 It will come as no surprise to you that the entertainment industry is currently in crisis. During the last few years, the music industry has suffered from a rapid deterioration of its sales, the sales of recorded music both in the United States and in the rest of the world, and in its revenues. This disease has not yet severely infected the film industry, but most observers, certainly most of the major participants in the industry, recognize that they are next in line. One of the major topics we'll be considering during the live portion of the iLaw program is what to do about this crisis. How can we address the problems that afflict the industry while preserving the substantial potential benefits of the technologies that have destabilized it? But in order to fairly assess the alternative possible solutions to the crisis, it's crucial that we understand the character of the crisis, the sources of the problems, and the ways in which the law and the relevant business people have sought so far largely unsuccessfully to deal with the threats. That's the purpose of this lecture, is to provide the foundation. The date is important. It's February 6, 2004. We're filming this lecture in Cambridge, Massachusetts, where it happens to be a terrible day outside. Freezing rain and gray snow is the order of business. My hope is that when you watch this lecture, the conditions are more pleasant. The reason for emphasizing the date is that some of the issues we discussed today will likely change in the next couple of months, and so we may have to update a few of the arguments when we meet you in person. So here's an outline of the program for the next approximately hour and a half. We'll begin with a discussion of the structure of the entertainment industry, focusing on the state of the law and the associated business models circa 1990, which is the moment at which technological challenges began to destabilize the traditional state of affairs. Then we'll turn to those technologies, describe what they are, and examine both the potential benefits they hold out and the problems that come with those benefits. And finally, the longest portion of the lecture will be addressed to what happened between 1990 and the present, how successive waves of technological innovation simultaneously held out potentially large benefits to consumers, while on the other hand, threatened the revenues available to the principal intermediaries in the music and film industries and the ways in which those industries have sought to respond. That's the plan, so beginning at the beginning. The American entertainment industry, first to be clear, for our purposes here we're going to refer, when talking about the entertainment industry, to the business for producing recorded music and film. It may help to provide some context to discuss how much of these commodities are currently consumed in the United States today. Well, here are the most current figures from the Census Department. Roughly speaking, the average American adult will, in 2004, watch over 800 hours of broadcast television, another 800 hours of cable and satellite television, 74 hours of videotapes and DVDs, either purchased or rented, 12 hours of movies and theaters, almost 1,000 hours of radio programming, 250 hours approximately of recorded music and other formats for a total of approximately 3,000 hours over the course of the year. That's about eight hours a day, 365 days a year. So it's something of a puzzle how people are able to watch so much, watch or listen to so much entertainment and still get their work done. But it appears that the explanation comes from two sources first, that many Americans spend substantial portions of their late afternoon and evening hours watching television and while they're doing other things, they tend to listen to recorded entertainment in the background. Still, eight hours a day, 365 days a year is a lot. One of the reasons for emphasizing this is as we were going to see as we go through these materials, the total financial value of the entertainment industry in the United States is not enormous, but its cultural value is enormous. Just by way of comparison, to give you a sense of how the manner in which Americans entertain themselves has been changing, here's the data on likely consumption during 2004 of other types of information. Americans will read, spend about 24 minutes a day reading newspapers, about 17 minutes a day reading magazines, 14 minutes a day reading books, and 17 minutes a day playing video games. The first three figures are going down and the last figure is going up. That's the United States. It's worth noting that the trends, although not quite so dramatic, are in the same direction in other countries. In Brazil, for example, even the poorest dwellings in the favelas surrounding the major cities typically have television antennae sprouting from their tops. Highland, a particularly grim recent study, suggests that in the northern provinces many families are now selling their daughters into sexual slavery. A recent study of this practice concluded that two-thirds of these families could afford not to do so, meaning that they did not need to sell their children in order to live, but preferred to buy color televisions and video equipment with the proceeds. So where does the stuff that fuels this consumption come from? Where do we get recorded music and film? To answer that question requires first a survey of copyright law, the relevant provisions of copyright law, in 1990 to repeat the moment when this technological revolution first took hold. In every piece of recorded music, there are two copyrights. First the copyright in the musical composition. That first copyright comes with four entitlements, four traditional entitlements captured in the United States Copyright Act in section 106. First is an exclusive right to reproduce the composition. So the holder of the copyright, in the first instance that's the composer, has an exclusive right to reproduce it in, for example, sheet music format or in mechanical copies, tapes, records, CDs, and so forth. Second, the holder of the copyright has an exclusive right to prepare derivative works. New arrangements of the music, for example. Third, the copyright owner has the exclusive right to distribute the product to the public by sale or otherwise. And finally, the holder of the copyright has the exclusive right to publicly perform it. So for example, if I play on my guitar a folk song, if the performance is in my living room to a small circle of friends, there's no legal problem. But if the performance is public, say in Harvard Square with my guitar case open for contributions, that's a violation of section 106 of the statute, a violation of the public performance right in the composer. That's the first of the two copyrights. Until 1972, as a matter of federal copyright law, it was the only one. But since then, there has been a second copyright in every sound recording. That's not surprisingly the recording, the copyright in the recording itself, held in the first instance, typically by the performer, although sometimes shared with the producer. Traditionally, certainly as of 1990, that copyright carried with it three entitlements, not four. So the performer or other holder of the sound recording copyright has an exclusive right of reproduction and an exclusive right to prepare derivative works. If you want to get very detailed, it's slightly different. It's slightly narrower in shape than that enjoyed by the composer, but similar otherwise. Also, the holder of the copyright in the sound recording has an exclusive right to distribute the recording, typically, again, by sale or gift or rental to the public at large. Interestingly, in the United States, one of the legs of this diagram is missing. So you notice here that the holder of a copyright in a sound recording does not enjoy an exclusive right of public performance. In most other countries of the world, certainly in Europe, for example, this entitlement exists, but it did not in the United States. The main significance of that gap is that when a radio station broadcasts a sound recording, it has to pay, through a series of licenses we'll discuss in just a minute, fees to the holders of the copyrights in the compositions, but it does not have to pay fees to the record companies, which typically hold the copyrights in the sound recordings. American record companies have long complained about this gap, and as we're going to see, a version was supplied in 1996 and 1998. Equally important, just as important as the entitlements we've just reviewed, are a series of exceptions and limitations. So you can think of the structure of the American copyright statute as, on the one hand, giving copyright owners an extensive list of rights, while with the other hand, withdrawing, cutting out of that bundle of entitlements a series of exceptions, limitations, and safe harbors. First of these is the first sale doctrine, which operates as a limitation on the exclusive right of reproduction. Roughly speaking, with some exceptions we won't pause to consider, section 109 of the American copyright statute says that, once the copyright owner has authorized the distribution of physical embodiments of his creation, say in our instance, a CD or a DVD, the first purchaser thereof can do whatever he or she wishes with it, can resell it, can rent it, and so forth. Second, cluster of limitations on these entitlements are known as compulsory licenses. So what a compulsory license consists of is an arrangement under which the law on the one hand gives the copyright owner an entitlement of some sort, one of the sorts we consider this far, but on the other hand, forces the copyright owner to surrender that entitlement to a third party in return for a state determined fee. So here's an example, is public broadcasting systems enjoy a compulsory license to publicly perform musical works without permission provided they pay the copyright owner's modest fees, an entitlement not enjoyed by commercial stations. Another example, the so-called cover license embodied in section 115 of the copyright statute provides that after a phono record of a musical work, by that meaning a tape or a CD, has been distributed to the public, any other person can make additional phono records provided that she pays a compulsory royalty, currently 8 cents per song per copy, unless the songs are long, and doesn't alter the melody or character. The last of the exceptions that we're going to consider here, the most general and the most important is the fair use doctrine. In just a minute I'm going to provide an explanation for how that doctrine works in the music industry, but for now we'll concentrate on its doctrinal structure. All right, putting these pieces together, how does the music industry operate? How are these entitlements deployed in practice? We'll begin with the composer, who writes a song, ordinarily the composer will assign the copyright in the composition to a music publisher. The publisher will then make money partly for its own purposes, partly to be passed along to the composer through a series of licenses, a reproduction license issued to a sheet music printer, a import license to a record importer, issued through and with the assistance of a major intermediary in this field, the Harry Fox Agency. Synchronization licenses, again issued with the assistance of the Harry Fox Agency to movie studios, and most importantly, mechanical licenses, again negotiated by the Harry Fox Agency, issued to record companies, which then combine the composition with the performance by a performer to create records. So that piece of the puzzle is applied by the recording artist, who as I indicated a minute ago enjoys a separate copyright in the sound recording, which the recording artist typically assigns to the record company. So notice the record company in this chart is both a licensee of the musical work copyright and an assignee of the sound recording copyright. That's not the end of the story. And the publisher, music publisher, will also issue licenses, public performance licenses through a different set of intermediaries in the United States, three of them, BMI, ASCAP, and CSAC, which collectively issue blanket licenses to radio and television stations who wish to broadcast music, restaurants provided they're large enough, and other establishments engaged in public performances. BMI flows through the system, through the channels cut by the legal rights in the opposite directions, so upward through the diagram. The only major exception to the correspondence between the revenue streams and the legal rights is this one, Payola, currently at a level of approximately $100 million a year in the United States, is money paid by the record companies, typically nowadays through independent promotion firms, to the radio and to a lesser extent television stations, to induce them to broadcast songs. Why do they do this? In order to inform the consumers of the availability of new recordings and induce them to buy more. All right, that's a quick sketch of the music industry. To repeat, this is 1990. Now let's turn to film. The law in this area is simpler. There's typically only one copyright, not two. That's the copyright in the motion picture. The holder of that copyright, the holder of the copyright in the motion picture enjoys all four of the entitlements we've discussed thus far, an exclusive right to reproduce the film, to prepare derivative works, which would include things like sequels or abridgments, the exclusive right to distribute it to the public, and the exclusive right to perform it publicly. Again, this entitlement is qualified by a series of exceptions and limitations. So the way in which this set of legal rights finds expression in business models is by organizing the affairs of the many people and institutions that combine to create a movie. They include a novelist, who ordinarily assigns his or her copyright, or the right to prepare a derivative work based upon his copyright, to a producer. Screenwriters, actors, directors, and composers who all enter into work for hire agreements, special kinds of contracts under which they contribute to the creation of the final movie, but surrender, assign their entitlements to the producer who aggregates these rights. Location owners, who, this is a debatable point in the industry, who, as a matter of custom, issue releases to the producer, releasing rights to include their buildings or parks in the background. Record companies, who issue master use licenses, and music publishers, who issue synchronization and performance licenses to the producer. So this elaborate array of rights is aggregated by the producer. Who is the producer, after all? Well, in some cases, the producer is an independent person or organization, but more often, nowadays in the film industry, the producer is an arm of a major motion picture studio. In circumstances in which the two organizations are separate, the producer typically enters into a perpetual worldwide distribution agreement, allowing the studio to distribute the film. How then does the studio distribute the film? Well, again, as in the music industry, it's done through a series of licenses. Major licenses are first, rental but plus performance licenses to theaters, and to HBO, the airlines, and other organizations that follow quickly in the wake of the theaters. Similar licenses to television networks, a business model we're going to come back to in just a minute. Sales of copies of the movie to video stores, which then rent them. Compulsory licenses governed by a separate provision of the copyright statute to cable companies. Mechanical licenses to record companies that prepare sound tracks. Derivative work licenses to merchandise manufacturers and advertisers. Again, money flows upward through this system, through the channels cut by the legal rights. Now, I promised a minute ago we'd come back to the fair use doctrine and illustrate how it operates as a substantial limitation upon copyrights in both the music and the film industry. The reason for emphasizing this doctrine is to come to play a substantial role in the struggles that happened between 1990 and the present. So here's the case that brought the fair use doctrine to bear most dramatically on the movie industry. Before I describe the facts, here's the doctrine itself. Section 107 of the copyright statute sets forth four factors that accord. When confronted with a fair use defense, must assess when deciding whether the defendant's behavior, although on its face illegal, should nevertheless be excused. And those four factors are first, the purpose and character of the defendant's activity. So for example, if the defendant is engaged in a commercial behavior, that counts against him. If the defendant is engaged in a transformative activity reworking the copyrighted material, that's favorable. If the defendant is engaged in making a parody, that's strongly favorable. Parodies are not automatically excused under American copyright law, but they come close. Finally, under this heading, a court is instructed to consider the propriety, the decency of the defendant's conduct. That's factor number one. Factor number two is the nature of the copyrighted work. Certain kinds of copyrighted materials are deemed susceptible of greater than average protection, others lesser than average protection. Specifically, fictional works are thought to be more strongly shielded than factual works. And unpublished works are given more protection than published works. So, if the defendant has copied a unpublished piece of fiction without permission, he's quite unlikely to be able to invoke the fair use doctrine. Whereas if he has copied or distributed or modified in some way, a published factual work, like a news broadcast, he's much more likely to be able to invoke the fair use doctrine. Third factor, how much did the defendant take? This is both a quantitative inquiry, literally how much did the defendant take, and a qualitative inquiry, how important to the plaintiff's work was the portion appropriated without permission. Fourth and finally, the impact of the defendant's behavior on the potential market for the plaintiff's work. Now, this is a subtle, tricky, much debated issue in copyright law. The subtlety, the ambiguity arises out of the many possible meanings with the phrase potential market. Some courts, as we will see, have taken the position that potential market means an average, ordinary, customary, or predictable market. Roughly speaking, the sort of thing that a plaintiff enjoying a copyrighted work would naturally or be expected to exploit. At the opposite extreme, some courts have taken the position that a potential market consists of any group of people who would pay for access to the material in question. And as we'll see, the doctrine differs significantly in application depending on which of those approaches one takes. All right, now, return to our chart here. This, to remind you, is the structure of the movie industry, circa 1990. Concentrate on the business model highlighted by the dotted rectangle here. The relationship between the studios and the television networks. So if we blow that section of the chart up, here's what we get. One of the ways to repeat that the studios make money is by issuing licenses to television networks to broadcast over the public airways their copyrighted films. The television networks release the information, release the programming in question to the public at large to viewers without charging them, but instead by embedding in the programming advertisements. Viewers pay, not the television networks, but instead the advertisers by paying higher prices for or simply buying more of the products advertised in the programs. The advertisers take a portion of this increased revenue stream and pay the television networks advertising fees, and the television networks in turn pay license fees to the studios. This long established familiar business model was destabilized in the late 1970s by the introduction of videocassette recorders originally by the Sony company. Sony sold VCRs to the general public and the members of the general public paid for them. What's the problem? The problem is that some members of the public use these new machines to avoid the advertisements contained in the programming that they recorded using their new machines and then watched later. This is old technology, but in the late 1970s and the beginning of 1980s, advertisements were commonly, not always, but commonly avoided in one of two ways. Either one member of the family would sit in front of the television when the show was broadcast live and press the pause button whenever an advertisement came on, enabling the other members of the family to watch the program free of ads. Or the program would be recorded automatically by VCR, by the subset of Americans who are capable of programming them and then replayed at a later time. And then the people watching would hold their thumbs on the fast forward button to speed through the advertisements. Either of these techniques, the advertisers believed would cut into the willingness and ability of the viewers to pay increased prices for, or larger volume of purchased products. The advertisers threatened to reduce their advertising fees accordingly. And the television networks threatened to reduce the license fees. They'd pay the studios. The studios, understandably unhappy with this state of affairs, looked around for some way to protect themselves. What could they do? Well, they could have brought a copyright infringement suit against the individual customers who were making copies of their shows on their VCRs. The behavior of the viewers, at least at first glance, appears to be a straightforward violation of Section 106 of the Copyright Statute. The viewers are making verbatim copies of entire copyrighted films. But the studios didn't want to pursue this course, in part because it would require bringing suit against millions of individual users, a very expensive proposition, and in part because, at least until very recently, the providers of entertainment thought it was a bad tactic to bring suit against their own customers. So instead, they brought suit against Sony. Formally, the suit asked for an injunction against continued sale and a manufacturer in distribution of videos cassette recorders. In other words, the students wanted to shut down the trade in VCRs altogether. Is that what they anticipated doing? No, had they won the suit, they almost certainly would have licensed the manufacturers of VCRs to continue engaging in this business, provided they paid a fee. So in practical terms, had the suit come out differently, you would have paid in 1985, $350 instead of $300. Something of that order of magnitude for your machine. The claim the studios made against Sony was not direct copyright infringement because after all, Sony is not making any copies of copyrighted shows. The claim rather was contributory copyright infringement. The argument of the studios was that Sony is facilitating or encouraging its customers to engage in illegal conduct. So the case wound its way up to the courts, took a long time. Finally made its way to the United States Supreme Court in 1983. Supreme Court was unable to decide it in its first year, held it over yet another year, and finally issued its opinion in 1984. The holdings of the final opinion are first, the manufacturer of a device that can be used to violate the copyright laws. In other words, that can be used for both legal and illegal purposes. Is liable for contributory copyright infringement if and only if the device is not capable of significant non infringing uses. All right, that's the standard for contributory copyright infringement. And then the Supreme Court held that one of the major uses of VCRs, namely time shifting, is a fair use. So these were two verbs, Jerren's actually invented in the case. Librarying consists in the Supreme Court's view of using your VCR to record a program off of the airwaves, and then put the tape of it on your shelf, hence the term librarying, enabling you to replay it repeatedly. That, a majority, perhaps all, members of the Supreme Court thought was illegal. The second major use of VCRs, which the Supreme Court called, following the arguments of the council in this case, time shifting, consists of recording a program once when you're not home and replaying it at a separate time. Supreme Court concluded that the latter use, although on its face, a violation of section 106 of the statute was nevertheless excusable as a fair use. So we return to the four factors discussed just a minute ago. The Supreme Court held that time shifting is a non-commercial activity, and therefore, under the first of the four factors, tilts in favor of the defendant. To be sure, time shifting does involve copying a fictional work, which is bad for the defendant, and involves copying an entire film, which again is bad for the defendant. But finally, factor four, the crucial, most important of the factors. The studios had failed, at least in the original trial record, to demonstrate that this behavior had any adverse impact on their existing traditional markets. So you notice here that two of the factors tilt for the defendant, two of the factors tilt against the defendant. But in light of the special strength in the Supreme Court's view of the fourth factor, it held that time shifting is a fair use. Because time shifting is a fair use, then plainly a significant, a major use of VCRs is lawful, and therefore, Sony, the manufacturer of the VCR, cannot be liable for contributory copyright infringement. That's a quick summary of the Betamax case, and we're going to come back to it several points in this lecture when discussing, for example, Napster and other modern cases involving claims of contributory copyright infringement. Okay, against this backdrop, we can return to our survey of the structure of the entertainment industry. This combination of legal rules and business models worked pretty well. Certainly, it was sufficient to enable the American entertainment industry to generate enormous amounts of material. Gradually, during the late 20th century, certainly through the 1980s, the American entertainment industry was expanding, expanding relative to that in most other countries. For better or worse, American popular culture is ubiquitous in the world today, largely as a result of the systems of law and business we've considered thus far. It was not a perfect regime, it had some flaws. These flaws would not have been sufficient, almost certainly, to provoke a change, but they're worth noting. Worth noting because when we consider how the entertainment industry might be rebuilt, it would be nice if possible if these disadvantages could be avoided. So what were these flaws? Well, first, high transaction costs. There are many players in both industries, and the cost of negotiating and managing the various transactions among them were substantial. Not prohibitive, clearly, but substantial. Next, an inequality in distribution. Both the music industry and the film industry in 1990 conformed in their patterns of distributing income to what economists refer to as a skew log normal distribution, fancy term for a pattern in which a very few number of players make enormous amounts of money and while the large majority of participants make modest incomes. So in the music industry, a few stars make quite a bit from their recordings, whereas the majority of musicians don't. So what explains this inequality in distribution? Two things. First, technology has, since the beginning of the 20th century, amplified the power of the people we're going to refer to as stars. So if one soloist is slightly better than another soloist, in the absence of technology for capturing their performances and distributing them to the world, they will enjoy roughly equal incomes. Because they will have roughly equal opportunities to perform live for audiences. But the technology of capturing well the performance of the two soloists dramatically increases the power of the slightly better one. Because if you can hear a slightly better recording, why would you purchase a slightly worse one? So technology has, throughout the 20th century, amplified the financial implications of the gap between the best and the not quite so good. The second contributing circumstance that helps to explain the inequality in the distribution of the fruits of the two industries is the concentration of both of them. So these are current numbers, but they were not much different in 1990. Today, five record companies control approximately 83% of recorded music. Seven film studios control approximately 90% of the film industry in the United States. And the firms in both industries have, for almost a century, collaborated commonly, typically tacitly, to avoid antitrust penalties in setting prices and arranging business practices. These things together, these circumstances together, create reasonably high barriers to entry. It's hard for newcomers to enter either market. So why is the concentration of the industry relevant to the inequality and distribution? It's because both industries, the principal firms in both industries, have come to the conclusion that they can make more money by producing a relatively few number of expensive products. Highly produced records, or blockbuster films, promoting them and then promoting them heavily, rather than producing a larger number of less expensive products. And the concentration, therefore, in the kinds of material generated by both industries amplifies the power of the stars who would appear in those relatively few products. Why should we care? Why was this troublesome, this inequality? Well, some people at the time and some people today regarded as unfair. The extreme gap between the winners and the hangers on in the two industries is troublesome from an equitable standpoint. And second, some economists, not all economists, believe that a regime that rewards heavily the few winners results in overpopulation of that industry because it draws an excessive number of people into the business, each of whom hopes typically overvalues his or her chances of becoming a star. And the last of the flaws in the traditional business model is homogeneity of content. As of 1990, the number of albums, as of that day, there were CDs, produced in the music industry each year had been going down, not up. In the film industry, over a period of about 30 years, the total number of major motion pictures released in the United States had been stable, despite the fact that the size of the viewing population had been steadily rising. So the array of products produced in both industries was either fixed or shrinking despite the fact that the market for both was growing. One important side effect is that there was less diversity than you might have expected or then had existed a few years earlier in the music and films available to the public. All right, that's the end of the first third of our analysis. We're now going to turn to technology. How had this traditional arrangement of legal rules and business models, how has it been destabilized by technology? We'll begin by discussing the technologies in question. What are they? Well, they're familiar to most of you. First, digital recording and storage systems. Music, sound recordings are now conventionally distributed in CD format, a digital format, no longer in analog formats of traditional tapes or long play vinyl albums and similarly movies when they're distributed in packages to consumers are now most often distributed in DVD format rather than analog VHS or Betamax tapes. That's the first of three technological changes that together have created this both opportunity and crisis. The second is codecs, compression, decompression systems. The most common by far in the music industry today is MP3. A common one, not quite so common one in the film industry is Divix. What decompression systems do is they enable digital recordings to be shrunk. For example, MP3 recordings occupy approximately 112th the space of a uncompressed file and that in turn makes it easy to transmit them over the internet and to store them on hard drives. The third familiar component of the technological revolution is a mechanism for distributing these now compressed digital recordings and that mechanism is the internet. So here's some data. As you can see in the left hand column, the households in the United States with internet access have been rising rapidly over the last four years. Currently approximately 70 million American households, that's roughly two-thirds of American households have internet access. Households with broadband access, either cable modems or DSL accounts, have been rising even faster. Currently 29 million approximately American households have broadband access and the rate is going up as I say, even more quickly than internet access in general. Here's a chart I find intriguing as a test of the common hypothesis that only teenagers surf the web. So this is data collected by the US Department of Commerce. It shows age distributions in internet usage in four successive time slices. So the bottom line indicates age distribution in 1997. As you can see in the left hand side of the chart indicated this arrow here, there's a bulge corresponding to the teenage years, meaning that a higher percentage of people make use of the internet if they're between the ages of approximately 13 and 20. But the working population, roughly between the ages of 20 and their 50s, is still fairly high in their usage of the internet. Rates of usage don't drop off dramatically until retirement. Now what happens in each successive year is, roughly speaking, the usage rates of the regular working population match those of the teenage population in the preceding year. And it keeps ratcheting up. And we can expect the trend to continue. Don't have data on 2003, but anecdotal information suggests the trend continues. In the world at large, the spread of internet access has been, in percentage terms, even more rapid. So these are data through 2002. You can see the extraordinary rates at which internet access has been growing already, more people connect to the internet in Asia than connect to it in the United States. All right, those are the three technological shifts that have been accelerating since 1990 that together have made possible dramatically new, and we're going to suggest, superior ways of gaining access to and enjoying recorded entertainment. And somewhat more specifically, three new ways of gaining access to and enjoying entertainment. The first is downloading, which consists of the transmission over the internet of a digital recording followed by storage of that recording on the recipient's computer enabling it to be replayed repeatedly on demand. So downloading enables people to collect permanent copies of recordings on their computers or other portable devices analogous to the collections people have had for many years of albums, tapes, and CDs. The second of the techniques is interactive streaming, which consists of at the recording of a recipient transmission over the internet of a digital recording, which is then played but not stored. So you can think of this as the internet equivalent of pay per view. And finally, non-interactive streaming is the same process, not at the request of the recipient. In other words, when the choice of programming is made by the broadcaster rather than the recipient, non-interactive streaming is roughly speaking the internet equivalent of traditional radio broadcast. These techniques, each of them harnessing in some way the combination of new technologies, offer us simultaneously large benefits but also some serious problems. So here are the benefits. The first, most dramatic, most easily seen is cost savings. So I'll focus here on the music industry. The suggested retail list price of a compact disc in the United States now is approximately $18. How's that money divided up? Well, the largest share of the pie goes to the retailer for an undiscounted CD. The retailer earns approximately 38% of the $18. A significant slice, 8% goes to the distributor of the CD, the organization that takes the physical disc from the manufacturer to the retailer. The green zones in this chart all represent different aspects of the income that ends up with the recording company. 14% goes to overhead charges in the recording company. 5% goes to the ANR department. So the folks in the record companies that locate artists sign them up and help them select their songs. 8% goes to marketing. 8% goes to manufacturing the discs. 1% goes to record company profits, a surprisingly small slice. 12% formally goes to the artist who recorded the CD. This is a figure we have to return to in just a minute. 4% goes to the music publisher, the holder of the copyright in the composition, and approximately half of that money is then passed along to the original composer. Alright, back to the recording artist. From the recording artist's formal share is typically deducted a series of charges. The cost of producing the CD, the cost of packaging it, a deduction for free goods, a deduction for breakage, which is a leftover of the time in which the containers in which recordings were housed, namely shellac 78s, did in fact break and transit, a deduction for reserves, recording costs, and a portion of the promotion costs of the CD. Bottom line is that the recording artist ends up with potentially less than 12%. How much exactly? Well, it varies by the number of records distributed and by the bargaining power of the musician, somewhere between zero for a starting unknown artist and 7% or 8%. Against this backdrop, the new technologies for distributing sound recordings offer us very substantial potential gains. So a little bit more specifically. Imagine here that sound recordings were distributed, not in the traditional, now traditional way of distributing physical CDs, but unencrypted copies distributed through the Internet. In such a regime, we would, first of all, save the cost of the retailer because this motive of distribution does not require any retailer. Copies are made instantaneously by the server to and provided to viewers on demand. For the same reason, you save the cost of the manufacturer because there's no physical disc that has to be manufactured, no disc and no associated artwork. Past this point, there are opportunities for cost savings, but they're a little more difficult to see. To recognize them requires a brief survey of the functions that record companies currently perform. So what are those? Well, five things. Record companies do five things. First is mentioned a minute ago. They select artists and help advise them on their repertoires. Second, they assist in the production of sound recordings, typically by hiring independent producers. Third, they promote the sales of the recordings. The most dramatic and expensive form of promotion is, as I mentioned a minute ago, through Payola. Fourth, they distribute the recordings to retailers. And finally, they engage in risk spreading. You can think of record companies as functioning in part as a funny sort of insurance system in which the majority of the insured have accidents. So in the record industry, the majority of products, majority of records, don't make any money. They lose money, and what the record companies do is take a portion of the profits from the winners and redistribute them to cover the losses of the majority of losers. So against that backdrop, the new technologies offer potential for cost savings in four of these five functions. So specifically, studio costs are dropping dramatically now because of the new digital recording systems and rapidly declining costs of the associated software. In promotion, record companies no longer need to depend as heavily on radio promotion. There are now more effective promotional vehicles available, including above all advertising through the internet. Distribution through the internet is much cheaper than the distribution of physical copies. Not free requires maintaining servers, but cheaper. And finally, because of the rapidly dropping studio costs and promotion costs, there should be in the new environment fewer losers. A smaller percentage of sound recordings should lose money and therefore the need for the insurance function will decline. Now turning back to our chart, that means that there are opportunities for cost savings in this slice, the distributor slice, this one, the marketing slice, and this one, the A&R slice. So notice, roughly speaking, two-thirds of the costs associated with distributing music from the manufacturers to consumers have been saved in our hypothetical new technology. That in turn makes possible either increased revenues for writers, meaning songwriters and artists, decreased prices for consumers, or both. So that's the first, and as I say, most dramatic of the potential benefits of the new technologies is by no means the only one. Next, the new technologies make possible elimination of the long-standing problem of over and under production of physical embodiments of recordings. Because the recordings are made, as I say, instantaneously on demand, you don't have to worry about producing too few or too many DVDs and CDs. Next, the new technology is substantially more convenient and precise as a mechanism for distributing recordings. It makes recordings available instantaneously. Consumers can get individual songs, not albums containing a fair amount of filler. They're able to use the new technologies to sample recordings, to try out recordings and see whether they like them. And finally, as Paul Goldstein, one of the scholars that works in this field, emphasized long ago, the new technologies make possible what he referred to drawing out of the writers as a celestial jukebox, a situation in which consumers could through any device connected to the internet obtain at any time any recording ever made. That's been technologically possible for quite a while. The impediment to our achievement of this is the legal system, which we'll consider in just a minute. Fourth of the potential advantages of the new technology is that it will increase dramatically the number of musicians or in the film industry filmmakers who can make a living and make their work available to the world at large. Why? Because the costs of producing recordings are going down, so more people can afford to do it. And in part because the internet makes possible, makes it possible for small-scale producers of sound recordings or film to reach and make enough money to earn a living from, niche markets. So suppose I produce an esoteric kind of music. There are a few people in Boston, a few people in San Francisco, and a few people in Singapore interested in hearing it. In the absence of a global, cheap, global distribution network, I won't be able to exploit any of those markets because each individually is too small. And in the aggregate, they're big enough to provide a market for my work large enough to enable me to make a living and therefore to induce me to apply my trade and make my stuff available. Partly for that reason, we get the fifth of the potential advantages in the new technology, namely cultural diversity. The array of stuff that will be available to consumers using these new technologies is much greater than the array available using the old systems. Already we can see this happening in the webcasting context. So try some time going, if you haven't done it already, to live365.com. It's the largest aggregator of webcasting stations currently available. You'll find there somewhere in the order of a thousand folk stations, a thousand jazz stations, a thousand classical stations, a thousand hip-hop stations and so forth. In this array of stations, you will find some that resemble closely the standard fare currently available on the radio. But you'll also find an extraordinary array of more specialized material. For example, under the folk heading, you can find the Makosa Jukebox offering 100% Cameroonian music, Paradise Sahemian specializing in Neo-Pagan music, Radio Free Boonie offering programs for little kids reading their stories. A great virtual babysitter if you're interested. Why do we care about this? Why is the diversity suggested by the emergence of this webcasting array important? Well, in part because it enables us to satisfy more precisely the taste of consumers, not all of whom like the fare that's available on Clear Channel. And in part, a somewhat more subtle reason, emphasized long ago by John Stuart Mill, the more multifarious the lifestyle's ideas and art forms on public display in a society, the more each of its members must decide for herself what to think and how to act, thereby developing her own mental and moral faculties and rendering the culture as a whole even more rich, diversified and animating. The sixth and last of the potential benefits of the new technologies for distributing entertainment is in my judgment the most intriguing. The scholars who work in this area refer to it using the slightly misleading label semiotic democracy. What that phrase refers to is the possibility that we would see a reversal of a trend that's characterized most Western cultures in the 20th century. That trend has been ever greater concentration in the power of producing cultural meanings and fewer and fewer record companies, film studios, advertising houses, political consultants and so forth. The internet in general has long been held out as an arena in which the power to create meaning would be democratized, spread more broadly. In some contexts we have seen gestures in that direction that have fallen flat. In the entertainment industry, we can see it happening. So here's one manifestation. Modified films. Digital technology and the internet is making it increasingly common for the consumers of films to edit them and then redistribute the modified versions. Examples include the many different modified versions of the first Star Wars movies known as Phantom Edits, each one of which modifies in some way the characters, typically by reducing the role of the buffoon, Jar Jar Binks, and expanding the role of the young Anakin Skywalker. The Batman movies have been recombined to a new collage known as Chaos and Gotham. Hop's dark version of artificial intelligence recapturing some of Kubrick's original vision is yet another example of a modified version of a film circulating now fairly widely on the internet. Each of these is instances in which consumers have taken a hold of a cultural artifact, in these cases major motion pictures, and begun to recode them, alter their meanings, participate more actively than was possible a few years earlier in the process of fashioning the cloud of entertainment through which we move. The mechanisms that make this possible are advancing rapidly, so many companies now are producing technology that facilitate this process. Clear play, trilogy, family shield, go to a TV or all examples. The most flexible is Final Cut Pro, a widely available software that enables consumers with a modest amount of money to spend to engage in quite professional-level editing of digital copies of films. There's now a user group. You use Google to find the Los Angeles Final Cut Pro user group. You can get a quick education on how to tinker with the stuff that shows up in your inbox. In the music industry, we see analogous techniques for modifying the recordings that come into our hands. Mad Player is one example. MP7 is another example, a compression-decompression system that enables the recipients of recordings to disaggregate them into their separate tracks. And so, on a simple level here, pull out the vocal track and make your own karaoke recirculated on the web. A final example of illustration of the beginnings of this semiotic democracy is amateur webcasting. Many of the stations that can be found in live365.com consist of people who just thought they would make a collection of songs available to their friends. It's that that has contributed to the diversity of material available to all of us. The indirect benefit of this habit is that more and more people are getting involved in the process of shaping our cultural environments. That's semiotic democracy. Alright, it looks pretty impressive so far. A very substantial list of benefits, but it's important also to recognize the dangers associated with the new technology. There are three of them. The first is familiar but worth emphasizing. The new technologies threaten the revenues of creators. Now, this is a little more complicated than it might at first appear. But some aspects of the new technology hold out potential advantages to creators. So network effects. The more people use the internet to gain access to recordings, the more valuable each recording is to the next consumer. Why? Because if you can discuss a particular song or movie with your next door neighbor, it's more valuable to you than if you only have access to it yourself. So network effects enhanced by the internet is potentially beneficial to the creators of entertainment. Another potentially beneficial effect is known as the exposure effect. Janice Ian, for example, thinks that widespread distribution of her material over the internet for free enhances her revenues rather than diminishes them because alerts more people to her music will then go out and buy it. Finally, some economists have suggested indirect appropriation, the ability of the first recipient of a digital copy to pay larger amounts for it is potentially a potential corrective to the adverse impacts of the new copying systems. All right, those are all possibilities, but in practice they have been swamped by the last one. Promiscuous reproduction over the internet of digital copies of sound recordings and films corrodes consumer demand because if you can gain access to a song for free over the internet, why buy it? That's the standard reading of this of the economy of this new technology and it is to a large extent true. So the current trends in this field are discouraging. Here's some recent data. Notice these are sales data in the United States of sound recordings. The total number of units sold peaked in 1999 and has been falling ever since. Not surprisingly, retail value, indirectly the determinant of how much the copyright owners in these recordings can make has also been falling since 1999 at an accelerating pace. On the global level much the same thing has occurred. This is a longer time spread. Notice that between the mid-1970s and 1990, long-play vinyl albums were a substantial portion of the products that were distributed housing sound recordings and they began to fall off actually they began to fall off in the mid-1980s and have been doing so steadily ever since. They were replaced for a long time by cassettes which throughout the world remain a substantial distribution channel but then in the late 1990s the vinyl record and cassette markets began to be displaced increasingly by compact discs. So the top line here is the aggregate performance the total number of copies of sound recordings distributed throughout the world with some dips here and here continues to rise until yet again 1999 and now since then has been falling. The film industry has not yet experienced the same devastating corrosion of its revenues. On the contrary, in the United States the film industry has been making ever more money during the years since 1999 and much the same thing has been happening in global sales of American films. Notice that they've been going steadily up but most observers certainly including most participants in the film industry recognize that as the availability of broadband internet access increases as the size of the pipes through which broadband access is supplied meaning the rate at which materials can be transmitted continues to increase as compression and decompression systems improve and as home entertainment systems proliferate the film industry will experience the same corrosion. So that's the first best known of the hazards associated with the new technologies but there are two others and they're related. Moral rights. This is a concern stronger in Europe than in the United States but still significant here. The creators of entertainment of all sorts worry that the new technology threatens the integrity of their products enables consumers to distort, corrupt them. So this is the flip side of the advantages of semiotic democracy. Initially when this behavior first made an appearance in Hollywood the creators of films appeared not to be terribly disturbed so George Lucas remarked when told about the first phantom edit the internet is a new medium it's all about doing things like that a generous tolerant response more recently Steven Spielberg commenting on the director's guild suit against clean flicks one of the firms operating this space said much more ominously no one is authorized to impose their truth on top of ours no matter how strongly they may disagree with it a fierce defense of the artists' moral rights. All right, that's something to worry about and for related reasons we should also pay some attention to the potential of the new technologies to destabilize popular culture blur cultural reference points reduce opportunities for parody and play and undermine cultural cohesion why in each case because the ability of consumers to modify and recirculate cultural artifacts means that in the near future we will have fewer and fewer stable, familiar, consistent pieces of entertainment available to us you'll have to specify well you saw the most recent Lord of the Rings movie last night which one was it? 3.1 version 3.2 version 3.3 and so forth okay so that's the end of the second of our three topics to recap here we've examined the structure of the entertainment industry circa 1990 the way it worked its strengths and its relatively modest weaknesses then discussed how technology between that began to make its appearance in the 1990s and continued to be widely available today has destabilized that traditional industry offering us simultaneously large potential benefits and significant problems our goal one would think in harnessing the new technology would be to structure business models and legal rules in a fashion that enabled us to reap as effectively as possible the potential benefits while avoiding as many as possible of the potential harms unfortunately that's not what we have done to date instead the changes we've seen in the legal system and associated business models since 1990 have had the effect of defending traditional business models against corrosion at the cost of limitation of the potential consumer benefits so the overarching characterization here is a defense of the old order a 14 year long defense of the old order it's happened through eight cycles of innovation and resistance eight overlapping cycles so eight times in the last 15 years we have seen a technological innovation that offers consumers some significant advantages followed by resistance sometimes the resistance has taken the form of litigation relying upon existing copyright law sometimes it's taken the form of law reform in which the organizations threatened by the new technologies have successfully sought a change in copyright law finally sometimes the resistance has taken the form of technological responses efforts to use counter technologies to shield existing business models against corrosion three different kinds of responses in most instances the resistors have prevailed now not all of these stories are complete the outcomes of a few remain in some doubt but for the most part the resistance has been successful so for the remainder of this lecture we'll be describing how each of these eight crises individual crises have developed and the legal and business responses alright here's the first of the eight digital recorders digital tape recorders and other digital recorders first a little bit of background analog taping of record albums has been common for a long time it was common in my youth it was common in my children's youth you'd buy a record you'd take your cassette tape recorder and hook it up to your amplifier and you'd make a copy of the long play album on your tape either a verbatim copy or more commonly for my daughters a mix you would take a few songs from one record a few songs from another and you'd distribute them to your friends this behavior very common for decades is probably illegal was probably illegal when it was engaged in there were some observers who argued that it was privileged lawful because of a tacit exception to the sound recording to the reproduction right for sound recordings embodied in the copyright statute perhaps tacitly legalized by the penumbra of the Betamax decision but each of those suggestions was fairly implausible in my view almost certainly a fair reading of the relevant copyright statute suggests that making a verbatim tape copy of a record and circulating it to your friends violates both the reproduction right of both the holder of the copyright in the composition and the holder of the copyright in the sound recording the record companies lost money as a result of this practice an RAA study circa 1990 estimated their losses and they ordered a magnitude of 1.5 billion dollars a year so it was not trivial but they tolerated it why? well in part because they were reluctant to sue their own customers and in part because of an important technological fact namely analog tapes are less good than the media from which they are taken and successive generations of analog tapes get worse and worse and worse so this technology though common among consumers was not a vehicle by which one could engage plausibly in large scale piracy and as a result to repeat it was tolerated and the process that destabilized this practice was the advent of first CD technology which made recordings available in digital form and then the development in the late 1980s of consumer versions of digital audio tape recorders and a little bit later mini disc recorders these new systems in combination offered consumers a mechanism of making copies of their recordings vastly more threatening to the revenues of the record companies why? because digital audio tape recorders make perfect copies of CDs and each successive generation of dad tapes is just as good as the originals so the record companies now abandoned their former tolerance posture and brought suit the manufacturers and potential importers of digital audio tape recorders same the basis of the suit was the same as the basis for the Sony litigation long ago contributory copyright infringement introducing these things into the American market would encourage consumers to engage in copyright infringement so the parties talked, argued and finally in 1992 entered into a compromise a compromise originally a tacit compromise among all the players but subsequently embodied at their suggestion in legislation specifically in the audio home recording act that statute had three components first a requirement that all digital audio recording devices contain a serial copyright management system which prevented them from making third generation copies with a dat recorder or a mini-disc recorder now distributed to consumers you can only make one generation of copies so first element of the compromise was the technological restriction second a compulsory royalty when digital audio recording devices are sold to consumers they come accompanied with a 2% tax 2% of the wholesale price is added on as a tax when the media used to make digital audio recordings blank music CDs are sold they also are taxed 3% of their wholesale prices and that money is turned over to a fund which is then distributed one third to the holders of copyrights and musical works and two thirds the holders of copyrights and sound recordings on its face this is quite attractive it's a sensible resolution of the crisis unfortunately it turned out not to be perfect for two reasons first the several year delay in which the party struggled over the right legal regime turned out to have been fatal to consumer demand so very few consumers nowadays have dat recorders by the time the dust settled consumers have moved on to newer technologies the second problem with the settlement is that it was too technologically specific so in 1999 the court of appeals for the Ninth Circuit held that the Rio which is an early mp3 player was not covered by the new regime it's also been deemed not applicable to CD burners that distributed as part of personal computers as a routine matter these days nor to blank data CDs which are by far more numerous than music CDs so the total revenue collected through this system is small four million dollars a year tiny compared to the roughly speaking eleven billion dollars eleven billion dollars a year collected by the recording industry so a promising possible outcome that turned out to be disappointing in practice that's the first of the cycles next encryption as the threat posed by the new technologies increased the companies working in this space began increasingly to rely upon encryption systems to try to prevent consumers from getting copies for free examples include the CSS system which was designed to prevent DVDs from being copied the secure digital music initiative a consortium of hardware and software manufacturers who sought to develop a secure encryption system for sound recordings the copy switch used by real networks to prevent the recipients of real audio streams from making copies of them all devices to try to use technology to restrict copying the problem is that each of these systems is vulnerable to hacking and so the developers of these encryption systems sought aid from the legal system they sought legislation that would penalize either with civil penalties or criminal penalties circumvention of the new technologies so the IITF white paper initially recommended that Congress adopt such a legal backstop to encryption in the United States when Congress during an election year refused to adopt this recommendation the backers of it turned to the international arena and secured through a long struggle a treaty provision under which contracting parties that includes the United States are signatory to this treaty were obliged to provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures which are used by authors in connection with the exercise of their rights armed with this new treaty provision the proponents of legal reinforcement of encryption came back to the United States and obtained the digital millennium copyright act a large statute that tries to do many many things but contains one provision relevant to our present topic namely section 1201 which forbids three related things first, forbids people to circumvent technological measures meaning encryption systems that effectively control access to a copyrighted work second, forbids the manufacture importation or trafficking in any technology that's designed for the purpose of circumventing access controls and finally, forbids the development and distribution of technology designed to circumvent encryption systems that limit copying in other words encryption systems that reinforce the entitlements enjoyed by copyright owners under section 106 of the statute those are the main aspects of section 1201 and it's backed up with substantial civil penalties and criminal penalties armed with this new statute the copyright owners have been with considerable success pursuing the developers or deployers of technology that break their encryption devices not perfect success as we're going to see but considerable success an illustrative case is the stream box litigation which the western district of Washington in 2000 held that a virtual VCR a device distributed by this company stream box that overrode the copy protection switch in real networks violates the digital money copyright act the big case in this area is Corley this was the most notorious of the pieces of litigation to understand it requires a little bit of factual background people who set up websites on the internet commonly combine a home page with subordinate websites subordinate web pages so suppose I want to set up a website I will customarily put on it some advertising so I can make some money and provide links to subordinate pages on which I'll put additional information that I think visitors might want to have access to so suppose one of the things I provide a link to is illegal material material that either the posting which violates the copyright statute or pornography or something else that means that a visitor can gain access to this illegal material in a straightforward manner by visiting my site clicking on one of my subordinate links and finding it complications arise, legal complications arise if a second website is set up elsewhere on the internet typically with its own advertising its own subordinate pages and a cross link to me enabling a visitor to gain access to the illegal material either by going to my site or by going to site number two clicking on its lateral link finding my subordinate link and finding the illegal material under these circumstances the plaintiff suppose here the holder of the copyright in the material that's been illegally posted has a straightforward cause of action against me for copying and reproducing without permission is her stuff does the plaintiff under these circumstances also have a cause of action against site number two that's the tricky issue that has been resolved differently in different contexts through litigation involving copyright infringement pornography and so forth on the internet the Corley case known as the Ramirdis case later the Corley case arose as a variant of this common troublesome phenomenon in which the illegal material in question was a decryption program developed by a Norwegian teenager that enabled recipients of it to break the CSS system that shielded DVDs against copying the program, the DCSS program developed by the teenager was circulated widely on the internet one of the organizations that gained access to it was an online magazine an online hacker magazine known as 2600 that posted copies of it on its site when a court ordered 2600 to remove the material in question from its own site it did so but provided lateral links to other sites where the decryption program could be found generating finally a variant of this common difficult problem so 2600 magazine is in the position of site number two and the illegal material in question is located over here does the behavior of 2600 magazine making available through these lateral links access to the decryption program violate the Digital Millennium Copyright Act the answer is yes held by the Southern District of New York in 2000 providing a link to a website that in turn enables visitors to download copies of encryption breaking software consists of trafficking in violation of section 1201A2 District Court went on to reject the various arguments made by the defendant to excuse its behavior the case was then appealed to the Second Circuit which in a major opinion in 2001 affirmed the lower court's ruling granting upholding the issuance of an injunction against the continuation of lateral links of this sort and rejecting a constitutional challenge to section 1201 is applied holding that the prohibition on trafficking in decryption software is in the language of constitutional law content neutral advances a substantial governmental purpose namely assisting copyright owners in preventing access to their property does not burden speech more than necessary for that purpose and therefore does not violate the First Amendment other constitutional challenges were raised in the Court of Appeals view too late to be considered so that's the core of the opinion an aggressive precedent applying section 1201 not just to the people who distribute decryption software but to the people who make possible users of the internet to locate websites where copies can be found the defenders of copyrighted material against decryption have not won every case so here are two examples of situations in which they have lost I mentioned a minute ago the SDMI initiative the consortium of hardware and software manufacturers who sought to develop a secure watermark system well after lots of time and effort they came up with some prototypes of watermarks which bravely they made available to the world at large encouraging people to try to crack them a computer scientist at Princeton University by the name of Ed Felton the assistant of his graduate students took up the challenge and succeeded and then decided instead of turning over the fruits of his research to the SDMI group he would publish it when his intent was made public the recording industry sent him a letter indicating that publication of the material could subject you and your research teams to action under the Digital Morning Copyright Act Mr. Felton responded with the assistance of the electronic frontier foundation by filing a declaratory judgment action and federal district court seeking a declaration that as applied to him the Digital Morning Copyright Act would violate the First Amendment curbing his freedom of speech his freedom to publish the fruits of his research had the case gone forward it would have been an extremely favorable factual record for presenting a constitutional challenge recognizing this the recording industry retreated retracted its tentative threat and insisting that it had no intention of preventing Felton from publishing his research and as a result the case was dismissed as moot so the Felton case did not issue in any judicial ruling any substantive judicial ruling but it did represent one instance in which an effort to invoke the statute fell the other circumstance in which an effort to rely upon the Digital Morning Copyright Act turned out in the end ineffectual involved the development by a Russian programmer Dmitry Skilyarov working for the company Elmcomsoft of software that defeated the copy protection system built into the Adobe e-book reader when Mr. Skilyarov at the time attending a hacker conference in Las Vegas was arrested for criminal violation of the Digital Morning Copyright Act there ensued a fierce public outcry and ultimately the prosecution failed the prosecutors dropped the charges against Skilyarov individually in return for a promise that he would cooperate in the prosecution of the company was in turn acquitted because there was inadequate proof of criminal intent so a victory for the I think of here as the hackers but not a terribly important one in the long run because notice it's founded upon a judgment on the part of the jury that the principles involved had insufficient knowledge of an intent to violate the statute nowadays very few defendants in criminal positions could plead analogous ignorance alright that's cycle number two cycle number three not quite so grand involves music lockers these are systems developed in the early 1990s that enabled people with digital copies of recordings on their hard drives to store them on servers to which they could gain access from other locations so these so-called lockers virtual lockers function in part as backup devices in case your hard drive crashed but in part they functioned as places where you could store music that you could then play on any internet connected device there were a bunch of them iDrive, Reoport, MyPlay MyMP3Stories.com, XDrive these were all systems that enabled individual customers to upload their music to the music locker the pioneering website MP3.com went one step better MP3.com developed two closely related services which offered an easier way in which consumers could load their virtual lockers called the BMIT service and the instant listening service here's the idea if I already own a CD argued MP3.com I ought to be able to place in a virtual locker on the MP3.com servers MP3 versions of the CD songs I already as they describe it, own so how would this work in practice would work by I would take my CD, place it in the CD ROM drive of my computer which would then communicate to the MP3.com website which would respond by loading into my individual locker MP3 versions of the songs in question or when I bought a CD through an online retailer connected to MP3.com at the same time that the physical CD is shipped to me MP3 versions of the same songs would be loaded into my virtual locker now the system is not perfectly secure as you can imagine it would be possible for me to obtain a CD from my next door neighbor loaded into my CD ROM drive and deceive this relatively simple security mechanism to thinking that I owned it already but the litigation that ensued here did not hinge on that fact the leakiness of the system rather the recording companies argued that in order to create the MP3 copies that it would then copy into people's virtual lockers MP3.com had engaged in copyright infringement specifically they had made reproductions of 40,000 CDs without permission a clear violation in their view of section 106 of the statute and the district court judge who heard the case agreed now MP3.com responded with an invocation of our old friend fair use defense okay it's true we violated section 106 of the statute by making without permission verbatim copies but our behavior should be justified excused for reasons analogous to the reasons that enabled Sony to escape liability in the Betamax case the district court judge responded no fair use defense is not available here and crucial to his reasoning is that the defendant is not the consumer it's the company remember in the Betamax case the person engaged in the copying is the owner of the VCR in the MP3.com case the defendant engaged in the copying is MP3.com the company in question so it's engaged argues Judge Rakoff in a commercial use not a non-commercial use the material being copied is creative and therefore subject to substantial protection the entire work is being copied and this practice undermines a potential licensing market otherwise available to the recording industry a generous expansive interpretation of the fourth of the fair use factors so MP3.com is unable to invoke the fair use defense the recording industry wins the company MP3.com responds by settling the suit with four of the five record companies for a fee of approximately $20 million per company plus future license fees but one of the five universal music pressed on didn't accept a settlement in the end was awarded damages in the amount of $188 million did settle the case in the wake of this award for $53 million plus warrants rights to purchase the company under certain circumstances again there was a public outcry efforts to secure some legislation reversing the decision which failed and finally in the wake of that failure MP3.com pioneering website in this space has collapsed universal music the last of the plaintiffs bought the company for $350 million although it resold it later to CNET networks and other music locker services in the wake of this decision closed up shop so an example of the pattern technological innovation potentially beneficial to consumers threatening to existing business models resistance both of the cycles involves webcasting so we talked some time ago about what webcasting consists of webcasters of various sorts pose arguably pose more substantial danger to record company revenues than do radio broadcasts which as I indicated have been permitted for a long time well because webcasting provides consumers music typically with better fidelity though not always than radio the variety of material available on the webcasting stations is greater it comes with fewer advertisements and finally and most importantly some of the webcasters are interactive in the sense that consumers can specify the musical work they want to listen to each of these factors but most importantly the last one is the likelihood that consumers will rely for their musical sources on webcasting rather than purchasing CDs and to the extent they rely more on webcasting there is a potential that sales of CDs would decline so pointing to this danger the record companies approached congress and argued yet again that they should be given an entitlement which record companies in most other parts of the world already enjoyed namely a public performance right and congress this time obliged in two related statutes the digital performance and sound recordings act effective in 1996 and the digital millennium copyright act another provision of that enormous omnibus statute adopted in 1998 congress provided to the record companies for the first time a narrow narrowly defined public performance right so remember this I hope this chart of the entitlements in the music industry the digital performance and sound recordings action the digital millennium copyright act added this leg to the diagram added a public performance right for the holders of copyrights and sound recordings limited however to public performances in the form of digital audio transmissions and then further complicated the scene by qualifying that entitlement with some limitations and exceptions embodied in section 114 of the statute this is very intricate but here is roughly speaking how it works the statute divides the universe of digital audio transmissions of sound recordings into three boxes there are level one transmissions which are exempt level two transmissions are forbidden unless and until the transmitter pays a compulsory license set by an arbitration panel and there are level three transmissions which are permitted only if the transmitter obtains a license after voluntary negotiation with the copyright owners three different levels examples exempt transmissions include such things as non-subscription broadcast transmissions licensed by the FCC those are at their core digital transmissions of radio programming over the airwaves not through the internet and store casting so if you go into a grocery store and you hear background music some of that will have come through a digital audio transmission does the local grocery store have to pay a license fee? no, they're exempt why? because it was the theory of the people who organized the statute and the congress that adopted it that you will not buy fewer CDs because you've been able to listen to music in the background in the store store casting does not threaten the revenues so they exempted these level two examples of level two these are digital audio transmissions for which one has to pay a compulsory license non-interactive subscription transmissions and non-interactive non-subscription transmissions so here imagine a webcaster almost all of those that are included in live365.com releasing information over the internet releasing recordings over the internet in streamed form chosen by the webcaster so long as they conform to some requirements you can't play Beatles music all the time you can't announce in advance your programming schedule and you have to make available copyright management information to the recipients as long as you abide by all those restrictions then you the webcaster get the benefit of a compulsory license you have to pay only a license fee set by an arbitration panel third category most important things here are interactive services they have to pay freely negotiated licenses they have to pay in effect whatever the record companies demand ok so far so good a reasonably sensible response to the apparent threat posed by webcasting to record company revenues the devil unfortunately is in the details specifically in the manner in which the arbitration panel set the royalties that the webcasters would have to pay so the proceeding that generated those rates terminates in 2002 the arbitration panel ordered webcasters in the future to pay the following fees example here is a regular non-commercial non-interactive webcaster has to pay 14 cents excuse me 0.14 cents per performance plus a premium of 9% seems pretty small right seems like a tiny figure well turns out for reasons we'll discuss in just a minute to be larger than you would expect there was controversy over these rates which prompted the librarian of congress on the recommendation of the copyright office to cut them down specifically to cut them in about half but the controversy continued why were people upset by these seemingly tiny fees of arbitration suppose that I operate a small hip-hop webcasting station distributing ad-free popular music to on average 10,000 listeners 24 hours a day that's about 15 songs an hour these are short hip-hop songs 360 songs per day 131,000 songs per year times the number of listeners that's over a billion performances per year I'll multiply that times the magnitude of the new license fees I will be obliged to pay to the organization that collects the money on behalf of the record companies a little under a million dollars per year plus an ephemeral license fee of $80,000 to just over a million dollars this to repeat is a small non-commercial hip-hop station stations in this general category a million dollars a year is prohibitive impossible and as a result in the wake of the final ruling in the case approximately a third of webcasters closed down controversy continued to intensify and congress intervened adopted a curious statute the small webcaster settlement act which nudged the record company into negotiating substantially lower fees for a short period with small webcasters so they've obtained a reprieve but only a brief one okay well what do we learn in retrospect from this process well the idea of using a compulsory license to on the one hand compensate the record companies for any losses they sustain as a result of the advent of webcasting while on the other hand keeping webcasters to flourish makes good sense problem was in the manner in which the rates were set so as a matter of procedure they were set by private arbitrators who no fault of their own did not have any expertise in this field next the expense of the process all the fees are borne by the parties which meant that small webcasters who didn't have the money to participate were neglected the statutory standard that the arbitrators rightly employed to set the rates could be construed in a fashion that's socially unwise so specifically the standard refers to the amount of money that a willing buyer would pay a willing seller for access to the music in question that standard construed to give rise to a fair market value in this context ignores the reasons why non-interactive webcasters should pay less than interactive ones ignores the need well recognized by congress to interpret the statute in a fashion that nurtures the infant webcasting industry and also last but not least neglects the hazard well recognized by congress in the legislative history that the record company companies if they can demand freely negotiated fees will act as a cartel finally this legislative structure unfortunately encourages the record companies for understandable reasons to seek high rates from the get-go because they know from experience that once rates are set it's very difficult in successive arbitration proceedings to get them to go up fast and so again for understandable reasons they demanded a high fee from the beginning may be sensible from their profit maximizing standpoint but regrettable from the standpoint of society at large alright that cycle number four cycle number five these two are the most familiar the next two are the most familiar centralized file sharing and decentralized file sharing so this is what's been in the news the most recently the peer-to-peer revolution led by the famous Napster organization so here's how Napster worked Napster established a website where they constructed a directory and an index and then made available to users free software music share software so user number one here connects to the Napster website downloads the music share software to his or her computer and provides a name and a password typically they're not always a fictional name and then uses the software to construct on his or her computer a library of mp3 recordings where do they come from well typically they are ripped from cd using free software available through the internet then having constructed this library on his or her hard drive user number one connects with the system again logs on and the software the music share software informs the Napster directory of the location where these mp3 files could be found now notice it's critical here that the Napster website does not contain copies of the mp3 files the copies remain on the users hard drive all the Napster website has is a directory so continue the story user number two does the same thing constructs his or her own library logs on the software lists user number two's library contents in the directory and now here's where peer to peer sharing occurs user number two sends a search request to the index the Napster index which say do you have any Joshua Redmond songs the index checks the directory and finds such a recording says yes indeed one of the users currently logged on has a Joshua Redmond recording on his or her hard drive you can find it at the following URL user number two then using that information connects directly to user number one and downloads the song to his or her own library that's peer to peer file sharing coordinated by the Napster website now I have to spend two users this would not have been troublesome trouble arose because the Napster system enabled millions of users simultaneously to connect to the system and distribute files among them alright that's Napster the pioneering website in this space other examples of centralized file sharing includes scour which is quite similar to Napster but allowed the sharing of digital files of lots of different sorts not just MP3 recordings Aimster and Madster Aimster subsequently renamed Madster is a Napster-like organization that piggybacked on AOL's instant messaging system hence the name Aimster's AOL AOL instant messaging system differed slightly in that it used encrypted files which were designed in part it appears to make it harder for the service to monitor what was flowing through its system but otherwise similar okay those are the first wave centralized file sharing systems they provoked a famous round of litigation so each of the three organizations we've just discussed was sued by the record companies and later in the case of the firms that transmitted video files also by the music industry the Napster case resulted in a declaration first at the district court level and then at the court of appeals at the district court level first the Napster was liable for contributory and vicarious copyright infringement partly because none of the defenses that its lawyers offered held up in the view of the court of appeals so now remember Napster is very similar in structure to Betamax in both cases the defendant is not copying any files the defendant is accused of facilitating the behavior of its customers who are copying copyrighted files in both cases the claim asserted against the defendant was contributory copyright infringement in both cases the defendant asserted that what its customers were doing should be excused as fair use and the Betamax case that defense succeeded and the Napster case it failed why? in this circuit none of the following activities could legitimately be considered fair first sampling Napster argued that one of the ways in which a system was used was to obtain copies of copyrighted songs which we then listened to once or a couple of times and then deleted the sampler if he or she liked the recordings in question would purchase legitimate copies next space shifting Napster argued that one of its functions one of the purposes of the system was to enable people to connect their home computers to the Napster website and then go to work logging on from their work and transfer files from their home computers to their office computers finally Napster argued that some of the copying done by its customers was permissive consisted of material the copyright owners of which allowed sharing none of these in the court of appeals opinion were persuasive none of these arguments were persuasive the latter two because the copying question was de minimis sampling is the issue that was most hotly debated and in which if one reads conscientiously the Betamax decision the Court of Appeals for the Ninth Circuit made a mistake so it's a controversial assertion but some an assertion that we may come back to in the live portion of the iLaw program the argument I'll offer here is that the Napster decision vis-a-vis the legal status of sampling was misguided why because sampling fairly analyzed is a non-commercial not commercial activity the customers of Napster are engaged in this behavior for the purpose of locating songs that they're then going to purchase not recirculating them to their friends and that activity does not corrode the potential market for sound recordings if the phrase potential market is construed in the narrow sense in which it was construed in the Betamax decision to avoid this outcome which would have led to the opposite result in the case the Ninth Circuit was obliged tacitly to deviate from the Betamax ruling to modify the conception of fair use to constrict it a modification that bodes ill for the future of the Fair Use Doctrine as a regulator in this space okay so to return to the Napster ruling none of the defenses claimed by Napster in the judgment of the Ninth Circuit held up neither the Fair Use Doctrine nor the digital one in copyright acts Safe Harbors nor its defense of copyright misuse bottom line Napster loses and unless and until it is able to purge its system 100% of copyrighted material it had to shut down unable to comply with that order Napster closed the Scour lawsuit was very similar in structure Scour actually had more substantial legal claims to be made but never got to make them because it didn't have the funds to mount an effective defense Scour went bankrupt even before the litigation began in earnest finally the Amster-Madster site did put up a more substantial legal defense but was held by the Seventh Circuit and an opinion by Judge Posner to be engaged in contributory copyright infringement there's some interesting nuances to the opinion but not worth dwelling on here same ultimate ruling contributory copyright infringement no fair use defense along the way the companies in question sought to obtain from the record companies licensing deals which had they been negotiated would have enabled us to reap very substantial social gains Napster at one point offered the record companies a billion dollars for permission to continue to engage in this activity Scour was involved in analogous negotiations when the suits were filed against it so there was a failure here of a licensing opportunity Decentralized file sharing looks similar from a consumer standpoint but has a different technology underlying it Examples the Genutella system a renegade project that actually issued from some unauthorized programming by AOL an AOL team a leading example improved Nutella systems like LimeWire and BearShare other examples the fast-track technology employed by several of the most popular companies Kaza, Morpheus, Music City Streamcast all one enterprise and Groxter another example and the most recent of the firms in this space Freenet and EarthStation 5 all used decentralized file sharing technology so what's decentralized well here's an illustration from the fast-track regime the way the fast-track system works to repeat several companies basically piggyback on the same technology is users interested in sharing files log on to a central server which provides them the address of a computer that's offered by its owner Supernode Supernodes are engaged in traffic coordinating knowledge concerning where files enjoyed libraryed on the computers of individual users are housed so each Supernode has a respective cluster of clients connected to it so suppose here that user number 3 wants to copy a file submits a search request to a Supernode which relays it to another Supernode which obtains the information in question from its connected users if the file in question is located then it is transmitted outside the system through the internet in general to the original requester no single central facility the central hub of Napster is missing here but through a combination of decentralized searches and Supernodes a very similar outcome so as I say lots of companies use variants of this technology the most amusing at least from a political standpoint is Earth Station 5 founded on the West Bank and its website proclaims that among its founders there are Jordanians Palestinians Indians Americans Russians and Israelis offering an example of cross-border cooperation certainly never achieved in the political arena somewhat different vein they then announced belligerently resistance is futile these systems have more than fully replaced Napster and the other centralized systems traffic through the decentralized networks is as great or certainly was as of the summer of 2003 as great as it was through Napster during its peak so not surprisingly the record companies once again brought suit this time however they did not fare so well so the suit against Caza originally brought in the Netherlands resulted in a trial court ruling in favor of the record companies but subsequently reversed by two appellate courts in the judgment of the appellate courts Caza had to be sure some illegal uses but also had sufficient legal uses to be justifiable to escape liability for contributory infringement similar outcome in the American suit brought against Grokster Judge Wilson District Court in California rules that Grokster streamcast not liable for contributory copyright infringement because there are various important non infringing uses of their systems and the decentralized structure thereof means that the networks themselves cannot discriminate between illegal transmissions and legal transmissions currently on appeal to the Ninth Circuit this is one of the decisions that may get modified before we meet in person in May okay last but not least partly because of the failure of these two suits against the services the record companies are now going after individual users there was a controversy still is a controversy concerning the mechanisms by which the record companies sought to obtain information about individual users initially they relied on a provision of the Digital Millennium Copyright Act that allowed them to file subpoenas with clerks and quickly obtained the names obtained from internet service providers the names of the persons whose IP addresses were being employed to share files in the Verizon litigation Court of Appeals for the District Columbia Circuit recently ruled that one could not rely on this streamlined subpoena system compelling the record companies to rely on a somewhat more secure procedural device of John Doe lawsuits and associated subpoenas that only slow did not stop the wave of lawsuits so so far there have been two waves of suits in each instance the record companies have after frightening the targets settled every one of the claims typically between $2,000 and $17,000 per defendant in the standard case pledges to stop the illegal behavior and acknowledges wrongdoing the prominence of these suits suggested by the fact that during the Super Bowl Pepsi managed to persuade a subset actually one young girl in particular to appear in an advertisement acknowledging that she had engaged in illegal behavior and then oddly pledging to continue to engage in downloading of material but only of songs that Sony made available for free when things rise to the Super Bowl level you know that the public attention is intense okay, we're nearly done here that's sixth of the eight cycles seventh involves CD burning yet another way in which consumers gain cheaper, more rapid access to recordings in a manner that threatens the revenue of copyright owners the technology necessary to engage in CD burning has been spreading fast in part because the devices have been improving and getting cheaper in 1995 a CD burner cost about $1,000 and recorded the CD only at about four times the speed with which it could be played by 2003 you could get a CD burner for $40 and it would record at 52 times playback speed and partly as a result CD burners are now included as a matter of course in personal computers sold in the United States about 120 million of them have now been sold so what do people do with CD burners? well, a variety of things some legal, some likely illegal so sometimes people use CD burners to make backup copies of commercial CDs in case they get scratched this seems most likely lawful as a fair use of the technology sometimes they make copies of commercial CDs to distribute to their friends this for the reasons I suggested earlier is probably unlawful not a fair use in part because it does indeed corrode the potential market for commercial CDs sometimes they make mixes they use CD burners to make new collections of songs this is substantially more likely to be lawful because it's transformative sometimes they use them to copy downloaded files to CDs so suppose that you obtain a collection of MPD recordings by buying them from iTunes you can use your CD burner to make a CD from them this is probably lawful at least if the files in question were originally obtained with permission similarly, using a CD burner to convert albums to CDs your old vinyl collection you want to preserve it for posterity or at least for your future uses that's probably legitimate behavior the opposite extreme is commercial piracy some CD burners are used to make mass produced CDs that are then distributed on the streets clearly illegal and finally a few people use them to prepare teaching materials so bottom line is these devices are susceptible of a wide variety of uses some are legitimate, some illegitimate what's been the response of the copyright owners well interestingly it has not been litigation perhaps on the theory that the devices have a significant set of legitimate uses the recording companies have not brought suit against the manufacturers of CD burners instead in this instance they've relied on technological protections so specifically they have developed a variety of systems whose purpose is to protect commercial CDs against reproduction so there are a bunch of these as of February 2004 they include key to audio, safe audio cactus shield, media clock and so forth what's happened well consumers thus far have devised ways to break the encryption systems turns out some are fairly easy to break the early encryption systems could be defeated by using a black magic marker to put a little stripe around the outside edge of a copy protected CD more recently the copy protection system offered by Suncom was defeated the discovery that one could by simply holding down the shift key on your computer override it briefly Suncom threatened to bring a digital millennium copyright infringement suit against the graduate student who made public this device but after derisive attention from the public media withdrew so so far this technological response has not been successful but that's probably only because the technology is thus far primitive as the copy protection systems improve you can expect the effectiveness of this initiative to go up alright that brings us to the last of the eight cycles which involves a subset of digital video recorders what are digital video recorders or DVRs these are machines that are the digital equivalent of VCRs the most common of them most popular is Tivo they're spreading fast so in 2003 there were 4.3 million in operation in the United States analysts predict there will be 17 million by 2005 38 million by 2007 people who get these machines love them the head of the FCC recently referred to them as God's machine why because they make it much more convenient to record and replay programming they also facilitate ad skipping it's much easier using a Tivo system to skip ads than it is using a VCR and that for understandable reasons worries the studios the studios have not to date however brought suit against the manufacturers of regular DVRs Tivo in particular why probably because of the shadow of the Sony decision which has been construed more broadly than the holding precisely establishes construed to suggest that consumers have a right to time shift programming broadcast to them over the airways whether it's using a digital machine or an analog machine a particular type of DVR however did draw the fire of the movie industry and that's replay specifically the replay 4000 and 5000 series two features of the replay systems in the opinion of the movie industry were excessively threatening the send show feature enabled customers to transmit to their friends material they had taped and even more dangerous was the commercial advance feature which made it possible for the user of a replay system easily automatically to delete the advertisements that was too much in the opinion of the film industry and they brought suit once again arguing that the manufacture and distribution of these devices constituted contributory copyright infringement and this time they prevailed sonic blue the parent company in the face of the litigation declared bankruptcy replay TV was sold to digital networks which then probably removed the offending features in question so those are the eight cycles each one to repeat involved some technological initiative which sought to capitalize on the new systems to make entertainment more easily accessible to consumers each one was met by resistance sometimes litigation sometimes law reform sometimes technology so where is it left us well in an unhappy state so the end result of this tumult is a world in which the distribution of recorded entertainment is characterized by high transaction costs including enormous cost of litigation in which the price to consumers of access to recorded entertainment remains strikingly high in which we do not have and have no prospect of seeing a celestial jukebox in which encryption systems digital rights management systems reduce the flexibility of the media we gain access to curtail fair uses and impede opportunities of a semiotic democracy in which tens of millions of Americans somewhere between 30 and 60 millions of Americans regularly violate the law and in which despite all of these efforts to protect against corrosion the traditional business models the fortunes of the record industry and soon the fortunes of the film industry are declining so that's the character of the crisis and the purpose of this lecture has been to suggest how we got there from this story we hope to infer lessons concerning how the crisis could be resolved and that's the topic we'll take up when we convene in May thank you