help, the software could, at the touch of a button, switch to a voice connection with an operator at the central database location.

{Some consumers might actually enjoy going to music stores, either because they like the atmosphere, because they prefer being around other people rather than closeted at home, or because they like to go to music stores with friends. Movies, for instance, haven't been entirely supplanted by VCRs, in part because of the better picture and sound quality, but in part because people like going to the movies.

But even if one sees going to the music store as a pleasure rather than an inconvenience, there will probably be few people to whom it will be worth the extra cost. People can socialize with friends, or mingle with strangers, in lots of other places. If home access provides a better deal on music, most people will just buy the music at home and do their socializing at a restaurant or the beach.}

Technology: The new setup will require a good deal of new technology, but what's not here yet is coming soon. Digitally recorded music is simply a collection of data, no different (from the computer's point of view) from your WordPerfect document. Music is already sent through the Internet. There's no reason it can't be sent down the infobahn to your home.

{Today such a transmission would take quite some time, because it goes over comparatively slow phone lines. The whole point of the infobahn, though, is to change all that. Data can be sent much more quickly through fiber optics than through wires; though no one is quite sure, people are generally expecting speeds tens of thousands of times greater than those of phone lines, which now generally ship about 1000 bytes per second. Cable TV wires can already be used for data transfers 50 times faster than those over phone lines.}

Once the music arrives—in data form—at your home, it will need to be recorded on some high-quality medium. Two familiar media are nonstarters: Normal analog audiotape is too low-quality, and today's CDs are read-only. But two recently introduced technologies—digital compact cassettes (DCCs) and MiniDiscs—might have what it takes. They both provide sound quality as good as that of a CD; you should be able to download music to them from your home computer, and then play it at home, in your car, or in your Walkman. Today this equipment costs a lot,  but prices are expected to fall as the technology improves and economies of scale kick in,  just as they did for normal CD equipment.  {MiniDisc recorders/players now cost $800; DCC recorders/players cost $400. Blank MiniDiscs start at $14 each, blank DCCs at $8.50.}

{Actually, you might even skip the recording step and instead play the music directly from a central database (just as you could today listen to music over the phone, only with much better sound quality). You might then be charged on a per-play basis rather than on a once-per-album basis. You'll still need to record the music, though, if you want to listen to it away from your infobahn connection, say in your car. And if transmission costs are high enough, you might prefer to pay them only once when you download the album, rather than every time you listen to it.}

To use this system, people will need computers, but there are already an estimated thirty million home computers installed today.  The cheapest ones now cost about $700,  and, of course, this money will buy you many more features than just home music delivery.

Moreover, even people who can't afford a home computer and a home digital recorder might still use the system through public music vending machines. These machines may cost more than the home versions, because they'll need to be more resilient (and probably more theft- and vandalism-proof); a fragile keyboard interface might have to be replaced by a touch screen interface, or by something similarly robust. Still, this shouldn't be especially difficult or expensive-consider ATMs, video games, and public lottery ticket machines. So long as there are millions of people who don't have home computers and home digital recorders, there'll be plenty of incentive to develop these technologies. Indeed, some music stores already have computerized music catalog machines (called Muzes)  that have these features-the music vending machines would basically be Muzes, plus an infobahn hookup, a credit or debit card reader, and a recorder.

Finally, it shouldn't be hard to charge people electronically for using the service. The system could ask for a credit card number when you access it-much as is done today for phone sales, or for cdconnection.com computer sales-and confirm it while you're shopping. Better yet, it could charge you through your infobahn provider, much as 900 numbers now charge through the phone company. Music vending machines could accept credit cards or debit cards. And even more convenient forms of electronic payment may soon be available; electronic payment could be a boon to many businesses, and the market demand for it has generated a good deal of research and investment.