Philip Greenspun's Testimony Against the Audio Home Recording Act of 1991 (before the Subcommittee on Patents, Copyrights and Trademarks, Senate Judiciary Committee) I am against S-1623 because it will [] not "promote the progress of the arts," the constitutional justification for expanding copyright [] artificially restrain the American computer industry, which, in the absence of legislation, would use the same blank media as digital audio recorders [] force American computer users to subsidize Japanese computer vendors, in the form of taxes paid on computer data storage media that will flow into the pockets of Sony and Matsushita [] destroy an emerging American industry manufacturing digital audio products and cause a substantial loss of manufacturing jobs [] hinder small American-owned record companies in their efforts to compete with large record companies, all of which are foreign or foreign-owned [] help create a Japanese monopoly on manufacturing digital audio equipment [] injure consumers, especially blind consumers (who do a disproportionate amount of audio recording), by subjecting them to price discrimination and taxes on computer data storage and noninfringing audio storage. Most of the benefits from these injuries would flow to foreign conglomerates such as Matsushita, Philips and Sony. [] increase the trade deficit as Japanese manufacturers and foreign-owned record companies displace American suppliers Illustrations of the harm done to Americans by S-1623 Joe Audiophile makes a live recording of his church's choir with his $500 "consumer" DAT recorder. A local CD manufacturer offers to press 100 CD's from the tape for members of the congregation for $500. Using a friend's "consumer" DAT recorder, Joe tries to copy passages back and forth until a one hour master tape is produced. He cannot because his DAT recorder has Serial Copy Management System (SCMS). Joe must now pay $5000 for two "professional" DAT recorders identical to the one he already owns but without SCMS. The Japanese receive a windfall profit of $4000; the trade deficit increases by $5000; there is no effect on infringement. June Wu has just finished designing a computer for American Laptop that will compete with units made by Sony and Matsushita. June's design uses compact, inexpensive Mammoth Magnetics magneto-optical disks for data storage. Without following each customer home, Mammoth Magnetics cannot determine whether the disks are "most commonly used for the purpose of making digital audio copied recordings" ($1001 (4)(A)). Mammoth's lawyers and accountants advise Mammoth to play it safe, collect royalty tax, and file quarterly reports under S-1623. June's customers thus pay a little extra every time they buy disks for their American Laptops. Who gets this money? A substantial portion goes to American Laptop's chief competitors, Sony and Matsushita, because they own two of the largest record companies. Jerry Teenager is a copyright criminal. He owns a $50 boombox with two cassette transports similar to the one I brought here today. He buys some prerecorded cassettes but also buys blank tapes and copies his friends' cassettes with his boombox. He is perfectly satisfied with the quality of recordings he makes effortlessly onto $1 tapes. Jerry is not going to run out and spend $1000 on a digital recorder and CD player so that he could copy $15 CD's onto $10 blanks. Jerry has a fixed budget for music and, even if home taping were eliminated, would not spend substantially more on prerecorded material. Unless he tries to get a job in the American computer or consumer electronics industry, Jerry's life will be completely unaffected by passage of S-1623; the record companies will not be able to get any more money out of Jerry, with or without S-1623. Julia Pirate is a copyright criminal. She sells 2,000,000 Michael Jackson cassettes every year. Julia's profit is $3 million/year. She paid $5000 for a "professional" DAT recorder because she understood that $500 consumer machines are hobbled by an act of Congress. DAT isn't useful in her piracy career, so she keeps her machine on her 75' sailboat. Commercial piracy will not be affected by S-1623; Julia gave $4500 in windfall profit to a Japanese company and added $4500 to the trade deficit. Jack Vicious flips burgers by day and is a guitarist in the punk group Twisted Weasels by night. The Twisted Weasels made a profit of $53.22 last year from 15 performances. Instead of listening to live Weasels, people would rather buy a recording of a popular foreign group on CBS/Sony records. Jack hopes to change that by buying a digital audio recorder and pressing a CD. But passage of S-1623 means he has to spend big bucks for a useful machine. Jack can't afford a "professional" recorder and the Twisted Weasels are doomed to obscurity. Jill Engineer runs Jilltronics, a 100-employee digital audio equipment manufacturer. Jill has to innovate to compete against a vertically integrated Japanese oligopoly that controls both hardware and software. After passage of S- 1623, Jilltronics is forced to re-engineer its products, although it barely has enough cash to operate as is. Every day, Jill fends off salespeople from Motorola, National Semiconductor and Texas Instruments touting their integrated circuits ("microchips"). Japanese vendors also call to offer her memory chips and other components used in computers and industrial products. These same firms are the only source of SCMS chips inexpensive enough to use in a consumer product. Yet, when Jill asks about them, everything suddenly becomes "difficult." Requested literature arrives after 11 weeks. After more than 30 telephone calls, Jill finally gets a quote: $100 for an input/output set of digital audio interface chips that implement SCMS. Jill has just discovered what the General Accounting Office recently concluded: because a chipset costs less than $2 to produce and is incorporated in low-cost, mass-market products does not mean that the Japanese will sell them to Americans. Unable to meet the requirements of S-1623, Jill lays off 80 of her employees and starts advertising non-SCMS products to professionals only, taking care to scratch a "P" on the front panel of each. Meanwhile, Jill tries to raise enough capital to build her own SCMS chips and return to the consumer market. Financiers won't even call her back because S-1623 allows people to sue Jilltronics for things over which it has no control (e.g. "the occupations of its purchasers" and "the uses to which it is put" -- $1001 (10)(B)(vii) and (viii)). Jill struggles along without capital until Nippon-California Records, owned by one of Jill's hardware manufacturing foreign competitors, sues Jilltronics for $6 million. It seems that 100 of Jill's customers were only "semi-profes- sionals" and copied some CD's onto "party tapes". Unable to afford litigation, Jilltronics files Chapter 7. An American business has been destroyed; 100 Americans have lost their jobs; tens of thousands of Jilltronics customers must now buy imported equipment, thus increasing the trade deficit. Holden Pierpont Preppie IV made his money the old-fashioned way: he inherited it. While a student at Harvard, he identifies a need for a record company to serve "discriminating classical music lovers" with recordings of young, unknown American artists: Snob Sounds. Snob Sounds's competitors are Sony/CBS, MCA (Japanese-owned), RCA (German- owned), Philips, EMI, Decca, and Deutsche Grammaphon (all European). To his competition, the $100,000 price of a Sony multitrack machine is chicken feed. However, Snob Sounds can only afford two microphones and hence shops for DAT machines. Passage of S-1623 means that Holden has to spend five times as much as he expected for his recording equipment. Furthermore, every time Holden buys a blank tape or a digital recorder, he pays a royalty tax that goes into the pockets of his huge foreign-owned competitors. Snob Sounds originally planned ten releases in its first year but can only manage three because of increased expenditures on DAT machines. Snob Sounds folds because it doesn't make a big enough initial impact. Holden goes to work for Daddy's bank and his four employees collect unemployment. The Japanese receive tens of thousands in windfall profits on Snob Sounds's DAT equipment; big, foreign record companies are protected from an innovative competitor; struggling American artists lose an outlet for their work. Jane Pollyanna comes up with a brilliant idea for making inexpensive, high-quality digital audio recorders for consumers. Despite facing competition from a Japanese hardware/software cartel with unlimited capital, Jane suc- ceeds in getting financing, building machines, hiring lawyers and accountants to comply with S-1623, and selling consumers. Her idea is so much better that even Sony and Matsushita are forced to re-engineer their products. Fortunately for them, every time someone buys one of Jane's recorders, royalty tax is paid to Sony and Matsushita's software arms. This money can be used to squelch Jane's company in many ways. With full control of hardware and software, the Japanese could simply advertise a new incompatible standard. Hit recordings henceforth would only be available on analog cassettes and this new media. No matter how many units Jane sold, software would no longer be produced for her machine. With nothing to play on Jane's machines, consumers go back to buying Japanese. Short-term Effects of S-1623 A Stone Around the Computer Industry's Neck Vast consumer markets create cheap data storage devices, e.g. audio and video recorders. This has often worked to the advantage of tiny American computer companies without sufficient capital to manufacture sophisticated mechanisms. Some of the first microcomputers used standard analog cassette recorders in place of the floppy disk drives common today. Many small American firms manufacture high-capacity computer tape drives based on consumer video and DAT transports. There is no way to distinguish between blank media used for audio and computer data storage. A tax that discourages the efficient development and use of computers in the United States is unlikely to make our economy more competitive worldwide. Unemployment in the American Digital Audio Industry Although Japanese firms will continue to dominate the consumer electronics industry, the increased popularity of digital audio equipment represents an opportunity for American firms. From an electrical engineer's point of view, there is little difference between a digital audio processor and the computer peripherals that Americans have successfully built for decades. By being creative, small American firms should be able to compete with huge Japanese firms. Economies of scale can be realized on much smaller volumes of digital audio equipment than with televisions, CD players or VCR's. Digital audio equipment can be produced in the same American factories that build computers. Most American audio equipment manufacturers are small and lack the resources to engineer custom integrated circuits (IC's). The large Japanese companies that dominate consumer electronics have ample resources to develop IC's that implement the Serial Copy Management System (SCMS) mandated by the proposed S-1623. By controlling the supply of SCMS IC's, the Japanese can control which American firms enter the market and at what retail price. After a decade building electronics in America, I have learned that, when I need a computer component, the Japanese are the world's most aggressive salesmen. When I need components critical to manufacturing consumer electronics, they either flatly refuse to sell, gently explain that "the guys in Japan will say no", don't return calls, withhold engineering data, or quote outrageous prices after weeks of delay. The recent GAO study, International Trade -- US Business Access to Certain Foreign State of the Art Technologies, confirms my personal experience. American digital audio manufacturers will be forced to choose between paying exorbitant prices to foreigners for SCMS chips or being sued for violating S-1623. Most American audio manufacturers are barely profitable-- the cost of re- engineering their products to comply with S-1623 even if SCMS IC's were free and widely available will put them out of business. Mandating SCMS or any other particular technology gives a tremendous amount of market power to mass producers, i.e. the Japanese. Furthermore, the administrative burden of filing reports, hiring auditors, studying the law, and keep- ing track of "orders of the Secretary of Commerce under $1022(b)" will fall more heavily on small domestic manu- facturers than on large foreign ones. Unemployment Among American Musicians and Record Producers While it is nice to think that Americans have a monopoly on creativity, a trip to a record shop reveals that most of the recordings are from foreign and foreign-owned firms and that many of the artists are foreign as well. S-1623 puts some money into the pockets of large foreign record companies and successful artists of all nationalities, but only by taking it from small American firms and unknown American artists. New record companies and average musicians are among the most cash-starved of all Americans. They are at a tremendous competitive disadvantage compared to established companies and artists. Recording stars earn money from concerts, movies, endorsements and licensing; some of these opportuni- ties are so lucrative that artists might do well giving away recordings free in order to get more movie and advertising contracts. By contrast, an unknown artist must produce a hit CD before any of the opportunities become available. S-1623 makes digital audio recorders capable of mastering CD's substantially more expensive, thus creating a new barrier to entry in an already concentrated industry. There will be fewer jobs for American artists and employees of American record companies. Slightly increased revenues for established record companies will not make up for the loss of variety and opportunities for newcomers. Full Employment for Lawyers Undergraduates at MIT often ask me for career advice. Since engineers make the same real salary they did in 1970, I usually tell avaricious students to become doctors. After reading S-1623, I am considering recommending law school. This bill would create a new government bureaucracy to be romanced by legal specialists with quarterly reports, annual reports, claims for payment, auditor's reports, and pleas for clemency. Simply interpreting "technical reference documents" and "orders by the Secretary of Commerce under $1022(b) (1), (2) or (3)" might constitute a lucrative business. Litigation is a sure path to riches for attorneys and S-1623 does not disappoint in this regard. Consider the case of Tiny Tunes, a small record company that mistakenly sets a single bit on a CD release incorrectly and thus violates $1021(c), rather unfortunate since $1031(d)(3)(B) provides for damage awards of up to $5,000,000. Misfortune for Tiny Tunes may be good fortune for its attorneys, who can charge any price to keep Tiny Tunes from being bankrupt by an S-1623 lawsuit. At least Tiny Tunes knows where it stands. Parts of S-1623 are so vague that nobody will be able to manufacture digital data storage devices or blank media of any kind without hir- ing an army of lawyers. In particular, the factors that distinguish a professional from a consumer unit are absurd, including such items as the letter "P" on the outside of its packaging, how it is marketed, and whether or not it has certain connectors. In practice, many professionals use consumer equipment and many consumers use professional equipment. Sony even coined a word for consumers who buy professional-quality equipment: prosumers. In the event that packaging, marketing and connectors are not vague enough, S-1623 states that a court may consider "the occupations of the purchasers of the recorder and the uses to which the recorder is put." Thus, a company may be sued at anytime because of factors entirely beyond its control and the company with the most lawyers will win. (Who can afford more lawyers, big Japanese companies or small American ones?) Consumers will be Bled; Japanese will Prosper S-1623 will force manufacturers to charge more for "professional" recorders that lack SCMS but cost about the same to produce as "consumer" recorders. Musicians, audiophiles, amateur recordists and professionals may have to pay over $1000 extra per machine just so they can go about their business. This is pure unearned profit for recorder manufacturers and will add to the trade deficit. If manufacturers got together to engage in this kind of price discrimination, they would be sued for violating anti-trust laws. With S-1623 forcing big foreign consumer electronics firms to make extra profit, consumers will have no recourse. Was Blind but Now I See (... what my Senator has done to me) Blind Americans purchase a disproportionate amount of audio recorders and blank tape. Digital audio equipment offers tremendous promise to the blind, not because of sound quality, but because of convenience, indexing, and the possibility of low-fidelity, long-playing tapes and disks. It is possible that special digital audio recorders might be designed for the blind that would not fit the definition of "digital audio recording device" in $1001(3). However, such a machine would have to use consumer media and therefore blind consumers would be paying royalty tax on media. Since blind people mostly record audio letters and talking books, it is unfair that they should have to pay a tax to foreign- owned record companies or even fellow Americans such as Michael Jackson. SCMS is a particularly nasty thorn in the side of blind consumers. When SCMS prevents a digital copy from being made, the sighted consumer notes the flashing "Congress says you can't do this" on the front panel and switches to the analog input. He then sets the recording level by watching two level meters while adjusting a knob. The sighted consumer ends up with a "nearly perfect" copy as opposed to a "perfect bit-for-bit" copy. A blind consumer cannot see the flashing SCMS indicator. Even if he did understand why the machine wasn't recording, he wouldn't be able to see the level meters and make a high-quality analog recording. S- 1623 prevents blind people from using consumer digital audio recorders for non-infringing purposes. Americans get a Cold Fish in the Face For decades, Americans have responsibly used photocopiers, VCR's, analog tape recorders and computers, all of which can be used to infringe copyright. Digital audio recorders can be used for hundreds of legitimate purposes. Congress's own Office of Technology Assessment determined that most home taping is non-infringing. No one has demonstrated any compelling need for this legislation, which robs American Peters to pay Japanese Pauls. Consider a taxpayer already reeling from the cost of bailing out the S&L industry. He walks up to his expensive new digital audio recorder, on which he has paid a tax, inserts a blank tape, on which he has paid a tax, inserts a tape of his daughter's school orchestra and tries to make a copy. The machine flashes a sign saying "your Congress decided that you couldn't be trusted with this technology." It all adds up to a cold fish in the face. Copyright for Sound Recordings The Congress shall have power to promote the progress of science and useful arts by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries. -- United States Constitution, Article i, $viii The right to not be murdered is an intrinsic right. Copyright, however, is an artificial concept created by the government to promote the progress of the arts so that society benefits. Copyright was created not to enrich authors but because it was thought that society as a whole would benefit if authors could earn more money from their creations. Successful authors often have interests at odds with those of society. For example, progress in the software development industry is being stifled by vigorous assertion of "Look and feel" copyrights, which keep innovators from competing with the likes of Lotus and Apple. Businesses cannot afford to adopt newer, superior products because their employees would have to learn completely different user interfaces. The shareholders of Lotus and Apple are well-fed, but not enough to compensate society for the sting of monopoly prices and, even worse, lost productivity due to use of obsolete products. Even a centuries-old industry such as publishing demonstrates that the strongest possible copyright is not optimal for society. Millions of copyrighted documents are Xeroxed every day. Why not tax photocopiers and blank paper? Infringement could even be halted altogether by banning photocopiers or requiring that every copy result in a FAX transmission to a central copyright bureau. Any of these schemes might increase the quality and/or quantity of authorship. However, most would agree that the costs to society would outweigh any benefits. Indeed, unknown authors are the ones who benefit most from low-cost photocopying because it has made self- publishing practical. Record companies do not have an inalienable right to squeeze every possible nickel out of American consumers. Any debate over whether to strengthen copyright for music must be decided on the basis of whether society will be better off overall. Will Michael Jackson produce better music if S-1623 increases his income by 1%? Or will we be deprived of a future Michael Jackson because an unknown artist could not afford a "professional" digital recorder? If S-1623 makes a composer slightly wealthier, will that make up for an in- creased trade deficit, lost American jobs and inconvenienced, somewhat poorer consumers? Is it worth shipping millions of extra dollars to Japan to reduce copyright infringement by a few percent (or to shift it from digital machines to cassette decks)? Should we pass laws that enrich foreign-owned record and consumer electronics conglomerates at the expense of American companies and consumers? Where's the Fire? "The most sensitive ear could not detect the slightest difference between the tone of the singer and the tone of the mechanical device," said a critic after hearing a live tenor and then a recording of the same man. Metropolitan Opera soprano Anna Case found that "everybody, including myself, was astonished to find that it was impossible to distinguish between my own voice, and Mr. Edison's re-creation of it." They were speaking not of Kiyoaki Edison, designer of DAT machines, but of Thomas Edison, inventor of the Diamond Disk phonograph. This technological wonder of 1915 was not digital but acoustic, i.e. purely mechanical with no electricity. The vast majority of people are extremely uncritical judges of sound quality. Most claims of improved sound quality amount to little more than advertising hype. Yuppies abandoned LP's for CD's because CD's are more convenient, not because of perceived higher sound quality. In many ways CD's have more distortion than LP's. The perception of higher sound quality was achieved through advertising, not engineering. Popular music so frequently copied by teenagers is particularly undemanding of recording systems. Radio stations in large cities often play music where the loudest sound is only twice as loud as the softest; the cheapest cassette recorder can hold a range of 1000 to 1. For most people, using a digital audio recorder for copyright infringement instead of a cassette recorder is about as much of an improvement as owning a Ferrari instead of a Chevy in a traffic jam. You get to pay $1000 instead of $100 for the machine, $10 instead of $1 for the tape and no one can hear the difference. Digital audio recorders are not dramatically more convenient than cassette recorders and the current generation being proposed by Sony and Philips offers lower sound quality than the best cassette recorders. Consequently, there is no reason to expect anyone except the Sharper Image set to rush out and buy these toys. After 15 years of extensive promotion, only 20% of American households contain CD players. Only a small fraction of these chose to buy one equipped with the digital output necessary to make digital- to-digital copies. Digital Audio Tape (DAT) machines, which sponsors of last year's DAT Bill were certain would entice all Americans into massive infringement, have been on the market for three years. Yet cassette-based units still outsell DAT by more than 100 to 1. The newest offerings from overseas are not even as good as DAT, recording only about 25% as many bits/second. Sound quality fanatics already own 10-year-old Nakamichi cassette decks that will outperform these latest digital gizmos. In the 16 months since I last testified before Congress on this issue, more than five new technologies capable of storing both digital audio and computer data have been introduced. By the time there are enough digital audio recorders in America to significantly contribute to copyright infringement, the multimedia revolution in computers will have made the fine distinctions among devices made in S-1623 laughable ("Is your Apple Macintosh MultiFrotz 2000 prominently marked with the letter 'P' on the front?", "Do you use your Panasonic Home Datavault more for recording nonmusical literary works, data bases, or other audiovisual works?", "Where was that 500 Gigabyte floppy disk you bought advertised? In Audio magazine? You'd better pay your royalty tax.") Why Nothing Resembling S-1623 Could Ever Work How to Circumvent SCMS or Any Other Scheme for $10 At last year's DAT Bill hearing, Mr. Leonard Feldman of the Leonard Feldman Electronic Labs testified that the SCMS system would be difficult to defeat. I testified that, in 1989, I inadvertently built a circuit that defeats SCMS and most likely any other copy restriction system. In an effort to simplify the design, only bits that affect sound quality were preserved. My circuit comprised only four chips, costing a total of under $10. Mr. George Wilson of Stanley Associates testified at the same hearing that he purposely built a device to circumvent SCMS that cost under $50 completely packaged. Both of us testified that, although an undergraduate electrical engineering background was necessary to design circumvention equipment, no specialized knowledge or components were necessary to construct such devices. In the long run, it will be possible to circumvent any copy restriction or debit card system with a single $1 "programmable logic device" chip. All a consumer would have to do is copy a program from a magazine article and spend ten minutes connecting the chip to RCA phono jacks and a $5 Radio Shack power supply. Before the decade is out, as part of the multimedia revolution, virtually all personal computers will be able to read and write digital audio. A simple 10 line computer program would then suffice to defeat copy restric- tions with the cheapest personal computer. Why Media Taxes Won't Work Media taxes won't work in the long run either. I once designed a machine that used video tape to store digital data. The same machine and tape could be used to store computer data, 80 simultaneous phone conversations, digital audio, talking books for the blind or television programs. You can buy 8mm and SVHS video recorders today that are capable of recording eight or more hours of digital audio on a single $8 tape. There is no way of determining what these machines are "most commonly used by consumers" for. Should SVHS and 8mm video tapes be taxed? How about the machines? Perhaps a few years of litigation over the intent of Congress as embodied in S-1623 might help. It will never seem fair to Americans to pay a tax on media that is primarily used for noninfringing activities. In the coming decades, every American is going to be storing, receiving, transmitting and manipulating digital data every day. Consumer digital data storage equipment will be ubiquitous. This equipment won't know or care whether the data being stored is audio, video, text, phone messages or still photographs. Less than 1% of the data will be copyrighted material that is outside of "fair use." Even if people could be convinced that a tax on blank media was fair, distributing it in the manner proposed by S-1623 is hardly likely to be perceived as fair, or to promote progress in the arts. Firstly, people often copy records that are out of print. Giving money to companies and artists who have re- cently sold records rather misses the point. Secondly, musicians and composers who need help are those who have not benefited from massive advertising, superstar promotion, movie roles, and Pepsi ads. Giving money to artists according to the number of records sold ensures that each receives an insignificant amount. Michael Jackson gets $100,000, which he might earn for a single concert; Mikael Jaacksen, Minnesota's most popular polka composer, receives $1.37, which is also what he typically earns for a performance. Finally, the "elephant in the room" is the question of why the record companies should get anything from royalty taxes. The fact that Americans are paying money to foreigners isn't so bad. Nor is the fact that users of American-made computers will be financing competing foreign computer vendors each time they buy blank media. What hurts is that record companies are part of vertically-integrated conglomerates that already make money every time a consumer makes any kind of recording. Sony and Matsushita sell blank media, every possible kind of analog or digital data storage device, and ancillary audio equipment. Philips manufactures recorders and other audio equipment, plus gets licensing fees for blank media that it standardized. Consumers who copy need to buy blank tape, a device for copying, and other devices for enjoying the copy. Unless the 12-year-old who copied a Michael Jackson song off the radio would have been willing to pay $15 for the Sony CD, Sony probably benefits from the copying because it sells media, recorders, amplifiers, and loudspeakers. Fair Long-term Compensation of Copyright Holders The most obvious way to compensate copyright holders for their efforts is the way they are paid now. The vast majority of Americans apparently think original recordings are worth $5-15 apiece. Some value the convenience of buying from a record store over borrowing from a friend or library. Some find that copying simply isn't worth the trouble. Some think that copying is unfair. Some value the booklet and other printed material that accompany the original. By capitalizing on the preceding factors and exercising some creativity, record companies will no doubt always be able to sell billions of dollars worth of original recordings. Any technological fix to compensate artists and songwriters will be easy to circumvent. Consequently, the best systems are those where the consumer realizes no benefit from circumvention. If we decide that musical performance and composition should be additionally rewarded by society, let us pay for them out of the general budget. One need then only survey consumers to find out what is being played and pay artists accordingly. We should keep in mind that consumer and society realize no benefit when a recording is copied. A songwriter should get more if his song is copied and played 200 times than if copied and played once. Automatic Surveying of Consumers Automatically surveying consumers should be straightforward. Almost all music played at home passes through a preamplifier, which is either a separate box or a circuit within a receiver. If one assumes that music is played 24 hours/day, 365 days/year, that the average song lasts three minutes and that at most 1000 billion songs need be distinguished, then a modest one megabyte of storage is necessary to store a year's worth of data on what was played. Every 12 months or so, the consumer would be reminded by the preamplifier to hook it up to a telephone line so that it could send in a report on what was played since the last report. A Census Bureau computer would determine how much to pay each artist. Is this system feasible? Yes, but it will take a few years to implement. Firstly, musical sources need to be tagged. CD's, DAT's, digital broadcasts and other digital sources are already equipped for such tagging. LP's and cassette recorders present difficulties, but the whole premise behind the clamor for S-1623 is that such analog sources are soon to be supplanted. In European countries, FM radio transmissions are tagged so that people can program car stereos to "look for some classical music". Implementing a similar scheme here would allow royalties to be paid based on radio listening and also allow consumer conveniences. Secondly, it would be necessary to insure that the system is proof against fraud. Although consumers have no incentive to defraud the system, artists do. An artist could theoretically feed bogus information to the central computer that his songs were being played hundreds of thousands of times. Public-key encryption, a technology that came into widespread use in the 1980s, would likely make it impossible for an artist to substantially corrupt the system. Thirdly, it would be necessary to insure that mandating the inclusion of specific technologies in preamplifiers does not injure American manufacturers of preamplifiers. Phasing in the system over several years would be helpful in itself. Funding a public-domain implementation of the technology would be even more helpful. The very existence of a public- domain implementation would ensure that no chip maker, foreign or American, would charge very much for survey chips. By the time musical sources were tagged and a substantial number of consumers ready to purchase digital preamplifiers, the marginal cost of adding an electronic surveying system will be minimal and certainly lower than the cost of adding a debit card reader (a scheme proposed by the Register of Copyrights last year). Plug-in modules could be employed for systems such as car stereos that are not easily connected to telephones. Note that the existence of a nationwide survey would mean that copyright holders and consumers could work together to ensure the widest possible distribution of copyrighted material. A teenager who distributed tapes of his favorite songs would be aiding the songwriters and musicians: every time one of his friends played a tape, the copyright holders would get more money. Most consumers would be happy to take a few minutes a year to call in their data since it means that their favorite artists will benefit. It is possible that music distribution will become more efficient and that, out of the $6.5 billion Americans currently pay for recordings, a greater percentage would go to artists. Note also that a survey system deals fairly with the question of compensating creators of out-of-print recordings. Only a tiny fraction of all recordings are still in print. If a media tax or debit card system is supposed to compensate record companies for taping that displaces purchases, why is it fair for a consumer to pay to copy one of the 99% of recordings that are no longer available? Yet if one decides that society should support artists whose work is being enjoyed by the public, it is perfectly natural and fair to compensate holders of copyright in out-of-print recordings. This would benefit new musicians and songwriters whose recordings may become out-of-print before achieving widespread public exposure. I am not necessarily advocating a comprehensive surveying system. As an engineer, it is not for me to say whether society should spend more to encourage musical composition and performance or whether composers and performers should get more and record industry middlemen less. However, as an engineer, I urge that the Congress not mandate half-baked technology that is destined to fail to serve artists or consumers and that will cost American jobs. Technology can be used to efficiently measure specific usage of copyrighted material and compensate copyright holders accordingly; it is painful to see an easily-side-stepped blunderbuss such as SCMS being seriously considered. Helping the Recording Industry It is not clear that Congress should do anything to artificially stimulate the recording industry. With annual U.S. revenues of $6.5 billion, it is not clear that Americans could or should be coerced into paying more for prerecorded music. Every dime that goes into CD's or royalty taxes might have been spent on something else. Now that the record companies are foreign-owned, money spent elsewhere is more likely to employ Americans than money spent on CD's or handed over to Sony, Matsushita, and Philips. However, even if one accepts the premise that the record industry is not getting its fair share of the American consumer's income, S-1623 is bad legislation. There are more obvious ways for the record industry to increase its sales than by running to Congress demanding passage of S-1623 and taxes on blank tape. Record companies could innovate. This is supposed to be a creative industry. Selling decades-old technology and then begging for government assistance is not particularly creative. CD's were designed in the 1970's and, although of- fering convenience and ease of handling, have higher distortion in many ways than LP records made in the 1950's. Millions of audiophiles worldwide continue to play vinyl LP's and put up with their shortcomings because of the CD standard's unavoidable distortion. Ford and IBM would not be very successful if they offered their 1950's and 1970's models in the 1990's. It is feasible to produce a "Super CD" containing more information and hence less distortion than 1970's CD's. Sony, Matsushita, and Philips would reap large profits as consumers purchased both Super CD's and Super CD players from them. All the currently envisioned consumer digital audio recorders would be unable to duplicate the sound quality of Super CD's and the whole issue addressed by this bill, i.e. the threat of perfect copies, would be moot. Record companies could innovate in non-technological ways. Since CD's cost so little to produce, companies could give away free CD's. Every Rolling Stones or Beatles CD would come packaged with a CD from an unknown artist likely to appeal to the same listeners. At a cost of $1 per CD, unknown artists would be introduced to millions of listeners. Copying both the featured and "freebie" CD onto, say, DAT tape would be pointless at a cost of two hours of time and $20 in blank tape. In the old days, when LP sales sagged, record companies splurged on posters, cover art and other printed material. High-volume color printing is inexpensive and hard to duplicate by consumers. Any consumer wanting the printed material would be forced to purchased the original CD. Record companies charge more and deliver less than they did in the 1950's. Why can't they try more elaborate packaging, cover art, posters, booklets, coupons for concert tickets, mini-biographies of popular artists, etc.? Finally, if their creative juices run dry, record companies could lower prices. CD's cost about the same to produce as LP records but are priced almost twice as high. There is currently no incentive to copy an in-print CD onto current digital audio media since the blank costs about as much as the CD. If prices on CD's are gradually reduced to the level of LP prices, no economy-minded person will infringe using a digital recorder for decades. Most people I know have a fixed budget for recordings and tend to spend a constant amount every time they walk into a record store: if CD's are half price, they buy twice as many. Thus, it is not clear that lowering prices would substantially reduce profits. For decades, publishers have innovated to compete against duplication technology. Book and magazine publishers have successfully responded to potential competition from photocopiers by printing higher-quality materials in color. Anyone with a personal computer and a hard disk can make a perfect copy of a $300 program in ten seconds, without even spending a penny on blank media. If one accepts the reasoning of S-1623's proponents, Microsoft would be a bankrupt shell instead of being worth $16 billion. There is no evidence that digital audio represents a unique challenge. Conclusion S-1623 is bad legislation. The Act will not encourage authorship, will hinder the use of computers in the U.S., will force American computer users to subsidize Japanese computer vendors, will create a new government bureaucracy whose mission will become undefinable with the next generation of computer data storage devices, will destroy the only realistic chance America has to get back into consumer electronics, will help create a Japanese monopoly on the manufacture of digital audio equipment, will make it more difficult for small American record companies to compete with foreign-owned giants, will injure consumers (especially blind consumers), and will increase the trade deficit. Appendix A: Software and the Trade Deficit A number of witnesses and Senators at last year's DAT Bill hearing implied that if we could reduce the trade deficit by encouraging authorship and increasing copyright holders' income. Even if every American could compose like Mozart and perform like Pagannini and copyright infringement were eliminated worldwide, we would still have to find other ways to reduce the trade deficit. Consumers have a fixed small budget for musical entertainment: exporting songs won't make up for importing cars. For every American who buys a Lexus, his counterpart in Japan must buy 3,000 CD's. The cold facts are that, as much of a symbol of American creativity as it may be, the record industry is a $6.5 billion drop in the bucket of a multi-trillion dollar economy. General Motors has more revenue in one month than the entire record industry does in one year. (Large or small, the record industry now sends it profits overseas to foreign parents. Thus, increasing record sales is not a very efficient way to reduce the trade deficit. Americans who want to share in the success of our creative community should buy stock in Matsushita, Philips, and Sony.) Appendix B: Helping Songwriters Songwriters get about five cents per song on a CD, cassette or LP. The production cost of a CD and LP is about the same, yet the record companies get twice as much for the CD. (That three companies distribute virtually all the recordings, even those from "independent labels", in the United States may contribute to the industry's success in getting consumers to pay $15 for a $1 item). Before coming to Congress because they think they aren't getting their fair share from consumers, perhaps the songwriters should try to get their fair share from the record companies. If CD's were priced the same as LP's, con- sumers would buy considerably more; record industry revenues would be comparable, consumers would enjoy larger music collections and songwriters would get much more money. Songwriters could bring down the price of CD's by starting their own discount record label, outside of the current distribution troika. Alternatively, the songwriters might negotiate a higher fee for songs distributed on higher-profit media. Appenix C: Witness Background My name is Philip Greenspun. I am in the Ph.D. program in electrical engineering and computer science at MIT and have worked full-time as an engineer or computer scientist since 1978. Prior to nestling into the groves of academe, I developed electronic and software products for large organizations such as Hewlett-Packard. After six years of working for others, I embarked on a quixotic effort to build things in Massachusetts, starting five companies since 1984 and burying two. The surviving three have grown to employ over 100 people and have annual sales of over $10 million. I have designed numerous circuits for recording digital audio, analog video and analog audio. With colleagues from the MIT Department of Electrical Engineering and Computer Science, I have physically constructed circuits that interpret SCMS-encoded digital audio signals. Classical music has been one of my passions since the mid- 1970's. My indulgences include a Boston Symphony Orchestra subscription and 2000 LP records. I was born in 1963, raised in Bethesda, Maryland, and graduated from MIT in 1982 with an S.B. in mathematics.